Svalgaard #3

Thanks to Leif Svalgaard for his continuing support of Svalgaard discussion, which is continued here (preceded by #2 here) and #1 here.

Continued here.


743 Comments

  1. jack mosevich
    Posted Jan 30, 2008 at 9:56 AM | Permalink

    I have very little knowledge of solar physics and therefore find
    it difficult to understand the implications of these threads.
    Could someone kindly summarize the implications, if any,
    in regard to solar forcing on climate?

    Thanks

  2. Posted Jan 30, 2008 at 10:57 AM | Permalink

    1 (jack):

    Could someone kindly summarize the implications, if any, in regard to solar forcing on climate?

    I would like that too :-)
    Jack, the issue here is that it is almost dogma that the sun is the cause of the climate swings the last few thousand years [LIA, MWP, etc]. This is so because we don’t know what else could do it [certainly not man-made CO2, if we except that last 100 years]. The problem is that the latest solar data seem to indicate [and this is still controversial] that the sun varies less than what we thought just a few years ago, so if we will maintain that the sun is still the culprit, then we have to crank up significantly the sensitivity of the climate to solar forcing. Most people [like Steve M] think that that is ‘impossible’. I don’t know if it is and actually came originally to this blog to find out, but it seems that few want to discuss this. I think we cannot maintain that we know what is going on if we just gloss over this problem…

  3. frank barling
    Posted Jan 30, 2008 at 11:49 AM | Permalink

    Leif

    Do you think it is still possible for the SC23 length to be over 13 years?- as a very long cycle may help settle some of the solar/climate questions.

  4. Posted Jan 30, 2008 at 11:53 AM | Permalink

    3 (frank): “SC23 length to be over 13 years?”
    I think that is unlikely, since we have already had the first spot of cycle 24. I’ll think 12.5 years at the most, but, hey, this is hard to say. A very small SC24 will go a long way to resolve many questions. The length of SC23 per se is not important, IMHO.

  5. SteveSadlov
    Posted Jan 30, 2008 at 12:30 PM | Permalink

    There is more to this than solar forcing. For example, the solar cycle will impart variations in the characteristics of the Earth’s magnetosphere and the overarching “magnetic field environment” of the solar system at large. I doubt that anyone would state with a good conscience that all the impacts of such characteristic variations are well understood. What do such variations do to cloud physics, to thunderstorm physics, to patterns of incident non solar origin cosmic radiation and matter, etc.

  6. jack mosevich
    Posted Jan 30, 2008 at 12:46 PM | Permalink

    Also, how does all of this relate to the Maunder Minimum
    and assumed link to the little ice age? After all, that occurence
    is one reason so many people believe in the solar link.

  7. Posted Jan 30, 2008 at 12:50 PM | Permalink

    5 (Steve S):

    the solar cycle will impart variations in the characteristics of the Earth’s magnetosphere and the overarching “magnetic field environment” of the solar system at large. I doubt that anyone would state with a good conscience that all the impacts of such characteristic variations are well understood.

    These variations are reasonable well understood [on par with or better than climate itself, I think]. What is not understood is the “impact” [if any]. This means that we are back to correlations and have left physics behind. Nothing wrong with that as such [if we really don't understand anything], but it was my understanding that climate scientists claimed they understood the ‘solar connection’ well enough to subtract it from the observed climate [which may be the unknown - hockey stick and all that] and to say something about the remainder. If not, it is impossible to say anything at all about what causes what. Conversely, one could subtract everything we claim to understand and try to ascribe the remainder to the ‘unknown’ cosmic causes. Either way, we have a problem.

  8. See - owe to Rich
    Posted Jan 30, 2008 at 1:08 PM | Permalink

    Re #4 Leif

    I am surprised that you say SC23 is unlikely to be more than 12.5 years long. I am, though, going by an observational rule which David Archibald often mentions, namely that the minimum nearly always occurs between 12 and 20 months after the first spot of the new cycle. By this rule, minimum should be between January and September 2009, which implies SC23 length of 12.5-13.2 years.

    Something I would like to study more is whether longer cycles correlate with longer waiting times after first spot.

    Do you believe that Archibald is correct with this, or is there some deeper solar physics which will plant the minimum during 2008?

    Rich.

  9. SteveSadlov
    Posted Jan 30, 2008 at 1:19 PM | Permalink

    Leif, excellent commentary. RE: “but it was my understanding that climate scientists claimed they understood the ‘solar connection’ well enough to subtract it from the observed climate”

    I am concerned that climate scientist so not understand all the impacts of the second and third order effects I alluded to. In fact, they generically concede that they do not understand how cloud physics affect the real world and make it different from the GCMs. I am also very concerned about: “What is not understood is the “impact” [if any]. This means that we are back to correlations and have left physics behind.” If I were the Lord of research purse strings, I would say, let us investigate the physics in these regards. I saw some work at UC Berkeley, regarding coupling / discharge between thunderstorms and plasmas further up. So, there are inklings of the needed work. But seemingly much more work left to be done.

  10. Bob B
    Posted Jan 30, 2008 at 1:21 PM | Permalink

    Leif, please comment on this paper:

    http://www.co2science.org/scripts/CO2ScienceB2C/articles/V11/N5/C2.jsp

  11. Posted Jan 30, 2008 at 1:23 PM | Permalink

    6 (jack): “relate to the Maunder Minimum and assumed link to the little ice age”

    When it was believed that TSI was significantly lower and that the solar magnetic field [HMF] was near zero during the MM, it might have been a good bet to say that there is a causal link between MM and LIA. Now, that there is a strong possibility that TSI and the HMF were not at those lows, we can only still believe that MM caused LIA if we postulate either 1) unknown other causes and/or 2) extreme sensitivity of climate to solar activity. I must have said this about 200 times in past comments. So, now we have these choices:
    a) our [new] knowledge of solar activity in the past is faulty and what we used to believe is correct [some solar scientists would take this position, as scientists are extremely conservative - nobody wants to rock a leaky boat {that may support funding and graduate students}; many cling to old beliefs years or even decades after they have been rebuked]
    b) examine the basic assumptions and calculations that led us to our earlier beliefs to see what might be improved or questioned [surely some are on shakier ground than others], or at least try to explore possible consequences of the new evidence. Is it ‘impossible’ or not? How dogmatic can we afford to be?

    Over at RC you will be lectured that Cosmic Rays cannot be ‘it’ because they show no long-term trend. Well, some cosmic ray researchers may disagree as to the trend, others will say that our proxies (14C and 10Be) show different trends or no trend and that we don’t know what the trend is as it is difficult to calibrate the proxies [not much different from climate proxies]. Solar and cosmic ray scientists are working hard to resolve some of these issues. What I ask of climate scientists is to be proactive and see if they could live with no solar trends. If not, the Earth’s climate itself becomes a proxy for solar activity and may provide useful constraints.

  12. Posted Jan 30, 2008 at 1:36 PM | Permalink

    9 (Rich): See on my website: When is Minimum.pdf (Sunspot Prediction Workshop for Cycle 24, Boulder 2007)

    The birth of the new cycle is believed to be fairly deterministic [controlled by internal dynamo processes] while the death of the old cycle is believed to be a fairly random process [Babcock model]. since minimum is determined by the interplay between and ordered and a random process, the minimum is hard to predict and may not mean much as a predictor element. Because of the random element any predictions based on strict cycles [however controlled] are not in my favor.

  13. Gerald Browning
    Posted Jan 30, 2008 at 2:01 PM | Permalink

    Leif (#12),

    Why can it not be the length of the period of relative solar inactivity
    that is crucial?

    Jerry

  14. John G. Bell
    Posted Jan 30, 2008 at 2:16 PM | Permalink

    We may also have a problem if we assume the remainder might not be due to the partly chaotic nature of the earth’s climate. Mann’s flat line climate history is a hard sell. Your flat line sun less so.
    Could it be that the period of the SCs is important to the climate in that a particular SC duration might fall into some natural cycle in our climate, such as ocean circulation, and over the course of many cycles convert a small input into a large effect?

  15. Chas
    Posted Jan 30, 2008 at 2:26 PM | Permalink

    Leif; what sort probability to you attach to the possibilty that cycle 24 might behave as Pete’s plot suggests, (in comment #454, ie having an overall TSI that is perhaps a watt/m2 less than cycle 23) ?
    And for good measure :-) what sort of probabilty would you give to cycle 25 and 26 showing similarly low integrated TSI’s ? Thanks.

  16. Posted Jan 30, 2008 at 2:34 PM | Permalink

    14 (Jerry): how would your mechanism work in detail?

  17. Posted Jan 30, 2008 at 2:46 PM | Permalink

    13 (Jerry): that should have been 13, not 14.

    14 (John): lots of things ‘could be’. How would your mechanism work? In general, I don’t think the length is important as such. I have commented on that in past comments.

    15 (Chas): I suggested to pete to make SC24 small to see what a small cycle would do. The argument for a small SC24 I have given also in a previous comment [with a nice graph, even]. A simpler argument may be that SC23 was just like SC13 and there does seem to be rough 100-year envelope [not strict cycle] to these things, so perhaps SC24 would be like SC14, too.

    Cycle 25 and 26? Statistically low cycles come in bunches and high cycles come in bunches, so if SC24 is low there is a good chance that 25 and 26 might be too. I don’t want to put a number on it, though. Dikpati and Co announced some time ago that they also had a prediction for SC25, but they have not said what it is. Hathaway predicts a low SC25 ["off the charts"]. All in all, the chance is good.

  18. Larry
    Posted Jan 30, 2008 at 2:56 PM | Permalink

    2,

    The problem is that the latest solar data seem to indicate [and this is still controversial] that the sun varies less than what we thought just a few years ago, so if we will maintain that the sun is still the culprit, then we have to crank up significantly the sensitivity of the climate to solar forcing. Most people [like Steve M] think that that is ‘impossible’.

    I won’t speak for Steve, but I don’t think he thinks it’s impossible. But people here with a control system theory background correctly point out that if the feedback sensitivity were that high, linear control system theory requires that the system will be unstable. The unstated assumption is that feedback is linear; i.e. the feedback gain is constant. The only way this this high sensitivity is possible is if the gain isn’t constant, and drops significantly as temperature increases, or in the past, when temperatures and CO2 were higher, the planet would have gone into runaway heating.

    This, of course, is exactly why the Mann hockey stick must be defended at all costs; high and constant gain can only be plausible if you make the audacious claim that the planet has never been warmer than now.

    The only other possible explanation is that there’s a significant and yet unidentified influence, which we can never rule out.

  19. Richard Sharpe
    Posted Jan 30, 2008 at 2:58 PM | Permalink

    Leif,

    Do those dips that Pete showed in posting 454 of the previous thread correspond to the Maunder and Dalton minimums? If they do one should still use caution.

    It’s a pity we do not have reliable data going back further.

  20. Jim Arndt
    Posted Jan 30, 2008 at 3:10 PM | Permalink

    Hi,

    Amazing what you can find on NASA’s websites.

    http://ulysses.jpl.nasa.gov/education/Ulysses_Wallsheet.pdf

    http://gammaray.nsstc.nasa.gov/colloquia/abstracts_summer06/presentations/Hathaway.pdf

    http://lwsscience.gsfc.nasa.gov/TRT_SunClimate.pdf

  21. Steve McIntyre
    Posted Jan 30, 2008 at 3:33 PM | Permalink

    #18. It is not obvious to me that high-energy low-entropy photons like shortwave solar radiation need have exactly equivalent forcing to low-energy high-entropy longwave infrared, the position taken by the IPCC, Hansen etc. For example, visible radiation can cause photosynthesis while IR radiation doesn’t. So equal amounts of energy at different wavelengths make a difference to a plant. Do they make a difference to a climate? Mathematically, it seems possible to me that solar could have very different sensitivity than infrared GHG forcing. I’m not familiar enough with the details to have any position on whether that possibility has any meaning in a physical context. Hansen thinks not. I don’t assume that experienced scientists are wrong; they’re probably right. I’d be interested in any formal analyses of this issue that people are aware of.

    #2. I agree that this is a very important issue and will try to spend some time on the matter.

  22. Posted Jan 30, 2008 at 3:36 PM | Permalink

    19 (Richard): Low SC24,SC25,SC26 would be classified as yet another Grand Minimum [the 'Eddy Minimum'] should it occur.

    20 (Jim): Yeah, these show pretty much the ‘party line’ and may be a bit dated [things move fast in this field]. If the solar changes were as great as depicted in the stuff from NASA we would not have much of a problem.
    Nevertheless, the presentations are well worth a visit.

  23. Stargazer
    Posted Jan 30, 2008 at 3:37 PM | Permalink

    Leif said “the issue here is that it is almost dogma that the sun is the cause of the climate swings the last few thousand years [LIA, MWP, etc]. This is so because we don’t know what else could do it [certainly not man-made CO2, if we except that last 100 years]. The problem is that the latest solar data seem to indicate [and this is still controversial] that the sun varies less than what we thought just a few years ago, so if we will maintain that the sun is still the culprit, then we have to crank up significantly the sensitivity of the climate to solar forcing. Most people [like Steve M] think that that is ‘impossible’. I don’t know if it is and actually came originally to this blog to find out, but it seems that few want to discuss this. I think we cannot maintain that we know what is going on if we just gloss over this problem…”

    Leif..For what it is worth I an very impressed by your (above) statement(a true scientist… looking). Like ‘many’ I too feel that there ‘must’ be a connection ‘somehow’ (Nobel prize for the finder…grin) and unless I read you totaly wrongly (and I do see wnat you say about the solar science)you too feel ‘in your water’ there is indeed a connection waiting to be found…after all William Herschel (astronomer, and discoverer of Planet Uranus) noted in 1801 that “when there are few spots on the Sun the price of wheat went up”. The fact that this was noted in 1801 implies that one ought not need to look too deep to find a Solar connection as he did …. BUT what is it??…. on a lighter note, It occurred to me to look at the current price of wheat, and guess what. http://www.investmentmarkets.co.uk/20070912-841.html

  24. MichaelJ
    Posted Jan 30, 2008 at 3:40 PM | Permalink

    Leif:
    Thanks so much for your patient explanations on Solar activity and the possible implications to Climate research. Only through open discussion can we find the mechanism(s) for Climate change and how they could impact man. It seems intuitive to most here that the Sun is the main driver of our weather and longer term Climate and not been given enough weight when Climate change is debated. Suffice it to say that a smaller change in Solar activity than previously thought could drastically alter our perception of our Climate history and highlights the extreme sensitivty to small changes in the Suns output.

  25. Jim Arndt
    Posted Jan 30, 2008 at 3:42 PM | Permalink

    Hi,

    #23 You might like this one.
    Solar Activity on State of Wheat Market in Medieval England

    http://xxx.lanl.gov/abs/astro-ph/0312244v1

  26. Posted Jan 30, 2008 at 3:56 PM | Permalink

    23,25: makes you wish for a bit more Global Warming :-)

  27. Jim Arndt
    Posted Jan 30, 2008 at 4:03 PM | Permalink

    Hi,

    #2 It could be not a direct influence such irradiance but an indirect cause such as the cosmic ray flux. When the sun activity is low then the magnetic field and solar wind is not as strong. This allows more comic radiation to hit the earth. The link is a little old.

    http://science.nasa.gov/headlines/y2006/10may_longrange.htm

  28. Posted Jan 30, 2008 at 4:04 PM | Permalink

    18, 24: I don’t think that there is such a hypersensitivity because if there were, we would have had a ‘run-away’ a long time ago over the ~3 billion year history of the Earth’s climate. [before that all bets are off, because conditions were just TOO different from now].

  29. Posted Jan 30, 2008 at 4:27 PM | Permalink

    28 (Jim): yeah, more cosmic rays, more clouds [but at one level, low, perhaps], higher albedo, less warming, dearer wheat, etc. But if there are no great trends in solar magnetic fields, then there would be no large trend in CRs and hence none in climate. So the question is the same: hypersensitivity? BTW, the cloud cover theory is not generally accepted and I don’t know what the current status is. Are there up-to-date measurements of this? Do they still match the CRs? Dunno. This latter reflects my low opinion of the assertion, but I’m always willing to change my mind provided good evidence shows up.

  30. Larry
    Posted Jan 30, 2008 at 4:30 PM | Permalink

    29, here’s where the control systems theory comes in: If you have a linear system, with a positive feedback greater than one, the system must run away. There are only three ways out of that: 1) the system is nonlinear and gain decreases with temperature, and will thus self-limit, 2) the system is non-linear in a way that increases gain as temperature, in which it’s quasi-stable now, but will hit what Hansen calls a “tipping point”. This implies that the earth has never been significantly warmer than now, or we’d have already gone there. 3) the feedback gain really isn’t greater than one, but something as yet unidentified is perturbing the system.

    21, I don’t think you followed the argument. The “gain” is a mathematical abstraction that has nothing to do with shortwave v.s. longwave radiation. These control system theory arguments are irrespective of mechanism, and are simply requirements of any dynamic system that feeds back on itself. If a 1 degree increase causes a subsequent 1.1 degree increase, regardless of mechanism, the system will go into a singularity (or in reality, hit some kind of limit at some point).

  31. Jim Arndt
    Posted Jan 30, 2008 at 4:39 PM | Permalink

    Hi,

    #30 This is the latest I have seen on it. Looks good but I don’t see much real world data.

    http://spacecenter.dk/research/sun-climate/cosmoclimatology/a-brief-summary-on-cosmoclimatology

  32. Posted Jan 30, 2008 at 4:43 PM | Permalink

    31 (Larry): Say it as it is. Does hypersensitivity cause run-away? as I surmised. If not, why not?

  33. Larry
    Posted Jan 30, 2008 at 4:59 PM | Permalink

    33, I think it should be obvious that if there’s a feedback mechanism that always responds to a one degree increase with a 1.1 degree increase, that after one lag time, that one degree increases causes 2.1 degrees, after two lag times, it causes 1.1*2.1 = 2.31 after three lag times it’s 1.1*2.31 = 2.54, etc. Yes, it runs away. That’s assuming, of course, that gain is fixed, and doesn’t change with temperature. I don’t think that’s a good assumption.

  34. VG
    Posted Jan 30, 2008 at 5:08 PM | Permalink

    Does anyone here know if there is any correlation data/studies between any solar parameter (TSI, Sunspots, Flux, magnetic etc) with atmospheric C02 or for that matter other greenhouses gases.

  35. Raven
    Posted Jan 30, 2008 at 5:28 PM | Permalink

    Leif,

    I curious whether you think these funding cuts for “space weather” science are an issue to be concerned about?

    http://news.bbc.co.uk/1/hi/sci/tech/7213917.stm

  36. Posted Jan 30, 2008 at 5:30 PM | Permalink

    31 (Larry): “1) the system is nonlinear and gain decreases with temperature, and will thus self-limit”

    This does not look like hypersensitivity to me.

    “2) the system is non-linear in a way that increases gain as temperature, in which it’s quasi-stable now, but will hit what Hansen calls a “tipping point”. This implies that the earth has never been significantly warmer than now, or we’d have already gone there.”

    I think it is clear that the Earth has been much warmer in the past, so we need not consider this one.

    “3) the feedback gain really isn’t greater than one, but something as yet unidentified is perturbing the system.”

    This must be what you consider the likely result, no?
    —————
    By trying the nail down the specifics, rather than to skate along the generalities, I’m trying to see where to go from here. One thing I have learned over the years is never to assume that what is obvious to me is also clear to others.

  37. Pete
    Posted Jan 30, 2008 at 5:32 PM | Permalink

    Larry(34) I was curious to ponder whether is there a perfect level of TSI which is the balance point. What if this was, as an example, 1365. Now, people often try and conclude a linear relationship between varying TSI and Global Temperatures. This was also used by F&L (2007) to show that a flat lining of TSI since 1985 could not have induced the measured GW 1985 to 2001.

    But, what if a level of TSI above the hypothetical ‘balance’ point induces a yearly increase in temperature that is out of balance and hence accumulates for as long as the TSI is above the ‘balance’. Say it was at 1365 for SC A and then 1365.1 for SC B. SC B is out of balance by .1 for a duration of 12 years. Each year it causes an increasing level of Warmth for the 12 years it is above the balance.

    This would obviously explain also why a stable, yet high, level of Solar activiy would still appear to cause an increasing level of GW.

    Anyway, someone feel free to explain a scientific reason this could not happen! I’m only pondering.

  38. Andrew
    Posted Jan 30, 2008 at 5:38 PM | Permalink

    VG, you might consider Jan Veizer’s paper here:

    http://www.sciencebits.com/files/articles/GACV32No1Veizer.pdf

    Especially figures 7 & 9
    Of course, whenever Celestial factors are correlated with temperature and CO2 is as well, it would make since for them to correlate with each other. Which makes it difficult to weed out causes, effects, and the relative importance of each factor.

    By the way, Leif, I’d be interested to here any comments you have on figures 8 & 9, seeing as this appears to qualify as large variation. The paper is a couple years old now, so how do you think your results might impact our knowledge of geomagnetic activity variation over longer time scales like those depicted?

  39. Posted Jan 30, 2008 at 5:40 PM | Permalink

    36 (Raven): Cuts in Space Weather Research:
    Yes this is very shortsighted, but we forget that the purpose of the various ‘space agencies’ is not to support Science and scientists, but to provide the aerospace industry with money, to wit this quote: “The research council argues that it will continue to fund solar-terrestrial physics satellites through its subscriptions to the European Space Agency (Esa)”. There is the same attitude in the US, where NASA pumps billions into the International Space Station that serves no [scientific] purpose. Fir the money it costs to send up a shuttle to repair yet another leak, a ground-based observatory could be run [and the data analyzed] for century. All this is well-known and there is little to be done about it, until the Chinese put a man on the Moon or on Mars…

  40. Posted Jan 30, 2008 at 6:05 PM | Permalink

    39 (Andrew): Fig. 8: just shows that the Earth magnetic field keeps cosmic rays at bay. This is not controversial and has little [if anything] to do with climate on timescale of centuries or shorter.
    Fig. 9: The caption says “Calculated intensity of solar irradiance (dots)”, but the scale [on right] says “Normalized solar modulation factor”. This is a term normally used about cosmic rays and not solar irradiance and therefore puts the paper in a bad light. [Almost caused me to stop reading right there].

    Fig. 13: Shows solar activity in the last 60 years being the highest in the last 8000 years. This is, of course, where I disagree. If solar activity was the cause of the warmings then one might expect the values around AD1000 in Figure 12 to be lower than today, since solar activity now is the highest ‘ever’. Finally, there is that debunked Friis-Christensen&Lassen graph again. So, I’m not sure the paper paints a correct picture of what is going on. But this is a big subject and there are certainly things in there that are right on. This is the problem with papers like this: the mixture of fact and ‘fiction’…

  41. Posted Jan 30, 2008 at 6:13 PM | Permalink

    31 (Larry): Say it as it is. Does hypersensitivity cause run-away? as I surmised. If not, why not?

    Why would the climate responding to solar variation be hypersensitivity? 0.1 percent is 1.37 W/M^2, 365 days a year. With a minimum, that may be more that 33 years. With CO2 doubled that may be, for the sake of argument, 6 W/M^2. Cloud cover changes may be multiple times that in the opposite direction. But the GHG is mainly air while the Solar is mainly ocean. Sol Y Mar!

  42. BarryW
    Posted Jan 30, 2008 at 6:30 PM | Permalink

    Has anyone looked at a possible correlation with particles other than photons? The aurora shows atmospheric affects but are the particle fluxes too transient or weak to impact the earth climatically?

  43. Eric McFarland
    Posted Jan 30, 2008 at 6:36 PM | Permalink

    One more tater for the soup:

    http://www.columbia.edu/~jeh1/mailings/20080114_GISTEMP.pdf

    Why is it either … or … could not both sun and gas (and other feedbacks) drive the climate? As a result, you could get some warming … or much, much more depending upon the mix. Answer seems sort of obvious … I guess.

    In any case, some times I get the impression that folks want to point to anything but gas … even believing in the “x” factor in the face of fairly clear physics and convincing circumstantial evidence … both of which point more in one direction and little to none in the other …

  44. cba
    Posted Jan 30, 2008 at 6:40 PM | Permalink

    29 Leif,

    I’m with you 100% on that hypersensitive runaway bit. There’s no way that tipping points as described can exist with high gain. Larry’s description seems right in that if there is positive feedback greater than 1, then anything could cause a minor increase and whamo – the system heads for the rail regardless. The question though of temperature sensitivity of the gain remains a significant unknown. There has to be temperature sensitivities of virtually every factor involved in the whole snarled tangled ball of twine but whether any are significant may not become apparent until that ball of twine is unsnarled.

    21 – as for Steve’s consideration on photon energy and equivalence, I posted on that previously (last thread probably about the middle) in that my suspicions are finger pointing in that direction – most especially if Leif is correct on the amount of variation for total solar energy. While I don’t think I’m onboard with that view, I will admit I don’t know enough solar theory at present to actually follow Leif’s arguments or make a realistic judgement but if time permits and the thread continues, enough may eventually soak in to get me up to some speed on it.

    I understand that while TSI is very constant in total energy, the distribution of that energy does have some significant variations – not so much the blackbody contribution directly but more the uV end which is being generated by the ultra high temperatures of the corona, evidently due to alfven waves (magnetic activity). I think the numbers are like 5 to 10% variation in the uV or some of the uV area. This is mostly past where there’s any contribution from the photosphere – although active magnetic fields are dropping photosphere BB intensity. Also, at least some consider the sun to be variable over all time frames.

    A simple example of effect would be very short uV is likely to be stopped in the upper atmosphere while long uV or blue visible is likely to hit the surface and can even penetrate into the ocean depths for significant distances – many feet, as well as scattered in the atmosphere. That same energy in red or worse in IR, could be stopped somewhere in the atmosphere, absorbed like outgoing longwave IR or it could reach the surface and be absorbed by land or sea. I don’t think the red or especially the IR is going to make nearly the distance in water and the albedo in the significant wavelength IR is substantially lower than visible light albedo.

    All this means that the energy incoming is going different places, either being absorbed or reflected at different altitudes (or depths). The example of photosynthesis is another good example of the non thermal ‘disposal’ of the incoming energy being dependent upon wavelength. There are possibly even some nonbiological actions accomplishing a similar result.

    Finally, I find the arguments and research of Shaviv and Svensmark quite interesting even if not fully complete in possible explainations. Also, Lindzens iris concept seems to fit in quite well with what is going and pointing the finger even more at clouds and cloud formation, modulated by CRs and impacted by solar magnetic field activity.

    From what little ‘playing around’ I’ve done with my pet project on radiative transfer, I’m not very convinced that ghgs actually have much of an effect, certainly not as much effect as other forcings of the same magnitude and that clouds and albedo are far more important in effect. A simple ‘back of the envelope’ calculation indicates that Mars has around 40 times the amount of co2 in the atmospheric column than does earth which is the equivalent of another 5 doublings – despite having only around 1% of the atmosphere. While low pressure absorption/emission has narrower line widths and temperatures also play a part, the simple BB calculation for Mars energy balance tends to be this massive amount of co2 GHG factors at somewhere between a loss of 7 deg. to a gain of no more than 10 deg. – depending on who’s average temperature value you use for Mars’ surface. My guess is more along the lines of roughly 0 effect ghg warming – about midway. Part of my view is the realization that absorption is related to emission for the ghg at temperature so the ‘emissivity’ rises with absorption, reducing the amount of warming necessary for reradiating energy. In this case emissivity is a function of wavelength – not a constant and is directly related to absorption.

  45. Posted Jan 30, 2008 at 6:46 PM | Permalink

    43 (BarryW): the aurorae do heat the upper atmosphere above 100 km or so. A very, very strong aurora imparts 1 TW (that is TeraWatt) = 10^12 W. The surface of the Earth is 5×10^14 m^2, so the aurora adds 0.002 W/m^2 for perhaps a few hours a few times per solar cycle. Not much. And we don’t know how to get that energy down to the surface.

  46. Larry
    Posted Jan 30, 2008 at 6:53 PM | Permalink

    Maybe I need to clarify something. Feedback doesn’t respond to forcing, it responds to temperature. Gain is dimensionless (or degrees per degree, if you prefer). Sensitivity responds to forcing, but feedback responds only to temperature. So it’s theoretically possible to have high sensitivity to forcing, and negative feedback, and lower overall sensitivity. Or conversely, it’s possible to have low sensitivity to forcing, and high positive feedback, and high overall sensitivity. What’s not possible is positive feedback greater than 1 (degree per degree). This would be unstable.

    The greenhouse theory posits a relatively low sensitivity to forcing (~1 degree C per CO2 doubling), but a relatively high feedback. That is possible, but constrains the gain to a value somewhere around 0.7 (for an overall increase of 1/(1-0.7) or about 3). If that 0.7 number were 1, the overall effect would become a singularity.

    Of course, it’s also possible that through some unknown mechanism, the solar feedback gain isn’t the same as the greenhouse feedback gain. I think this is what Steve was driving at. If that’s the case, then it’s not truly a feedback, but a markedly different sensitivity to forcing. Teasing sensitivity and feedback apart is a nontrivial task, and what Schwartz (and Lucia) was trying to do.

  47. Posted Jan 30, 2008 at 7:00 PM | Permalink

    46 (me): There is an average background auroral energy influx of ~25 GW, or 0.00005 W/m^2. One can play with the fact that the influx is concentrated in the polar regions etc, but I don’t think it has any effect overall. And there are TWO polar regions so one might double everything, but still minute…

  48. David Archibald
    Posted Jan 30, 2008 at 7:10 PM | Permalink

    Re #8 and time of minimum, the above graph is adapted from Jan Janssens’ work. Solar cyces 10 to 15 had an average of 66 months from first spotless day to month of solar minimum. For solar cycles 16 to 23, that was halved to 33 months. First spotless day for this minimum was 27th January, 2004. The accumulated spotless days for this minimum are plotted in green. So far, we are plotting on the 10 to 15 line. If this relationship holds, month 66 is July 2009, so that is my pick for the month of solar minimum. I believe that Dr Svalgaard has opted for July, 2008. July 2009 for the month of solar minimum would make Solar Cycle 23 13 years and two months long.

  49. David Archibald
    Posted Jan 30, 2008 at 7:14 PM | Permalink

    Wrong image in #49 above. Please remove it. This should be the correct one.

  50. Andrew
    Posted Jan 30, 2008 at 7:18 PM | Permalink

    Thanks, Leif. Its by a geochemist, not a solar scientist, so I can see where he got some stuff mixed up. Its also old, circa 2005. Just wanted to see if I could get a comment on some things. Thanks.

  51. David Archibald
    Posted Jan 30, 2008 at 7:27 PM | Permalink

    Re #47, above is plotted the estimate of climate sensitivity to doubling of CO2 of Australia’s most credible climate scientist, Bill Kininmonth, and the US’ most eminent climate scientist, Richard Lindzen, against that of the IPCC range, with the Stefan-Boltzmann result for reference. Based on their wealth of knowledge and experience, the eminent scientists think that negative cloud and evaporation effects will keep the rise to 0.6 degrees. The IPCC modellers go the other way. Somebody is wrong.

  52. Jim Arndt
    Posted Jan 30, 2008 at 7:35 PM | Permalink

    Hi,

    #41 (Lief)I agree with you that the MWP was warmer than today and seriously challenges the fact that the solar activity is at the highest in 8000 years. The only other explanation is a saturation point or a negative feedback kicks in.

  53. Posted Jan 30, 2008 at 7:44 PM | Permalink

    52 (David A): what I need is a similar graph for ‘solar forcing’

  54. Eric McFarland
    Posted Jan 30, 2008 at 8:03 PM | Permalink

    At least we have a warning about all of the coming solar mischief:

  55. Gerald Browning
    Posted Jan 30, 2008 at 11:27 PM | Permalink

    Leif (#16),

    We have discussed before that the relative variation in the TSI is thought to be around .1 to .3 % over 5 years. Define epsilon = .1 so that the relative variation in the TSI would be on the order of epsilon^3,
    i.e. in scaled terms the TSI perturbation would look like
    H = H_0 + epsilon^3 H_1 where H_0 is the TSI from the standard solar constant. Then this variation would not be seen by any climate model that has relative continuum and numerical errors much larger than the variation of TSI, especially in the tropics where those relative errors are O(1).

    However, it has been shown mathematically that the vertical component of velocity w is in balance with the total atmospheric heating near the tropics. This is because the scaled equations for slowly evolving atmospheric motions (not sound or gravity waves) with a time scale of a day or so have a large coefficient, e.g. epsilon^(-2),
    in front of the vertical velocity w and total heating term Q in the
    potential temperature equation, i.e. the two terms must cancel to order epsilon^2 for slowly evolving solutions. Thus for balanced solutions it must be the case that w = Q + O ( epsilon^2 ).
    Now consider the perturbation equation for the TSI
    variation in H of O(epsilon^3). Because the variation of H (epsilon^3 H_1) is multipled by epsilon^(-2) and epsilon^3, the variation in the forcing
    of the perturbation system is unlikely to be seen for a period of a year or more if we use the standard integral estimate for the external forcing.
    Thus the impact would not be seen instantaneously and the longer the period of time of the variation, the more likely the impact on the vertical velocity (the band of storms near the equator).

    These arguments can be refined if the concept fits the pattern. :-)

    Jerry

  56. J. Marshall Lancaster
    Posted Jan 30, 2008 at 11:32 PM | Permalink

    http://www.worldclimatereport.com/…/neptune_temps.JPG

    tamino.files.wordpress.com/2007/01/solar3.JPG

    Pete #38 I agree with your general contentions and I too am suprised not many are seeing, sustained record high TSI as the GW culprit. I believe,and I refer you to the second chart, you will see a divergence in the last part of the century. What I see is, the high TSI baked the earth and over time caused the increase,like a baking oven. Others look and see a lack of correlation here. The first chart of course shows there is a correlation after all.

  57. Posted Jan 30, 2008 at 11:46 PM | Permalink

    56 (Marshall): but there is no sustained record high TSI, so what gives?

  58. J. Marshall Lancaster
    Posted Jan 31, 2008 at 12:02 AM | Permalink

    Lief #57

    Sorry I ment 8,000 year high,never the less the same principle of high TSI having a baking effect remains. I believe this and other anomolies about the sun will be seen to be the culprit of GW. I am not a AGW skeptic I am only stating its degree is still after all these years not known.

    Lief, with TSI dropping and the delay of sc24 what do you think is the probability of a minimum ? Do you think very many climate scientists and such will change there views at that point ?

  59. Posted Jan 31, 2008 at 12:03 AM | Permalink

    57 (me): I should have supplied a graph. Look at the last page of
    http://www.leif.org/research/TSI (Reconstructions).pdf

    The increase from 1900 to 1950 did not happen in TSI. so, no correlation.

  60. bender
    Posted Jan 31, 2008 at 12:11 AM | Permalink

    #46 Keep plugging away, Larry. They’ll get it yet. Hypersensitive in the low end, loss of gain in the high end due to nonlinear feedback. Control system theory. Lots of gain when you need it, but ultimately no runaway. What’s the mechanism? No one knows. Daytime clouding + afternoon rain + night-time clearing? Study those clouds.

  61. Posted Jan 31, 2008 at 12:13 AM | Permalink

    58 (Marshall): there is no 8000 year record high either, but let that slide. I think you ask if there is possibility of a new ‘Grand Minimum’. Yes I do think so. Will people change their mind? I’m not so sure about that.

  62. stargazer
    Posted Jan 31, 2008 at 2:53 AM | Permalink

    Leif you said “I think you ask if there is possibility of a new ‘Grand Minimum’. Yes I do think so.”

    I am impressed again, you are a famous solar astronomer and the world should

    here is a letter I sent to a newspaper on this very thing

    The term ‘Climate Change’ was until recently used to describe ‘Natural Warming’ over time… and ‘Global Warming’ was used to describe Human caused warming (AGW).the term ‘Climate change’ has been hijacked by the AGW’ers. And it makes me smile when I get described as a ‘Climate Change Denier’ Why??? because climate change is the very thing that I (and many scientists) KNOW has happened time and time again over hundreds, thousands,and millions of years. The climate changes… that is what is does…Its the AWG’ers who go out of their way to ‘pretend’ the climate has ‘flat lined’ until humans came along and ‘made’ it change. The AGW’ers are the real denier’s, they get rid of the ‘Roman’ warm period.the ‘medieval’ warm period, the little Ice age. and call what’s left the ‘hockey stick’ graph.. this is nonsense!!!!

    WHAT if our civilization (exactly as it is now) had occurred during the Holocene Climatic Optimum. when the climate WAS warmer than today (no-one really disputes this) would ‘you’ complain that ‘we’ had warmed the climate by an extra one degree f…. And what if you could know that in some ‘alternative universe’s future civilization’ the temp was STILL lower than yours (despite the supposed added man caused C02.warming ) And what if our civilization had occurred in the LIA…I’ll bet No-one would complain we had warmed things up a bit… This is the point…I think that we are in for a period of cooling.. and this will be a disaster that will make the worst warming predictions look good…

  63. stargazer
    Posted Jan 31, 2008 at 2:56 AM | Permalink

    missed words

    #’and the world should know this fact’…

  64. See - owe to Rich
    Posted Jan 31, 2008 at 3:14 AM | Permalink

    Leif, re #17 and not believing in importance of cycle lengths.

    Have you read my paper at CA Forum->Surface Record yet? (I did link it on this thread, but for the first time I’ve been snipped!)

    Anyway, with an R^2 of 0.87 for solar length and CO2 combined, I am conviced that solar length is important. However, I do say that it should only be a proxy. What we want solar physicists, such as yourself, to do, is to work out how that proxy works. Then we should get an even stronger correlation, and be able to throw the solar cycle length data away.

    Does that sound reasonable?

    Rich.

  65. Pete
    Posted Jan 31, 2008 at 3:33 AM | Permalink

    Leif, your comment “The increase from 1900 to 1950 did not happen in TSI. so, no correlation.” is the exact point I was trying to get across! If you angle your statement to “Did the constant but higher than a balancing point value of TSI 1900-1950 cause a decadal increase” – is the point I was trying to state.

    How can I rephrase it. If you put your foot on the accelerator in your car to Maximum, the physical position of your pedal is the same even though you are constantly increasing in speed with respect to time. The actual physical position of your pedal is not directly related to a specific speed.

    With TSI, people always try and say a specific value in TSI must equal a specific temperature on Earth. But why?

    I tried anyway!

  66. Louis Hissink
    Posted Jan 31, 2008 at 5:41 AM | Permalink

    Dear all

    The solar connection is not principally the radiative component that drives the diurnal variation, but the underlying plasma one driven by galactic Birkeland currents etc.

    Sources here http://public.lanl.gov/alp/plasma/papers.html

    The earth’s background temperature is dominated by the thermal state of that solid bit under your feet, not the ephemeral gas we breath.

    As the discovery of continental crustal rocks outcropping in hitherto assumed spreaded ocean floors, (see http://www.ncgt.org, latest issue [subscribe to get it or email me for a copy])has dealt a death blow to plate tectonics theory, so also to the underlying basis for ignoring the solid earth in modelling its thermal state in climate studies. Volcanism is not the result of plate tectonics but another ignored source of energy.

    As Peratt has noted on P2, of http://public.lanl.gov/alp/plasma/downloads/PerattAntiquityZ.pdf, millions of amperes of electrical current leave and enter the upper atmosphere of the earth (see his Fig 2 cartoon). Electrical currents must complete circuit and they so through the earth.

    Guess what? Electric currents pasing through resistive loads generate heat, the thermal background that the dirurnal solar radiation varies about, even on a decadal sense, and perhaps longer.

    Birkeland proposed this EM connection 100 years ago, so it isn’t a new theory, but an ignored theory, and in its case, verified experimentally by measurement.

    Think about it.

  67. Louis Hissink
    Posted Jan 31, 2008 at 5:57 AM | Permalink

    #66

    Some grammatical errors, third last sentence, and it’s not the solar radiation the fluctuates, but the diurnal variation of it…..(just in case a nit-picker wants to pick nits).

    And I seem to have omitted a “do” in front of a “so”.

    So there :-)

  68. Jonas Birgersson
    Posted Jan 31, 2008 at 6:17 AM | Permalink

    #46 (Larry)

    Yes, a positive feedback greater than 1 would lead to a run-away climate by any forcing.
    So, this is obviously not the case.

    An unknown effect (associated with solar activity) explaining LIA does not require a high feedback because that effect will work as a forcing on the climate, not as a feedback.

    It seems to me that the control theory argument against cosmic rays explaining past climate is thus invalid.

  69. Louis Hissink
    Posted Jan 31, 2008 at 6:23 AM | Permalink

    #68

    Cosmic rays are essentially charged particles in motion.

    Some of us call that electricity, so perhaps there is some confusion of a terminological kind between astrophysicists and electrical engineers?

    Oh and space does transmit electricity….but we all knew that.

  70. Hector Pascal
    Posted Jan 31, 2008 at 7:10 AM | Permalink

    #46 Larry
    An observation from a long time lurker and weather freak from the Top End of OZ: the diminishing feedback mentioned by Larry as mechanism #1 has to be clouds insofar as the tropics are concerned. Any one who has lived in the wet/dry tropics for a few seasons will tell you that clouds drive temperature and not the other way round.
    That is, in the wet season, clouds = cool as in when the monsoon is here; Tmax = 29/30 C, Tmin = 27/28 C and SST = 30C.

    During the dry, clouds = warm as in Tmin climbs closer to mid 20s and SST = 23C.

    In the transition from beautiful dry season to the wet, humidity builds from 30 odd % to about 70 ish % (won’t argue any corrections). The common expression is that a hot wet blanket is thrown across the Top End in the “build up” (local colloquialism). From there on, it is variations on a theme as clouds wax and wane according to the Madden-Julian Oscillation and random local convection. Throw in El Nino and La Nina and you have got a perfect suite of drivers for temperature.

    Incidentally, I have a graph of levels in Darwin River Dam dating back to the late 70’s that shows an interesting picture of catchment inflows (integrating rainfall across an area that smooths out event “noise”). There are some almost decadal cycles in dam inflows, but there seems to be no match to solar cycles as far as I can see.
    OT, I have examined Bureau of Meteorology (BOM) and GISS Darwin temp records (dating back to 1880’s) and there are some interesting “corrections” to Darwin temps by BOM. I will post on an appropriate thread if there is any interest.

    Google Hector the Convector for an interesting perspective on tropical thunderstorms that daily punch up to, and sometimes beyond, 18 km. I used to work under Hector and appreciated the daily afternoon “solar umbrella” he created at the tropopause.

  71. Posted Jan 31, 2008 at 7:48 AM | Permalink

    64 (Rich): The solar cycle length is a [poor] proxy for solar activity itself. Try your correlations again but use the size of each cycle instead, filtered the same way.

    65 (Pete): TSI translates into temperature because the Earth eventually will have to re-emit the energy received from the Sun. I’m sure this has been discussed ad nauseam elsewhere in this blog.

    66 (Louis): The Birkeland currents penetrate down to 100 km height, where they turn around and return back up the field lines. So, these currents appear as adjacent sheets: one down, the other one up. They do not penetrate to the surface and cause heating there.

  72. Pete
    Posted Jan 31, 2008 at 8:29 AM | Permalink

    Leif, Good answer. The word you use ‘eventually’ is the buffering effects of various systems, and the time it takes for Earth to reach ‘equilibrium’. If it ever does. Some think the Earth is always striving to reach a point of equilibrium but can never keep up. From what I understand, there is a fairly poor understanding of the processe involved in this and the buffering? Unless you invoke the ‘Black Body’ theory, which as we may know means Earth has no Albedo, which is ridiculous. A conundruum really. More like a ‘Greyish Body’ but with some tweaks and caveats.

  73. cba
    Posted Jan 31, 2008 at 8:48 AM | Permalink

    72 Pete,

    the BB theory isn’t as bad as you’re assuming. Earth has a nice 30% or so albedo that is primarily in the visible which is where there is essentially no BB emissions. Once you go down into the IR, you’re in totally different territory for reflectivity versus absorption. One doesn’t need to travel far at all in that direction to start hitting 96-98 % absorption from just about any surface material of any consequence. Most of the albedo reflections are coming from clouds as most of the earth’s surface reflectivity is significantly below 30% to begin with. Throw in the wavelength change for important albedo vs emission and you’re way into excellent blackbody territory.

  74. Larry
    Posted Jan 31, 2008 at 9:15 AM | Permalink

    60 (bender), you’re getting my point. Even if we understand nothing about mechanisms, certain things have to be true of the mathematical description of the overall system. Certain combinations of parameters can be ruled out. The original point of this thread (in its first incarnation) was the ramifications of the hypothesis that solar output has been a lot more constant than we may have thought. You put that together with what we know has to be about overall stability criteria, and we get a very puzzling picture. Something very strange and unintuitive has to be.

  75. Richard Sharpe
    Posted Jan 31, 2008 at 9:32 AM | Permalink

    Hector,

    Were you there on the evening of December 24, 1974?

  76. Posted Jan 31, 2008 at 9:41 AM | Permalink

    75 (Richard): stay on topic, please.

  77. Bruce
    Posted Jan 31, 2008 at 11:27 AM | Permalink

    The Canadian Space Agency’s radio telescope has been reporting Flux Density Values so low they will mean a mini ice age if they continue.

    Like the number of sunspots, the Flux Density Values reflect the Sun’s magnetic activity, which affects the rate at which the Sun radiates energy and warmth. CSA project director Ken Tapping calls the radio telescope that supplies NASA and the rest of the world with daily values of the Sun’s magnetic activity a “stethoscope on the Sun”. In this case, however, it is the “doctor” whose health is directly affected by the readings.

    This is because when the magnetic activity is low, the Sun is dimmer, and puts out less radiant warmth. If the Sun goes into dim mode, as it has in the past, the Earth gets much colder.

    Tapping, who was originally from Kent, says that “Typically as you go through the ten or eleven year solar activity cycle you see the numbers go up or down. The lowest number is 64 or 68. The numbers 71 or 72 are very low, but they usually start to go up. We are at the end of a cycle, but the numbers still haven’t gone up. We have been joking around coffee that we may be seeing the Sun about to shut down.

    http://www.britsattheirbest.com/001645.php

  78. Posted Jan 31, 2008 at 11:39 AM | Permalink

    77 (Bruce): The flux has not started to go up because we are not at minimum yet. Once we are past minimum, the flux will go back up.

  79. Gerald Browning
    Posted Jan 31, 2008 at 12:20 PM | Permalink

    Leif,

    Did I lose you on #55 or are you still contemplating the statements?
    I will be happy to expand on any of the statements if you want more clarification.

    Jerry

  80. Bruce
    Posted Jan 31, 2008 at 12:25 PM | Permalink

    What did the Flux do during the “minimums”?

  81. RCB
    Posted Jan 31, 2008 at 12:28 PM | Permalink

    Interesting article at: http://www.ucar.edu/news/releases/2006/sunspot.shtml
    Not sure if it was mentioned before.

    Your work is excellent, sir. I have shared your website with many friends who are interested in this subject.

  82. Posted Jan 31, 2008 at 12:33 PM | Permalink

    79 (Jerry): No, you did not lose me on #55. I hear you, just have not connected the dots.

    80 (Bruce): the flux measurements started in 1947 and at each ‘regular’ sunspot minimum it always goes down to about 65. We have not had any flux measurements during a ‘Grand Minimum’ [like the Maunder Minimum], but I would expect it also to be stuck at 65 at such times. This statement is, of course, the crux of the matter, namely that my contention is that at every minimum [Grand or not] we always reach the same ‘ground state’ of the Sun. Not everyone agree with me on this [yet :-) ].

  83. Mark
    Posted Jan 31, 2008 at 12:41 PM | Permalink

    #81

    That article you referenced is way out of date. More and more solar physicists are now predicting a weaker SC24. Here’s a more up to date forecast.

    http://www.swpc.noaa.gov/SolarCycle/SC24/Press_Briefing_Biesecker.ppt

    Haven’t heard much in the way of further official updates but SC23 is dragging on . . .

  84. Posted Jan 31, 2008 at 12:42 PM | Permalink

    81 (RCB): Thanks for the kind words and your interest. The post you refer to is, of course, the HAO-prediction, which we are putting to the test for cycle 24. Too early to call yet, but my own take is that the prospects don’t look too good, but we must let the Sun tell us. There could always be surprises in store when we don’t know well enough what is going on.

  85. See - owe to Rich
    Posted Jan 31, 2008 at 1:03 PM | Permalink

    Re #12, Leif’s “when is minimum”. I looked at this and found it interesting, but not compelling. I prefer David A’s two statistics, namely length of time after first new sunspot, and graph of cumulative spotless days (see #49), which I have been plotting myself. But we shall see…

    Re #30, the possibility of non-linear gain. This is an argument that I make at Discussion point t. of my
    article on the CA Forum (where I’d love to see more comments if you cared to drop by). Reproduced here, it is:
    A point which is independent of the modelling here is that of non-linear response to forcings. It is sometimes asserted that sensitivity will increase with warming. Actually, I think it is the reverse. As it gets warmer and snow and ice caps shrink, the albedo certainly goes down because there is less snow reflecting solar rays. But the regions in which this change happens get nearer to the poles, where the quantity of solar power to be reflected is much smaller. Conversely, as ice caps grow to lower latitudes, for a given extra area they are reflecting more solar power. This is presumably why ice ages can occur, but it would be wrong to conclude that melting ages can occur in the same way.

    Re #43, Eric M seems to want both “sun and gas” to be considered. He should definitely read my article…

    Re #51, David A plotting votes for CO2 sensitivity. Please add a marker for me at 1.4! (Again, from my article, and slightly lower values would not surprise me.)

    Rich.

  86. Gerald Browning
    Posted Jan 31, 2008 at 1:59 PM | Permalink

    Leif (#82),

    Please let me know how I can help fill in the dots. I know that I left out a number of details and am willing to fill in with more explanations and refs. Some of this goes back to discussions I had with Steve M about
    movements Steve thought he saw in the cloud bands near the equator over time (see intro on another thread).

    Jerry

  87. Posted Jan 31, 2008 at 2:10 PM | Permalink

    86 (Jerry):

    Now consider the perturbation equation for the TSI
    variation in H of O(epsilon^3). Because the variation of H (epsilon^3 H_1) is multipled by epsilon^(-2) and epsilon^3, the variation in the forcing of the perturbation system is unlikely to be seen for a period of a year or more if we use the standard integral estimate for the external forcing.

    For someone not intimately involved with these ‘standard’ matters I need quantification of unlikely [10%, 1%, 0.00001% ?] and of ‘year or more’. Where does that time scale come from? In my field [almost] everything can be quantified by a ‘back of the envelope’ calculation that makes a claim at least plausible. This is what I need here.

  88. beng
    Posted Jan 31, 2008 at 2:42 PM | Permalink

    One of Pielke’s recent papers here shows some impressive albedo changes in E. U.S. since ~1650 AD, tho I think he means a decrease in W/m^2, not an increase, for increasing albedo. From just looking at the albedo charts, there seems to be at least 15 W/m^2 decrease in the net solar energy absorbed over the period — compared to 3.7 for CO2 doubling.

    The albedo did generally decrease from 1920 to 1992, so that coincides w/20th cen warming.

  89. Gerald Browning
    Posted Jan 31, 2008 at 3:34 PM | Permalink

    Leif (#87),

    If the TSI variation were too small, then the integral estimate (involves maximum of the variation over the period of time) would
    indicate that the TSI variation would play little or no role over the time period. But that is not the case here because the TSI variation,
    though small, is multiplied by a correspondingly large coefficient in the perturbation system forcing near the tropics.
    Thus the TSI variation cannot be discounted and can even lead to a linear
    increase in its impact during the time period. This is also the reason
    for the importance of the length of time of the variation. I did this calculation some time ago and as I recall the period of time was on the order of 5 years. Also you might look at the discussion on the Possible ITCZ Influence thread.

    Jerry

  90. Gerald Browning
    Posted Jan 31, 2008 at 3:49 PM | Permalink

    Leif (#87),

    These are not back of the envelope calculations. They depend on
    known mathematical estimates for hyperbolic (or incompletely parabolic)
    PDE systems and considerable work on slowly evolving solutions
    of the atmospheric equations in manuscripts by Heinz Kreiss and me.
    As long as the TSI variation is no smaller that we discussed,
    it can have an influence on the ITCZ, especially over a longer time period.
    That time period can be narrowed down using the above info.

    Jerry

    Jerry

  91. See - owe to Rich
    Posted Jan 31, 2008 at 4:50 PM | Permalink

    So, I was all set to give up the blog for the night. But then I took a bath, and pondered Leif’s advice at #71 to study solar cycle sizes rather than lengths. Well, I had just looked at the Wolf data on wikipedia, but I know that others have struggled to correlate those with temperatures – which isn’t to say I won’t try in the future. Can anyone point me to concise data files for sunspot numbers, which would be better than looking at graphs?

    So I came back to thinking about why cycle length might be important, despite Leif’s protestations. Why are we seeing 50-year records for cold broken this autumn and winter (though not in England, yet)? It can hardly be due to the peak size of Cycle 23, which around a Wolf number of 150-160 was not at all shabby. Surely it has to be connected with the late onset of Cycle 24, whose first spot (Jan. 2008) arrived about 3 years later than it should if Cycle 23 had been sub-10 years like Cycle 22. Well, that’s just a theory, but on the blog I see a good deal of support for it.

    How would this work? Well, we don’t really know how the sunspots drive the climate – it doesn’t look like TSI alone is enough. But people surmise that they might affect the weather and the albedo.

    Why might timing be more important than quantity of sunspots? Is it possible that a new cycle, with the change in polarity, somehow invigorates the climate? In other words, is it possible that the ramp up into the new cycle has a more profound effect than the stale old spots of the preceding cycle? If so, then delaying the cycle may delay this warming effect, hence the Earth cools. Or, it could be that a small integral of sunspots over an extended minimum is the important point. But we should bear in mind that Cycle 19, peaking in 1957, was ultra strong, yet did not apparently push temperatures up – but at 10.7 years that cycle was the longest since Cycle 14, fiftysome years earlier.

    All we would need now is a way to make this into a falsifiable hypothesis. Possibly Leif will consider this idea just a bit too Rich.

    Rich.

  92. Posted Jan 31, 2008 at 5:08 PM | Permalink

    91 (Rich): http://sidc.oma.be/sunspot-data/

  93. Posted Jan 31, 2008 at 5:11 PM | Permalink

    89 (Jerry):

    TSI variation, though small, is multiplied by a correspondingly large coefficient in the perturbation system forcing near the tropics.

    What coefficient? Where? gotten how?

  94. Posted Jan 31, 2008 at 5:14 PM | Permalink

    90 (Jerry): send me the G.L.Browning and H.-O.Kreiss paper on this, please.

  95. Gerald Browning
    Posted Jan 31, 2008 at 5:42 PM | Permalink

    Leif (#94),

    What manuscript are you referring to. The one for all scales of motion near the equator or the one for mesoscale storms in the midlatitudes (assuming you have looked at the ITCZ thread)? The large coefficient appears in both.(Note that neither discusses solar impacts. I have done that on my own.) I have reprints, but no electronic versions. I can provide references
    that can be obtained electronically thru online electronic journals or thru
    your university library.

    Jerry

  96. Posted Jan 31, 2008 at 5:58 PM | Permalink

    95 (Jerry): Here is my problem: if somebody asks me why the TSI has a ‘large coefficient’ I wouldn’t know what to say [and I would like to]. You know best what manuscript is best for me. To get an electronic reprint you go to a Kinko’s. For a few dollars they will scan the reprint you have and email it to you.
    When I try to access any of your JAtS papers I get this:

    Server Error
    We’re sorry. We could not locate the resource you requested. If you believe this is in error, please contact us at helpdesk@allenpress.com. Be sure to include the complete URL you were attempting to access when you encountered this error.

  97. a scientist
    Posted Jan 31, 2008 at 6:31 PM | Permalink

    32 (Leif)
    Does hypersensitivity cause run-away?

    I don’t think your question was fully and clearly answered, although I did see attempts by Larry (33), captdallas2 (41), and bender (60).

    Hypersensitivity may result in large excursions, but it does not cause what control theory calls runaway. At the risk of oversimplifying, let TE be Earth’s temperature anomaly, S the insolation anomaly (maybe including magnetic impact), and let G be the greenhouse gas anomaly. A control theory equation would be

    d(TE)/dt = -A * TE + B * S + C * G

    Hypersensitivity to insolation would be a large value of B. That could drive TE to a large value, but it could stabilize there, provided only that A > 0.
    Runaway can happen only when the value of parameter A is negative. The sign of the parameter A is determined by the feedbacks in the climate system, and has little to do with B, the sensitivity to solar forcing. Those who propose positive feedbacks within the climate system (via water vapor and clouds, for example) are suggesting in effect that A is negative; Lindzen (among others) finds that implausible. (I believe that parameter A is one minus the gain, but I’m just a dynamical systems guy, not a control theorist, so perhaps Larry can help me on this.)

  98. Posted Jan 31, 2008 at 6:44 PM | Permalink

    A scientist– Out of curiosity, when climatologists use “positive feedback”, do they actually mean A goes negative? Or do they mean there are physical mechanisms that lower the value of A?

  99. Larry
    Posted Jan 31, 2008 at 7:13 PM | Permalink

    98, He’s defining “A” as the negative of what I’ve been calling the “gain”. So when the climate scientists talk about a “positive feedback”, yes, that would make “A” negative. The usual situation in a control loop is to have the feedback be negative, specifically so that the loop stabilizes.

  100. Posted Jan 31, 2008 at 7:49 PM | Permalink

    Larry–
    I guess I’m puzzled by terminology. The steady solution (aka equilibrium) for that equation is:

    -A * TE + B * S + C * G =0

    or
    TE = (B S + C G)/A

    TE is supposed to be the equilibrium temperature, right? The constants B, S, C, and G are all supposed to be positive. And the “extra” temperature with greenhouse gases is supposed to be:

    TE = CG/A

    Now, here’s why I’m puzzled.

    Even climatologists who talk about positive feedback always suggest the equilibrium temperature is positive, but larger than without feedbacks. That’s consistent with A remaining positive, but just being smaller than expected without feedbacks.

    Aside from all the runaway business and lack of stability etc. associated with negative values of “A”, you really can’t describe any sort of equilibrium temperature with a negative value of A.

    Describing it, and saying you predict say, 1.5C without feed backs and 4.2 C with feedbacks implies “A” is positive and remains so whatever amount of feedback they say causes the equilibrium temperature to be 4.2C with a gain.

    So…. it seems to me they aren’t using this quite the same way we use it in control theory.

  101. Posted Jan 31, 2008 at 7:57 PM | Permalink

    100 (lucia) lucia, Larry, and others: please resolve this issue as I’m very interested in this [and welcome, lucia] and want to be able to explain it to others [therefore must understand it myself first].

  102. Raven
    Posted Jan 31, 2008 at 8:17 PM | Permalink

    Roy Spencer mentioned that the modellers usually omit the Planck response when discussing feedbacks even though it is included in the models:

    http://www.climateaudit.org/?p=2571#comment-195584

  103. a scientist
    Posted Jan 31, 2008 at 8:23 PM | Permalink

    I was just trying to nudge this toward a mathematical analogy, to show that hypersensitivity and runaway are separate issues. I may not have chosen the best equation for understanding how feedback relates to runaway. I would welcome any suggestion of a more suitable equation or mathematical analogy.

  104. John Lederer
    Posted Jan 31, 2008 at 8:51 PM | Permalink

    Would this press release of a paper which suggests that volcanic activity has a marked effect on the climate effects of solar activity be of any help to the discussion?

    http://www.unisci.com/stories/20022/0613022.htm

    The scientists started out with the assumption that a low level of cosmic rays on Earth resulting from high sunspot activity would lead to less cloud cover and less rain, with resulting high dust levels.

    “This was true for the first three or four solar cycles we studied, from about 1930 to 1962, but then the correlation reversed itself, demonstrating that the mechanism couldn’t be what we thought,” said Ram.

    It turned out that during those 32 years of positive sun/dust correlation, there was relatively little explosive volcanic activity worldwide. The researchers found that the same conditions existed between 1860 and 1882. Each of these relatively “quiet” periods came to an end with increased volcanic activity.

    For example, in 1883, the Indonesian volcano Krakatau erupted in one of the deadliest volcanic disasters, killing 36,000 people. At exactly the same time, the data started to exhibit low dust concentration whenever there was high sunspot activity, a correlation that violated the scientists’ original assumptions.

    “By carefully studying the timing of other volcanic eruptions, we found that they coincided with all of the correlation reversals between sunspots and climate,” said Ram.

    A chart in the paper shows how six major volcanic eruptions between 1800 and 1962 occurred during precisely the same years when there were reversals in the correlation between sunspot activity and climate.

    That revelation provided a further insight into how sunspots affect climate.

    “All energy comes from the sun, but the change in visible radiation from the sun during any one solar cycle is less than one half of a percent,” explained Stolz. “Scientists have said it’s impossible that so small a change could influence any signal in the climate. But here we have evidence to show that it’s not just radiation energy from the sun that is affecting climate, it’s the solar-modulated cosmic rays that have a strong influence because of their impact on cloud cover.”

    With fewer clouds, and therefore less rain, the scientists reasoned, maximum sunspots should cause levels of atmospheric dust to rise.

    “That is true sometimes,” said John Donarummo, Jr., UB doctoral candidate in the UB Department of Geology and a co-author on the paper.

    But, the researchers discovered, during periods of high volcanic activity, high sunspot activity also results in high levels of atmospheric dust.

    According to Donarummo, it long has been known that volcanoes add more dust and more sulfates to the atmosphere.

    The UB team discovered that these additional sulfates cause cosmic rays to have a more pronounced effect on Earth by spurring the formation of small droplets in the atmosphere that, in turn, cause the formation of a type of cloud that does not produce rain.

    “During these times of high volcanic activity, the sunspot/climate correlation reverses and dust levels rise, even in the absence of high sunspots,” explained Stolz.

    I think this is the paper:

    http://www.agu.org/pubs/crossref/2002/2002GL014858.shtml

  105. Pat Keating
    Posted Jan 31, 2008 at 9:15 PM | Permalink

    Feedback

    A linear system has vout = g* vin.
    If we add feedback,
    vout = g* vin + a* vout

    Thence vout* (1 – a) = g*vin
    or
    vout = g*vin/(1-a)

    If the feedback coeff is negative, the gain is reduced.
    If it is positive, say 0.9, the gain is increased tenfold. If it is greater than 1, you get runaway.

  106. Gerald Browning
    Posted Jan 31, 2008 at 9:37 PM | Permalink

    Leif (#96),

    If you go to the American Meteorological Society and do a search on articles by Kreiss under JAS the following manuscript appears:

    The Role of Gravity Waves in Slowly Varying in Time Tropospheric Motions near the Equator
    G. L. Browning, H.-O. Kreiss, and W. H. Schubert
    Journal of the Atmospheric Sciences
    Volume 57, Issue 24 (December 2000) pp. 4008–4019

    A hard copy can be obtained from this issue in your library or
    on line if your library subscribes to the electronic version of the journal. The manuscript contains references to the midlatitude case.

    I had kinkos make a copy of Sylvie Gravel’s draft manuscript and it was very badly done (skewed lines).

    Jerry

  107. Posted Jan 31, 2008 at 9:49 PM | Permalink

    With no offense to the control theorists, which cost me a ton of money with adaptive feed forward controls a couple of decades ago, the universe is simple. The solution boils down to the major players, the sun and the sea. Solar variation is only 1.4 Watts per Meter squared. With an eleven year cycle, the impact on climate is often lost in the noise. The noise is created by decadal oceanic oscillations that may or may not be synchronized. Unsynchronized, solar influence is somewhat evident, pre-1980. Synchronized, the solar influence is buried in the chaos that is climate.

    During a solar minimum, the 0.75 watts per meter squared less incoming radiation (assuming half) sounds insignificant. But over a couple of decades that is a crap load of Watts with big exponentials. The ocean driven oscillations don’t have to require huge changes in forcing to reverse. If they are seeking equilibrium, minor solar changes can change them. That is the weird thing about equilibrium, it don’t take much to change things. The reverse is of course true.

    0.1 percent change in solar irradiance sounds like nothing, but what is required to change things?

    Tsonis with his chaotic analysis, indicated that synchronized oceanic oscillations have a fairly large impact on global temperatures.

  108. Posted Jan 31, 2008 at 10:02 PM | Permalink

    104 (John): There is something with those volcanoes. I’m giving a paper next week at the SORCE meeting in Santa Fe, NM, on a reconstruction of TSI from the heliospheric magnetic field [HMF] deduced from 10Be in ice-cores from Greenland. The reconstructed HMF has strange dips [see Figure on page 2] at times of major volcanic eruptions. The HMF is reconstructed from 10Be whose production depends on solar modulation of cosmic rays, but whose deposition on the ice depends on sulfuric acid aerosols and how efficiently these rain out of the atmosphere. These dips are NOT due to solar modulation [as we have other ways of inferring HMF].

  109. J. Marshall Lancaster
    Posted Jan 31, 2008 at 10:19 PM | Permalink

    Lief one needs to go back over 8,000 years in order to find a time when the Sun was, on average, as active as in the last 60 years.(see link)

    http://www.futurepundit.com/archives/002432.html

    You can see by this paper there is a high in sunspot activity, I agree with Pete that continued high output has baked the Earth and explains warming. This warming period we are in is the 6th since the Holocene maximum, they were all explained by the sun,this one is no different. We will see the true importance of the sun to GW as time goes on.

  110. Posted Jan 31, 2008 at 11:23 PM | Permalink

    49,85 (David A, Rich): Simplicity is often a compelling argument. And certainly the Figure in #49 is a lot simpler than my messy and involved analysis in the “When is Minimum” paper. But simplicity can be deceptive when it stands in the way of reality. The beautiful curves of #49 hide a very messy reality. Below [actually the bottom part] is a plot of the accumulated number of spotless days since the first such day after each sunspot maximum since cycle 9 [there are too many missing days in earlier cycles for a good count]. Large cycles have low numbers of spotless days are colored red. Small cycles [blue] have many spotless days, so there is a natural division into two groups as Jan Janssen also saw. The count for cycle 23 is light green as on #49. But instead of the nice two curves of #49 we have a very broad [3 years], disorganized distribution and it is not clear where SC23 belongs. The reason for this is that the very first spotless day can be influenced by too many random factors [even observational error].

    The upper part of the plot uses a standard technique to reduce the noise. Since we have many hundreds of spotless days we can afford to discard the first few. The upper portion shows the accumulated count of spotless days since the tenth spotless day. Now the spread is a lot less [only 1 year]. Within that spread the red and blue curves wiggle without any clear organization and it is hard to pick one that ‘matches’ SC23. Maybe SC14 [heavy blue curve] is the best analogue. From the first spotless day, the number of spotless days grows as we approach minimum, but as we pass minimum, the number of spotless days declines as we deeper into the new cycle. The curve will thus have an inflection point about ‘halfway up’ which could define the minimum. If SC23 were to be a large cycle on par with the ‘red’ cycles it would seem that SC23 is already at or past the halfway point, but we have only had one tiny new-cycle spot, so clearly there is still a way to go before minimum. If SC23 were to be small, the halfway point might only be reached after 40 months [after 2004/10/24 that was the 10th spotless day which would put minimum in February of 2009. Were SC23 to be of intermediate size, then the minimum would be intermediate between the times mentioned above. So prediction of minimum is intimately bound up with prediction of the size of the cycle. Anyway, you can draw your own conclusions from the graphs.

  111. Posted Jan 31, 2008 at 11:42 PM | Permalink

    109 (Marshall): The 8000 year record-high activity myth is based on a mis-calibration of the sunspot number. This was discussed extensively in the earlier comments on this topic. There is direct evidence from 14C in tree rings that “solar activity around AD 1150 and 1600 and in the late 18th century was probably comparable to recent satellite-based observations”.

  112. bender
    Posted Jan 31, 2008 at 11:46 PM | Permalink

    #107
    I don’t why the control theorists would take offense at this remark. It does not imply that control is not happening, just that it would be very hard to detect. (Although I suppose that might annoy the control empiricists.)

    captdallas2, I encourage you to present this point of view to the gatekeepers at RC. Let me know how you make out.

  113. Posted Jan 31, 2008 at 11:46 PM | Permalink

    110 (me): Folks, the plot is there, but it refers back to my website that may have some update problems right now. Should be better shortly.

  114. usoskin
    Posted Feb 1, 2008 at 3:21 AM | Permalink

    Dear all,
    please avoid biases: when referring to Muscheler’s comment on Solanki et al. (#109), please don’t forget about the Reply by Solanki et al. (2005). Actually, it was Muscheler et al. who made an erroneous calibration of radiocarbon data.

  115. usoskin
    Posted Feb 1, 2008 at 4:11 AM | Permalink

    see also “Solar Proxy” comment #125

  116. a scientist
    Posted Feb 1, 2008 at 6:50 AM | Permalink

    100 (lucia)
    105 (Pat Keating)

    Many thanks for your contributions. lucia is probably correct, the definitions and/or notations may differ between climatology and control theory. In climatology, it seems (at least in one paper I have) that runaway occurs when the sum of the feedbacks is 1 or greater, as in the 1/(1-a) put forth by Pat Keating. It seems iteration, not continuous evolution, is the paradigm in climate.
    More help in understanding this would be most welcome.

  117. Hector Pascal
    Posted Feb 1, 2008 at 7:18 AM | Permalink

    #76 Leif :Richard is just checking my credentials as a tropical cloud watcher
    #75: Richard – yes I was present when Cyclone Tracy flattened Darwin.

    Perhaps we should take any discussion of wet/dry tropics and Hector the Convector to a more appropriate thread?

  118. Erl Happ
    Posted Feb 1, 2008 at 7:34 AM | Permalink

    Leif,
    A link to the paper mentioned in 108 would be much appreciated.

    Is there any evidence of a link between geomagnetic activity and vulcanism or seismic activity and if there is, which way does the link operate?

    What is the physical basis of elecric current induction between the atmosphere and elecric power grids and the Earth itself? Can one anticipate a consistent relationship between the global atmospheric electric circuit and the induction of currents in terra firma or the oceans? Are any parts of the Earth more susceptible to this influence than others?

  119. Posted Feb 1, 2008 at 8:05 AM | Permalink

    118 (Erl): #108 has the link. Click on ‘paper’.
    Geomagnetic activity has nothing to do with seismic activity.
    Current induction: Strong currents can flow at 100 km height. These currents have a magnetic field of their own which can be felt at the ground and below. The currents fluctuate and hence their magnetic field fluctuates. A fluctuating magnetic field induces a further current in any conductor within the field. The Earth is conducting at depth and sea water is a conductor as well, and there are man-made conductors [power lines, pipe lines, telegraph wires in earlier times, etc]. In all these conductors induced currents will thus flow. These induced currents can blow transformers, melt wires, upset communications, etc. Since the conductivity varies from place to place, the effect will vary too.

  120. Posted Feb 1, 2008 at 9:49 AM | Permalink

    118 (Erl): “global atmospheric electric circuit” is a different beast maintained by thunderstorms and not really part of the currents caused by geomagnetic activity.

  121. Pat Keating
    Posted Feb 1, 2008 at 10:25 AM | Permalink

    116 scientist
    I’m not sure what else can be added. If there is internal feedback within the system to a forcing (vin in my post) then you will get the 1-a factor. The CO2 raises the temperature, which raises the water-vapor level, which either:
    (i) raises the temperature some more, and raises the water-vapor level, and so on (positive feedback), or
    (ii) causes clouds, which lower the temperature, which reduce the clouds, and so on (negative feedback).

    You can, of course, look at it as an iteration — after all
    S= 1+a+a*a+a*a*a+…= 1 + a*S,
    so that S = 1/(1 – a). You can do the same thing with any system with feedback. (As an aside, this is how Feynman diagrams tend to sum)

    It seems to me that our energies should be devoted to figuring out whether the climate feedback is positive or negative, not arguing something which is well-established, like the feedback equations.

    In my view, the feedback is positive at low humidity levels and few clouds (polar regions) and negative at high humidity levels (tropics).

  122. Bob B
    Posted Feb 1, 2008 at 10:49 AM | Permalink

    Re #121. I think you have to define the variables more clearly. In electrical engineering where we had Black of Bell Labs, the discoverer of feedback theory, we talk in terms of voltage ,current and power. If you write Vout=Vin a/(1+a). “a” is defined in terms of voltage gain. For climate analogies you have power, Temperature, cloud and water vapor absorbtion, reflectivity etc. Temperature is a scaler and some of the other variables may be phasors.I would like to see one of the climate experts here put the proper variables in a simple equation to show the positve feedback effects.

  123. Larry
    Posted Feb 1, 2008 at 10:50 AM | Permalink

    105, 116, Pat got it on the money in 105. And it really is that simple. That part that I think confuses some people is that some think that all positive feedbacks are unstable, when as Pat showed, it only runs away if the feedback gain is greater than 1. The overall system response is r = 1/(1-g), where r is response and g is gain. It should be obvious why the response blows up at 1.

  124. Larry
    Posted Feb 1, 2008 at 10:54 AM | Permalink

    122, that may be another source of confusion. The climate feedback is defined in terms of temperature. It can also be rearranged in terms of power, but the common feedback multiplier used is degrees per degree, i.e. dimensionless.

    Because the conversion between power and temperature involves Stefan-Boltzmann, it’s not a simple linear transformation (but we already knew that it’s a fudge to apply linear control system theory to the nonlinear climate).

  125. Posted Feb 1, 2008 at 12:32 PM | Permalink

    Larry–
    The other problem is scientists identification of the “feedback” as the magnitude of capital A in 97.

    Systems do run away if capital A in comment # 97 is negative, but capital A is not the feedback.

    Up to a certain point, positive feedback averaged over the planet’s surface, just increases the equilibrium temperature compared to what you expect with the current albedo, cloud cover etc. If feedback got too large, then the system would run away.

    The use is the same as in control theory– I just got confused by #97 saying capital A is the feedback.

  126. Posted Feb 1, 2008 at 1:01 PM | Permalink

    125 (many): so will somebody post a one page summary that everyone can agree one? We can iterate a few times until that point is reached.

  127. Larry
    Posted Feb 1, 2008 at 1:40 PM | Permalink

    Leif –

    http://en.wikipedia.org/wiki/Control_theory

    The confusion over the sign of “A”, as will become apparent from the article, comes from the fact that if you configure a control loop to stabilize something (think cruise control), the gain has to be negative, so the convention is that the gain term is subtracted from the set point. The conventional wisdom in climate theory is that feedback exists not as a device, but as a natural mechanism, and it’s positive; i.e. it’s an amplifier.

    Heretics like Lindzen have proposed mechanisms where there can be natural negative feedback (iris effect), but most feedback mechanisms proposed are positive, i.e. amplifying. An example would be higher temperature causing more water vapor in the atmosphere, which would cause more water vapor greenhouse effect.

  128. J. Marshall Lancaster
    Posted Feb 1, 2008 at 2:15 PM | Permalink

    111 Lief
    Never the less the TSI is very high and was during the last part of the century. The principle holds true that there is a baking effect from continued high TSI and it probably caused the El nino in 1998. Do you hold that there is no accumulated effect ? It would answer a lot of questions.

  129. HMcCard
    Posted Feb 1, 2008 at 4:43 PM | Permalink

    Many comments on this thread relate to feedback in linear systems. Complex systems have multiple positive and negative feedback loops that make them inherently nonlinear. However, there may be narrow limits in some demensions where the system is linear and conditionally stable. Beyond those limits, the system may even become unstable until another feedback loop (negative) achieves dominant control of the complex system. Indeed, intermediate conditions may become chaotic. Inital conditions are ususally important in these instances.

    Since the planetary and climatic systems are inherently nonlinear and chaotic, it seems to me that self-similarity and un-predictability can co-exist.

  130. Pat Keating
    Posted Feb 1, 2008 at 4:45 PM | Permalink

    126 Leif
    As far as I’m concerned, #105 is it. If anyone has a problem with it or a question, I will be happy to respond, but it really is quite simple, especially for an electronic circuit, for which it was first derived.

    To transfer it to climate modeling, vin is the forcing, g represents the system, which ‘transforms’ vin to an ‘output’ temperature rise vout, which in turn acts within the system, to cause feedback. This is simple-minded and the system is really very complex and non-linear, but that’s the essence.

    I would add that there is no a priori reason why the feedback coefficient for solar effects has to be the same as for GHG effects.

  131. Posted Feb 1, 2008 at 5:03 PM | Permalink

    Pat– I agree, #105 is it. My confusion was in the linear ODE written by “a scientist” before you commented. The text said capital “A” in that equation was the feedback as used by someone or other. But that didn’t make sense to me because it didn’t match the verbal discussions about feedback we read in climate science literature..

    Anyway, it turns out the capital “A” in scientists ODE is not the feedback in control theory (which you later denoted with a small ‘a’.) That capital “A” is also not feedback in climate science.

    It IS possible to describe what ‘feedback’ has to do with that ODE, but I need to get ready to go out to a restaurant and then attend my dance lesson. (Some things matter more than control theory! :) )

  132. Pat Keating
    Posted Feb 1, 2008 at 5:54 PM | Permalink

    Yes, the capital A bit was confusing. I’m not sure where it came from.
    Enjoy your dinner and your dance lesson. I guess you’re not knitting tonight…..

  133. Posted Feb 1, 2008 at 6:18 PM | Permalink

    128 (Marshall):

    Never the less the TSI is very high and was during the last part of the century.

    Measurements of TSI started in 1976-78, so what is the basis for your statement? One can only read your statement as also saying or implying that in the first part of the century TSI was very low. Now, I know that there are reconstructions of TSI based on the faulty sunspot data, so no need to rehash those.

  134. Raven
    Posted Feb 1, 2008 at 6:22 PM | Permalink

    Dr Svalgaard,

    I have seen it stated that the sun was 30% weaker 4 billion years ago. Does the new information from TSI proxies change that understanding?

  135. Richard Sharpe
    Posted Feb 1, 2008 at 6:59 PM | Permalink

    I have seen it stated that the sun was 30% weaker 4 billion years ago. Does the new information from TSI proxies change that understanding?

    4 billion years ago is 4 million millennia and 40 million centuries.

    As Leif has already said, that is an imperceptible increase per century.

    Perhaps someone is claiming some sort of step-wise increase every now and then?

  136. Raven
    Posted Feb 1, 2008 at 7:04 PM | Permalink

    Richard Sharpe says:

    As Leif has already said, that is an imperceptible increase per century

    I realize that, however, the new data is forcing people to reconsider old ideas about solar variations in the recent past. I am wondering if this also affects existing ideas about solar variations in the distant past.

  137. maksimovich
    Posted Feb 1, 2008 at 7:07 PM | Permalink

    I have seen it stated that the sun was 30% weaker 4 billion years ago. Does the new information from TSI proxies change that understanding?

    Faint sun paradox

  138. J. Marshall Lancaster
    Posted Feb 1, 2008 at 7:08 PM | Permalink

    LIEF,

    I simply would like to know the answer to one question I think Pete also would like to know.

    Do you think a continued high TSI and sunspots can over time create an increase in temperature, greater then the measured output should cause, because of a baking effect ?

  139. cba
    Posted Feb 1, 2008 at 7:16 PM | Permalink

    HMcCord 129

    It would seem that there could be many factors involved in feedback. There are also many factors which can and do vary for many different reasons. It should be inconceivable that the net feedback is anything but substantially negative so as to smooth out all these random and independent ‘bumps in the road’ because if we’re right next to some mythological tipping point, then the earth would never have maintained its unstable precarious balance for all this time otherwise. Inverted pyramids don’t stay inverted.

  140. Raven
    Posted Feb 1, 2008 at 7:28 PM | Permalink

    maksimovich says:

    Faint sun paradox

    Thanks. Those keywords allowed me to find answer. The 30% weaker sun is an expectation that comes from nature of stars and would not be affected by the new understanding of short term TSI variations.

    But this raises another puzzler: if the climate is extremely sensitive to TSI variations then how can we reconcile the much hotter climates in the past with much lower TSI levels? GHGs? That sounds like a weak explanation there is no clear correlation between GHGs and temps in the historical record. Comic rays? A weaker sun would allow more cosmic rays and therefore cause cooling so that does not work well either. How about different effects for different types of radiation? Did the sun in the past produce different types of radiation?

  141. maksimovich
    Posted Feb 1, 2008 at 7:33 PM | Permalink

    Biological modification

  142. Pat Keating
    Posted Feb 1, 2008 at 7:36 PM | Permalink

    139 cba
    That’s one way to look at it. However, when I look at the Vostok data, I see an oscillator output with a>1 and the oscillator going from one rail to the other. In a sense, the rails finally provide negative feedback (e.g. no more VDD from the power supply) to stop it going further.
    I suspect that we might have ‘hit the upper rail’ with our current nice climate, where an excess of humidity causes negative feedback.
    Guess where we go next!

  143. Posted Feb 1, 2008 at 7:40 PM | Permalink

    137 (Max) yes, this is a normal evolution for a star like the Sun. As they grow older, they grow hotter and more luminous.

    138 (Marshall) you got my name right finally :-)
    You say ‘continuing high TSI..’, but what if TSI and sunspots are not higher now than in the 18th century [as I think] then there is no ‘continuing high’, but just a small [~0.07%] wobble up and down with perhaps a weak 100-year quasi-period. I don’t know what that would do. That is why I’m here to ask that very question.

    Now, there are many [most?] people in the solar community that will disagree with me on the absence of long-term drifts and exceptional activity, so I can soften my stance a bit by asking: what would be the impact IF I were to be correct. If it is such that current understanding[?] demands a higher solar variability in order to account for LIA and MWP and the like, then it becomes doubly important to investigate my claim [at least to me] lest you dismiss my research out of hand [what some people might do/have done; as I also do to some research I come across].

  144. Posted Feb 1, 2008 at 7:46 PM | Permalink

    141 (Pat): But the Vostok data [and climate, of course] is a driven or forced system responding to orbital changes in what looks to me a rather direct way, no feedback or other complications needed.

  145. Raven
    Posted Feb 1, 2008 at 8:06 PM | Permalink

    144 – My understanding is the orbital forcings were too small to explain the observed effect without adding feedbacks. However, if the TSI variations are small and the climate is very sensitive to TSI variations then it would be possible to explain the ice ages without feedbacks.

  146. Dennis Wingo
    Posted Feb 1, 2008 at 8:06 PM | Permalink

    (#11) Leif

    b) examine the basic assumptions and calculations that led us to our earlier beliefs to see what might be improved or questioned [surely some are on shakier ground than others], or at least try to explore possible consequences of the new evidence. Is it ‘impossible’ or not? How dogmatic can we afford to be?

    Leif

    Today I spoke with Otha H. Vaughn, the NASA Scientist that funded the research that was responsible for the discovery of “vertical” lightning, from the tops of clouds to space. He said that there is virtually no ongoing research into this phenomenon at all right now and it would be fairly simple to implement. No one has even put out any papers on this phenomenon as it was recorded by the BATSE instrument on the Gamma Ray Observatory. There are also X ray instruments in orbit right now that can provide this data at a somewhat lower energy but there are no Research and Analysis funds out there to support this research. This is one aspect.

    Another aspect is ultraviolet absorption in the atmosphere. Ultraviolet light is orders of magnitudes more energetic than IR wavelengths and there has not been any systematic bi-directional measurements of this absorption in the atmosphere. We are not even accurately recording the Earth’s energy balance at this time and have terminated the single measurement based on the varying intensity of reflected light from the Moon.

    The bottom line is that we are not doing the research necessary to look beyond just TSI measurements. It is quite clear that there is a solar/climate connection but throwing up one’s hand to say that “TSI variances are too small to explain it” is not an adequate response.

  147. Pat Keating
    Posted Feb 1, 2008 at 8:25 PM | Permalink

    144 Leif

    But the Vostok data [and climate, of course] is a driven or forced system responding to orbital changes in what looks to me a rather direct way, no feedback or other complications needed.

    Sounds fine to me, but I thought it was you who was saying the solar effect is not large enough?

  148. Richard Sharpe
    Posted Feb 1, 2008 at 8:47 PM | Permalink

    I am wondering, in my simple view of things, why we are not talking about energy storage and transport through the atmosphere and hydrosphere rather than temperature. Because of the properties of water and my (perhaps mistaken understanding) that the energy content of wet vs dry air are different, and the fact that the oceans can store enormous amounts of energy when there is a rise in insolation and release it later when insolation drops, it seems that the important thing to understand is energy transport.

    117: (Hector Pascal): Send me email at rsharpe@richardsharpe.com. It is amazing to come into contact with someone who shared those events (along with about 40,000 other people) and who has similar interests.

  149. Erl Happ
    Posted Feb 1, 2008 at 9:05 PM | Permalink

    146 Dennis

    The bottom line is that we are not doing the research necessary to look beyond just TSI measurements.

    That is a telling comment. From what I can see (post to come)from a close study of tropical cloud cover between July 1996 and December 2000 the variation is not explained by feedback mechanisms from surface warming. In the broad the tropical sky is clearer when the Earth warms and less so when it cools, but there are violet swings from month to month. The data also suggests that relatively clear skies do not always bring surface warming. This suggests that the dynamics of the heating and cooling process within the period of a solar cycle owe something to phenomena above the tropopause. The upper atmophere response to UV light is something to investigate, if only to rule out this dynamic as a possible influence on the penetration of other wave lengths.

    The variation in the interception of radiation by the atmosphere is the obvious frontier in solar science. A better understanding will elucidate the ENSO phenomenon and once and for all invalidate the pernicious and dismissive notions of ‘climate noise’ and ‘internal oscillations’.

  150. Posted Feb 1, 2008 at 9:22 PM | Permalink

    147 (Pat): The orbital effects are a hundred times as large as the TSI variations.

  151. Posted Feb 1, 2008 at 9:53 PM | Permalink

    146 (Dennis): Yes, I have also lamented that there is no money for research, but such is it. The albedo measurements are very important and should be continued. Palle is working on setting up a global network for this.

  152. J. Marshall Lancaster
    Posted Feb 2, 2008 at 1:46 AM | Permalink

    Leif,

    Along with a baking effect from consecutive years of high TSI there is the cosmic ray effect and its interuption of cloud forming. What is your opinion of Henrik Svensmark’s studies and their impact on GW.

    Here is an extensive interview with Henrik Svensmark.
    :http://discovermagazine.com/2007/jul/the-discover-interview-henrik-svensmark/article_view?b_start:int=0&-C=

  153. bender
    Posted Feb 2, 2008 at 3:12 AM | Permalink

    #129
    From a control theory perspective it doesn’t really matter how the nonlinearity in the feedback term arises, just that it does. The point is that runaways that happen in hypothetical linear system models may not happen in their nonlinear analogues, and may not happen in real-world nonlinear systems. Climatologically, the question is what would be the mechanism behind the supposed nonlinearity. e.g. Tropical clouding & convection in ITCZ?

  154. Andrey Levin
    Posted Feb 2, 2008 at 6:10 AM | Permalink

    runaways that happen in hypothetical linear system models

    Bender, thanks for the illustration.

    Black body, gray body, or even colored body emit IR radiation as fourth power of absolute temperature, even through tight radiative window. Only illiterate moron could believe Hansen that any radiative logarithmic positive feedback could overcome 4-th power negative feedback and produce runaway climate heating.

  155. Jordan
    Posted Feb 2, 2008 at 6:17 AM | Permalink

    This is a great thread. Thanks to Leif and others for making it so interesting.

    If I may be permitted one long-ish post to reply to some of the discussion about feedback. I’ll cut it there because we should not allow the thread to lose its focus.

    Back in #97, a scientist introduced the following first order differential equation:

    dTE/dt = -A.Te + B.S +C.G

    Problem here is poor choice of units. The units on the LHS are rate of change in degrees/time (deg C or F). The units on the RHS must be the same. So if the anomalies are in degrees, the B and C parameters must be ‘per-unit-time’ dimensions. That get things off to a bad start. Too obscure.

    Perhaps a better representation would be: A.(dTE/dt) = -TE + B.S +C.G

    A is now a time constant. Parameters B and C are dimensionless and we can have a more clear discussion about why they have any particular value.

    For example, in #98, lucia set the LHS to zero in order to assess steady state (s-s) response. If we use the second of the above equations, we have: TE = B.S + C.G

    Note how there is no factor ‘A’ in the denominator, and there is no need to debate s-s amplification attributable to A. Amplification, if it exists, needs to be accounted for in the derivation of parameters B and C.

    A common, simple example (one I’d like to use for illustration) is an electronic low-pass filter:

    RC.(dVo/dt) = Vi-Vo.

    Using lucia’s method of setting the LHS to zero, this circuit has no s-s amplification.

    In #105, Pat makes the leap into difference equation: Vout = g.Vin + a.Vout.

    This a discrete representation of a feedback system and I believe the step-over to discrete is likely to be confusing. There is not a nice intuitive mapping from continuous to discrete representations.

    To show this, my simple low pass filter can be represented using the following recursion at the k’th time step kT:

    Vo(kT) = (1-b).Vin(kT) + b.Vo(kT)

    It is important to note that this is an approximation to the continuous equation – there is nothing to say that the discrete equation will reproduce the exact value of the continuous response at each time interval. It is always worth bearing in mind that discrete models introduce their own errors.

    For my simple low-pass filter, the parameter ‘b’ should be equal to b=exp(-T/RC).
    We can change the value of ‘b’ by changing the R and C devices. But this not available to climatology. Climatology needs to analyse the climate response using empirical methods and use the results of that type of analysis as a-priori knowledge when formulating the mathematical models in the GCMs.

    Alternatively, we can change the value of ‘b’ by altering the step interval T. But model error is sensitive to this – particularly as the step interval increases towards the value of the dynamic parameters (T much shorter than RC).

    If I claimed to have “adjusted the feedback” (parameter ‘b’) by doing no more than altering the time step T, would you be convinced if I also claimed that my RC filter will now amplify the input signal? Nah.

    Final point. I know that my simple RC filter has no s-s amplification. To get that into my difference equation, my mathematical modelling exercise needs to go through two steps. Firstly derive the parameter ‘b’ which represents dynamic response (RC and my chosen step interval T). Secondly, derive another parameter which secures the s-s response, and that gives me an input scaling factor (1-b) in the difference equation.

    In my equation, you can alter the value of b for different reasons, but it does not affect the model’s s-s amplification.

    Ps. Good point is a post further above (can’t find it right now). If there has been past changes to inputs and the dynamics have not yet settled down, in principle, an output can be changing with no(or insignificant) observed change in the input. I think it can be true for relatively short period of observation. Or, at least, short compared to climate response times.

    As I said above, that should be my final word on this thread. Maybe get back to interesting discussion about solar influence.

    Hope it was helpful.

  156. Jan Pompe
    Posted Feb 2, 2008 at 6:32 AM | Permalink

    Bender and all,

    the problem I have with much of the discussion about control theory is that it was developed for industrial systems, nuclear reactors and machines which all have internal power supplies or energy sources ( although it may be generated outside the plant) that permit gain >>1. In fact in most formulations and analysis it is tacitly assumed that whatever energy is needed it will be available. In any case if it isn’t we’ll just put in a bigger transformer, sub station or generator and leave it to the power transmission specialists to deliver it. The upshot being that most of the naive quotes regarding feedback .eg. To/Ti = A/(1 + BA) really won’t work all that well for any case where A (assumed)>1 linear or not.

    I have put a derivation for a black body or body at thermal equilibrium as per Kirchoff’s law on the BB

    Differentiating wrt to Ts gives a temperature sensitivity dTe/dTs = ~.048 that is for very 1K change on suns temperature we get .048 on planet earth.

    then differentiating wrt R dTs/dR = Ts/2* sqrt(rs/2R^3) = 9.3141e-10K/m which leads to an expected variation of ~4.2K for the annual orbital variation of 3%

    Please check the numbers I am getting on a bit. It kind of verifies what Leif has been saying about the comparison of orbital effects and TSI effects.

    Leif – Is there a series available of solar surface temperature measurements I have not been able to find one and have spent several hours looking.

  157. Posted Feb 2, 2008 at 7:30 AM | Permalink

    156 (Jan):

    Is there a series available of solar surface temperature measurements

    Back in the very first post of this thread I wrote:
    Line 1:
    The Total solar Irradiance (TSI) has several sources. The first and most important is simply the temperature in the photosphere. The hotter the sun, the higher the TSI. Some spectral lines are VERY sensitive to even minute changes in temperature. Livingston et al. has very carefully measured the line depth of such temperature-sensitive lines over more than 30 years spanning three solar cycles [Sun-as-a-Star Spectrum Variations 1974-2006, W. Livingston, L. Wallace, O. R. White, M. S. Giampapa, The Astrophysical Journal, Volume 657, Issue 2, pp. 1137-1149, 2007, DOI; 10.1086/511127]. They report [and I apologize for the somewhat technical turn my argument is taking, but if you really want to know, there is no avoiding this], “that both Ca II K and C I 5380A intensities are constant, indicating that the basal quiet atmosphere is unaffected by cycle magnetism within our observational error. A lower limit to the Ca II K central intensity atmosphere is 0.040. This possibly represents conditions as they were during the Maunder Minimum [their words, remember]. Within our capability to measure it using the C I 5380A line the global (Full Disk) and basal (Center Disk) photospheric temperature is constant over the activity cycles 21, 22, and 23″. I have known Bill Livingston [and White] for over 35 years and he is a very careful and competent observer.

    I could get his series of data, but since they show a constant temperature, there would not be much point in getting a long series of constants numbers [+random noise].

  158. Pat Keating
    Posted Feb 2, 2008 at 8:03 AM | Permalink

    150 Leif
    Please excuse my confusion. I was under the impression that you did not believe that insolation effects are large enough to explain the global temperature variations we observe. If I understand you correctly in 144 and 150, you believe that TSI effects are not, but insolation changes due to orbital effects ARE sufficient to explain them, including the 10-12C changes in the Vostok record.

    Is that a correct understanding of your position?

  159. Pat Keating
    Posted Feb 2, 2008 at 8:08 AM | Permalink

    155 Jordan

    In #105, Pat makes the leap into difference equation: Vout = g.Vin + a.Vout.

    That’s not a difference equation, where did you get that idea? vin and vout are normally continuous functions of time.

  160. Posted Feb 2, 2008 at 8:22 AM | Permalink

    158 (Pat): yes, because the orbital effects are MUCH, MUCH larger than the TSI variation.

  161. Jan Pompe
    Posted Feb 2, 2008 at 8:25 AM | Permalink

    Leif #157

    Thank you for your response and patience I do remember that earlier post. It’s just that I would have liked a look if it was readily available but as you say it is probably pointless. Even the rather large expected (black-body earth) variation due to orbital variation is barely noticeable if it exists or not swamped out by other effects.

  162. EW
    Posted Feb 2, 2008 at 8:29 AM | Permalink

    For a totally confused biologist please specify: Which are those orbital effects? Milankovich cycles or something else?

  163. Pete
    Posted Feb 2, 2008 at 8:36 AM | Permalink

    #146 Dennis Wingo:
    “We are not even accurately recording the Earth’s energy balance at this time and have terminated the single measurement based on the varying intensity of reflected light from the Moon.”

    Dennis, I thought Big Bear Observatory was activily measuring this via the Earthshine project? I pointed Leif to an updated paper on this only a few weeks ago.

  164. See - owe to Rich
    Posted Feb 2, 2008 at 9:02 AM | Permalink

    I hope y’all will find it helpful if I show how the different “control theories” at #97 and #105 are in violent agreement, with A = 1-a, and the former supplying the time evolution of the process. It is possible to do this from first principles, as follows.

    We are trying to model a temperature anomaly T with respect to a constant forcing F, where F = BS + CG in #97. In the finite evolution of #105, with feedback constant a, we can write

    T(0) = F, T(1) = F+aF, T(2) = F+a(F+aF), … T(n) = F(1+a+…+a^n) -> F/(1-a)

    Then T(n)- T(n-1) = F + (a-1)T(n-1).

    To turn this into an infinitesimal from which to derive a differential equation, write

    dT(t) = dt (F+(a-1)T(t)), which is

    dT/dt = -AT + F, with A=1-a, as at #97.

    The solution to this with T(0) = F is

    T(t) = (1 – (1-A)exp(-At)) F / A

    So this shows how the constant forcing F changes temperature anomaly from
    F to F/(1-a) as time increases.

    DTH?
    Rich.

  165. Pat Keating
    Posted Feb 2, 2008 at 9:05 AM | Permalink

    160
    Thanks for the response, Leif.

    I guess the AGW view is that orbital effects are enough to start deglaciation (say), but it requires a lot of CO2 to drive the temperature up to say 1800 levels.

    It seems I have badly misunderstood your position.

  166. See - owe to Rich
    Posted Feb 2, 2008 at 9:09 AM | Permalink

    Hey, I just posted #164 before being able to see #155. Sounds like we are saying just about the same thing (but mine’s shorter, which may be good or bad).

    Rich.

  167. Posted Feb 2, 2008 at 9:46 AM | Permalink

    162 (EW): Yes, Milankovich cycles

    163 (pete): I have been in contact with Enric Palle [albedo-guy]. Here is a clip from a email from Enric:

    “In the past months I have moved from Big bear to the Canary Islands where we are planning to install a new earthshine telescope. We also wanted to let some time pass before comparing the data and calibrating a single time series.

    However, a different way to help that might be interesting to you (and us) would be by hosting an earthshine telescope. Currently we have two robotic telescopes one in Big bear and a second being shipped to the Canary Islands, and a third manual telescope in Crimea that has been taking data for a couple of years now. However, if fundings are available, we are planning to expand to a network of robotic telescopes hopefully to between 6 and 8 stations. Our telescopes are now almost
    fully automatic, and all the components can be purchased ‘of the shelf’. The idea is to replicate the existing telescopes for which we already have the software. A full telescope, including shipping and installation costs about $80,000. Do you have a place/suggestion where to put one?”

    So, the albedo monitoring project is alive [and hopefully well].

    165 (Pat): I’m not sure what my ‘position’ is that you are referring to. I think that Milankovich effects are the main driver of ice age related variations [combined with suitable arrangements of continents, ocean currents and other details]. These effect are very large [100 times TSI variations], and I don’t know if one can scale from the large [orbital] to the small [TSI]. Explain to me, what you think [and thought] my position[s] is[are]. (if it is important).

  168. Raven
    Posted Feb 2, 2008 at 11:57 AM | Permalink

    Leif Svalgaard says:

    I think that Milankovich effects are the main driver of ice age related variations [combined with suitable arrangements of continents, ocean currents and other details

    Your view directly contradicts claims made by AGW advocates who say that Milankovich effects are too small to explain the ice ages and insist that the real cause was CO2 induced positive feedback. The CO2 sensitivity measurements used into models today are, in part, derivied from analyzing temperature changes during the ice ages and assuming that almost all of the temperature changes can be attributed to CO2. If they are wrong wrt to the Milankovich effects then the future warming predicted by the models is likely overestimated.

    Here are RC’s delcarations on the topic:

    The reason has to do with the fact that the warmings take about 5000 years to be complete. The lag is only 800 years. All that the lag shows is that CO2 did not cause the first 800 years of warming, out of the 5000 year trend. The other 4200 years of warming could in fact have been caused by CO2, as far as we can tell from this ice core data.

    The 4200 years of warming make up about 5/6 of the total warming. So CO2 could have caused the last 5/6 of the warming, but could not have caused the first 1/6 of the warming.

    I am I reading too much into your comments?

  169. Pat Keating
    Posted Feb 2, 2008 at 12:13 PM | Permalink

    167 Leif
    Re your position: I understood (correctly, I guess) that you felt that the TSI effects were too small. But I had no idea that you feel that the orbital effects are large enough to explain 10-12C without feedback.
    I am not quarreling with your position, just surprised (like Raven in 168).

  170. EW
    Posted Feb 2, 2008 at 12:29 PM | Permalink

    The reason has to do with the fact that the warmings take about 5000 years to be complete. The lag is only 800 years. All that the lag shows is that CO2 did not cause the first 800 years of warming, out of the 5000 year trend. The other 4200 years of warming could in fact have been caused by CO2, as far as we can tell from this ice core data.

    That’s the thing which puzzles me the most. The warming is started by a cause X (Milankovich). After 800 yrs another factor Y (CO2- or something else yet?) appears, which may increase the warming. So – is there any possibility to separate the X- and Y-caused parts of the total warming effect? Or are we here firmly on the ground of “could have been, as far as we can tell”???

  171. Raven
    Posted Feb 2, 2008 at 12:42 PM | Permalink

    EW says:

    So – is there any possibility to separate the X- and Y-caused parts of the total warming effect?

    In my opinion – no. Especially since the majority of the CO2 amplication comes from water vapour feedback which could easily be attributed to the Milankovich effects instead of the CO2. Personally, I think the RC argument is an ad hoc attempt to salvage the CO2 hypothesis after the 800 year lag showed up in the new ice core data. Unfortunately, I had to take them at their word when they claimed that the Milankovich effects are too weak to explain the changes without CO2 feedbacks.

  172. EW
    Posted Feb 2, 2008 at 1:06 PM | Permalink

    RC argument is an ad hoc attempt to salvage the CO2 hypothesis after the 800 year lag showed up in the new ice core data.

    And a very often repeated one. Lag is not problem, CO2 did the resat…but if there is no possibility of modeling or separating the effects, does it mean that we are standing here with empty hands, with practically no knowledge about the mutual effects and interactions of the factors X, Y, Z…?

  173. cba
    Posted Feb 2, 2008 at 1:11 PM | Permalink

    Pat,

    As to guess where we’re going, my guess is it’s going up and it’s going down (multiple times) and ultimately, it’s going way up in the very extreme long run. I think the cloud factor (or Lindzen’s Iris) is a rather important factor in this. The orbital variations are definitely part of the whole on the long term. The melankovich cycles do tend to jive with timing for the most part although there may be problems with some of it. Just our current aphelion, perihelion variations amount to something like a theoretical difference during the year of almost 100W/m^2 in insolation at the TOA.

    From what I understand, the majority of our albedo is coming from clouds, not from the surface. Snow does offer something along the same ballpark as clouds so large scale icesheets would tend to have a dramatic effect on this probable (probable at least in my mind) regulating mechanism which then creates a bit of a short circuit so to say that defeats the mechanism that would otherwise allow more energy in to balance the heating.

    Additionally, there are such things as plate tectonics going on that are changing the layout of land and oceans. This would affect and disrupt and require rerouting of ocean currents when such things as central america connected the north and south american continents.

    Considering that just the earth’s axis tilt is sufficient to establish summer and winter variations over very short time frames for large amounts of the globe, having variations in orbital eccentricity, tilt angle, etc. could conceivably have some fairly serious ramifications. Also, notions of long times required for effects to settle seem totally blown out of context as well since large areas undergo such effects twice each year.

    Some consider the sun to be a variable star over all time frames. TSI certainly hasn’t changed much over the last 30 to 40 yrs and according to Leif, much longer than that. That doesn’t mean that the composition of that radiation doesn’t change significantly in some wavelengths and it doesn’t mean that energy at wavelength X has the same effects on planet earth as that same amount of energy at wavelength Y. Variations can be both how much is converted to thermal versus into biological processes as well as where that energy is deposited. As I recall, uV radiation can vary by 5 to 10% even though the total energy coming in is varying by substantially less than 1%.

    At present, our eccentricity variation on incoming between NH and SH of nearly 100 W/m2 TSI is vastly greater than the purported 3-4 W/m^2 attributed to the expected co2 doubling. While one is expected to be consistant while the other varies with time, there should still enough difference to make comparisons possible between NH and SH.

  174. cba
    Posted Feb 2, 2008 at 1:22 PM | Permalink

    Leif,

    What is the target for the albedo measurement (earthbased) you posted? Is it the dark area of the moon or an orbiting target or atmospheric glow?

  175. Posted Feb 2, 2008 at 1:52 PM | Permalink

    168-171 (many): I’m not a paleoclimatologist [or whatever the correct word is] and don’t have strong opinions on the causes of ice ages. As most scientists in the Earth-Sun-Planet fields I have what I think is a reasoned opinion based on my general knowledge of these things. If someone says I’m wrong on this or that, that doesn’t surprise or upset me [just tell me where and give understandable reasons if on topic]. So, again, I have no “position” on the ice ages. My understanding is as follows:

    Planetary [incl. lunar on the oblate Earth] perturbations cause several orbital [and spin axis] elements to fluctuate around their average values on time scales of tens [and more] of thousands of years. The distribution of land and sea [including ocean currents] also changes with time but on million year time scales. When things are lined up just right we get an ice age. To me, the ice age lasts millions of years [as long as the land/sea distribution is favorable]. When the orbital elements are favorable for a small warming [e.g. when the land-rich Northern Hemisphere has winter when the Earth is closest to the Sun], the Earth warms up a bit. That drives CO2 in the oceans out of solution and the sea warms a bit more because of the added CO2 which drives more CO2 out, etc, so we get a ‘mini’ run-away that melts [most of] the ice sheets. Eventually that CO2 goes back into the ocean [when the orbital elements no longer favor warming] and we have a slow slide back into a cold period with extensive ice sheets again advancing. We live in one of these warm periods, which will come to an end some thousands of years in the future. The CO2 we are adding now will cause a spike in CO2 concentration now, but that spike will be absorbed into the ocean some hundreds of years from now [when we have run out of oil and no more keep adding CO2] and we will then continue our sure slide into a cold period with extensive ice sheets. This is how I see it, but, again, this is not a subject for discussion [off our topic of solar variability and climate] and I don’t feel strongly about any of this and don’t want an argument. Anything else you want to believe is fine with me, unless you can show how it fits within our topic.

  176. Posted Feb 2, 2008 at 2:00 PM | Permalink

    174 (cba): As Leonardo da Vinci pointed out, the Earth shines on the Moon, so the areas of the Moon that are not in sunlight must be lit solely by reflected Earthlight. Since there is nothing on the Moon to make it change its albedo, any change in the reflected light ["old Moon in the arms of the new"] must be a change of the Earth’s albedo [ignoring or correcting for the minuscule changes in TSI].

  177. cba
    Posted Feb 2, 2008 at 2:40 PM | Permalink

    I guess the 12% or so lunar albedo averaged over the same area(s) would probably do a decent job once corrected for fraction of earth light present. It just seems like such potentially ‘grubby’ variation riddled data compared to a TOA direct measurement from satellite. Granted a 16-20″ robotic telescope for $80k is a lot cheaper than a geosynch. orbit satellite, but then no one needs continually modify the data to correct for the changing geometry. For some reason I thought there was already quite a bit of ability at getting direct albedo measurements from existing satellites – along with TSI which is not directly avail for the passive reflector.

  178. Posted Feb 2, 2008 at 2:56 PM | Permalink

    177 (cba): correcting data for varying geometry is no problem. This is done already for TSI and for F10.7 flux, where two numbers are given: one for the actual measured data and one corrected for the varying distance to the Sun. The beauty of the Moon is that it has constant calibration, something we cannot say for the TSI data. To wit the ‘acrimonious’ debate between the ACRIM, PMOD, and SORCE TSI-series.

  179. EW
    Posted Feb 2, 2008 at 4:12 PM | Permalink

    The CO2 we are adding now will cause a spike in CO2 concentration now, but that spike will be absorbed into the ocean some hundreds of years from now [when we have run out of oil and no more keep adding CO2] and we will then continue our sure slide into a cold period with extensive ice sheets.

    I got interested in paleoclimate when studying its effect on evolution of grasses and fungi. Seems there was always a lot of change so I can’t quite get the AGW uproar now. Anyway, about the future I see it similarly as you – hardly a case for runaway greenhouse.

  180. cba
    Posted Feb 2, 2008 at 4:59 PM | Permalink

    Leif,

    I gather that it pretty much has to be a Bond albedo – visible only as the atmosphere is going to absorb and scatter the other stuff, depending on wavelength. I suppose with other data, such as instantaneous cloud cover that one could possibly have a great of information to glean but my mind just keeps going back to all the problems associated with the light reflecting and coming right back into the atmosphere again. It’s not as bad as the first time because that’s where the main attenuation happens and most of whats left has been filtered already before hitting the moon.

    Currently, the numbers I’ve seen all tend to say 29-31 % albedo – save one number I stumbled across the other day and misplaced where which was a real outflyer at 36%. I think these are all total albedo rather than Bond and represent reflected/scattered light divided by TSI. It looks though that in atmosphere measurements (aircraft/ground based) tend to call themselves albedo despite the fact that there is attenuation of the incoming light in the atmosphere and attenuation of outgoing light in the atmosphere making it more of a reflectivity that needs adjustment to become actual albedo.

  181. Chris Knight
    Posted Feb 2, 2008 at 5:16 PM | Permalink

    #104, #107 Volcano contribution to Dalton & Late Maunder minima.
    I wonder if stratospheric dust and lower level volcanic haze (as well as increased cloudiness due to the aerosols,atmospheric instability due to variation in stratospheric temperatures and Jet stream irregularities) confounded the observations of sunspots in the 15th – 19th centuries. There is thus the possibility of poor atmospheric viewing conditions causing small sunspot groups to be missed, and periods of lack of observations due to cloudiness.

    The presence of volcanic aerosol and particulates may also be responsible for adsorption of stratospheric ions changing their usual distribution patterns when they eventually are precpitated to the ground.

    If there is anything in this at all, perhaps there has been some observational bias in ascribing too much to solar forcing of these climate events.

    Finally, there is a known tidal influence on vulcanicity and seismic events. Should we be looking at the lunar contribution as well? Apsidal cycles, Nodal cycles, nutation and Length of day variation due to lunar tidal braking?

    The major eruption time line (not exhaustive) from Greenland/Antarctic ice core sulphate/ash data is:

    Maunder 1645-1715

    1640-1641 Kelut Indonesia, Parker Phillipines Komagatake Japan*
    1653 unknown S hemisphere
    1661 Long Island Indonesia*
    1667 Shikotsu (Tarumai) Japan*
    1671-73 Gamkonora Indonesia, San Salvador El Salvador
    1680 Tongkoko Indonesia, Krakatau Indonesia
    1689 Chikurachki Kurile Islands Russia?
    1691 Reventador Ecuador
    1694 unknown (Serua or Gunung Api Indonesia)
    1711 S hemisphere
    1715 N hemisphere
    ————

    Dalton 1785-1825

    1783 Laki Iceland VEI4, DVI2300, IVI0.19
    1788 unknown tropical
    1795 unknown S Hemisphere
    1808/9? unknown (possibly 2 contemporary unrelated eruptions) VEI6, DVI1500, IVI0.19
    1812 Soufriere, Caribbean
    Awu Indonesia
    1815 Tambora Indonesia* VEI7, DVI300, IVI0.50
    1817 Raung indonesia
    1822 Galanggung
    ————

    dates from:
    http://www.ub.es/gc/Documentos/Prohom_CHAPTER_2.pdf from doctoral thesis (english translation)
    PROHOM DURAN, Marc J. (2003): Incidència de les grans erupcions volcàniques en el clima de la Península Ibèrica i Balears.
    http://www.ub.es/gc/Castella/Marc%20Prohom%20Duran.htm

  182. MGW
    Posted Feb 2, 2008 at 5:37 PM | Permalink

    In Svalggard #2, posting 264, Leif had referenced recent work by White and Liu. I apologize for this delayed response to that posting. I read the thread just today and felt I might be able to offer my insight in hopes that it could clear up a question posed by Leif.

    Leif’s posting gave the reference:

    JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113, C01002, doi:10.1029/2006JC004057, 2008

    Resonant excitation of the quasi-decadal oscillation by the 11-year signal in the Sun’s irradiance

    Warren B. White

    Scripps Institution of Oceanography, University of California, San Diego, La Jolla, California, USA

    Zengyu Liu

    Center for Climate Research, Gaylord Nelson Institute for Environmental Studies, University of Wisconsin-Madison, Madison, Wisconsin, USA

    He questioned the authors’ use of forcing: “I’m a bit at a loss why they assume a 2 W/m^2 solar cycle variation while direct measurements of TSI shows a 1 W/m^2 change.”

    I think I can explain. The use of 2W/m^2 for the quasi-decadal change in forcing is used, instead of the usual 0.1% of the mean solar constant (~1.0 to 1.4 W/m^2), so that after the authors took into account the averaging of energy delivered around the globe and the albedo(divide by 4, multipy by .70), the amount of surface radiative forcing delivered to the upper ocean would be ~ 0.5W/m^2. The numbers don’t seem to come out exactly (this I can’t justify), but this value of 0.5W/m^2 is aligned with what they found in their 2006 study. In that study they found that due to changes in latent and sensible heat fluxes across the air-sea interface, changes which were due to forcing from the stratosphere-down (change in lapse rate), the diabatic heat storage in the upper ocean, observed from peak to trough of the quasi-decadal solar cycle, was equivalent to about 0.5W/m^2 (~0.05K). Because the model in the referenced 2007 study lacked capacity for ozone absorption in the lower stratosphere, and thus could not perform the physics necessary to produce this observed result, White and Liu ’08 manipulated the solar constant to effect this ultimate change in indirect solar forcing.

    I do not speak as an expert, but I have read many of White’s papers. The 2W/m^2 caught my attention too. But the reasoning is given in the text. But, had one not read the 2006 paper, their reasoning would not have been so clear.

  183. Chris Knight
    Posted Feb 2, 2008 at 5:47 PM | Permalink

    #181 Link (http://www.ub.es/gc/Documentos/Prohom_CHAPTER_2.pdf)- appears not to work directly try via google from

    http://64.233.183.104/search?q=cache:m9vniUgfUQgJ:www.ub.es/gc/Documentos/Prohom_CHAPTER_2.pdf+gamkonora+tongkoko+laki&hl=en&ct=clnk&cd=7

  184. Philip Mulholland
    Posted Feb 2, 2008 at 8:11 PM | Permalink

    cba #177

    Using earthshine (lunar reflected earthlight) to measure the earth’s albedo is an interesting scientific experiment with a number of benefits compared to albedo measurements made with an artificial satellite in low earth orbit. The first (and perhaps least significant benefit) is that the moon, orbiting at a height of 240,000 miles, observes a larger fraction of the earth’s disc (89 degrees visible) than the geostationary satellite, at a height of 36,000 miles does (81 degrees visible). A much more obvious benefit is that a geostationary satellite, located over Africa, never observes the weather over the Pacific Ocean, whereas the moon sees the earth rotate, turning once a day 240,000 miles above* it, so the moon effectively scans almost the whole earth once every 24 hours.
    (* He stood on the moon and looked up at the earth. “Wow! The Moon!” Gene Cernan, Apollo 17 Astronaut).

    At the time of new moon on the earth it is full earth on the moon, so the new moon is located above the earth’s local midday and receives the maximum possible earthlight. The problem for an earth observer of course is that the new moon occurs during daytime and so we need to measure the earthshine in the waning phase and waxing phases of the moon when the sky is dark before dawn (waning crescent) and after dusk (waxing crescent).

    The third benefit of measuring the moon’s earthshine is that the annual seasonal variations can be recorded. In December the new moon is high in the southern hemisphere’s daytime sky and so the earthshine includes light from the fully lit Antarctic continent. We are therefore including measurements of the earth’s southern polar albedo at this time. Correspondingly in June the new moon is now high in the northern hemisphere’s daytime sky and we are therefore including the lit Arctic albedo in our earthshine measurements.

    ‘I saw the new moon late yestreen
    Wi’ the auld moon in her arm,
    And if we gang to sea, master,
    I fear we’ll come to harm.’

    The Ballad of Sir Patrick Spens

  185. maksimovich
    Posted Feb 2, 2008 at 8:32 PM | Permalink

    re 181 (ck)

    The average annual increase of CO2 went up from about 0.9 ppmv/year during the 1960s to about 1.5 ppmv/year during the 1980s. The annual CO2 growth rate has kept fluctuating since the start of direct measurements in 1958. Many fluctuations appear to be related to ENSO events.The drop of the CO2 growth rate between late 1991 and late 1993, however, cannot be directly linked to an ENSO event.The rise of atmospheric methane and of nitrous oxide temporarily slowed down simultaneously. Mt. Pinatubo’s 1991 eruption may have played a role, why? IPCC 95, p 75-6

  186. Posted Feb 2, 2008 at 10:13 PM | Permalink

    Folks: Tomorrow morning I’m off to the
    SORCE Science Meeting. So, I’ll have less time to respond to comments. I’ll try to report on the highlights of the meeting as time permits.

  187. _Jim
    Posted Feb 2, 2008 at 10:24 PM | Permalink

    Andrey Levin at 154 says (in response to a post by bender):

    Black body, gray body, or even colored body emit IR radiation as fourth power of absolute temperature, even through tight radiative window. Only illiterate moron could believe Hansen that any radiative logarithmic positive feedback could overcome 4-th power negative feedback and produce runaway climate heating.

    The essence to understanding the above statement is I think: look at the ‘movement’ in the 4th power Planck spectral curves for different temperatures of the ‘black-body’ earth emitting/radiating through the LW IR ‘window’ that exists down in the 8 .. 15 um wavelength region.

    As temperature increases, there is a pronounced increase in the amplitude of that part of the spectral ‘curve’ that develops in that IR window, and it has a pronounced effect on the amount of IR radiated (IR imagery from satellites of water and land masses make use of this window, and it seems to work quite well out to geosynchronous orbit heights anyway).

    This became apparent months back – when inspecting ‘data’ (Planck spectral curve for a given temp, atmosphereic ‘transmission’ capability vs wavelength ala MODTRAN and work cba/Charles performed) plotted on the relevant graphs.

  188. cba
    Posted Feb 2, 2008 at 10:27 PM | Permalink

    184, Phillip,

    I don’t necessarily see those as real benefits. Two satellites would cover the same area not possible with one. Full earth shine as you mention is during daylight where a ground based scope is battling scattering (blue sky) which is something that helps with the albedo amount. Going at greater angles is going to toss in a bit of limb darkening I’d bet plus the moon is going to do its wobble bit shifting around the point directed at the earth’s center, changing the precise nature of what the moon albedo is and angles related to where the earth is coming from. Granted it’s probable that programming could accomodate all this mess, but to me it seems like a poorer way to go, even if far cheaper. Also, there is pretty much total integration for the moon which means one would lose the resolution that could help build models as to albedo of clouds vs. forest vs. sand etc. It would also seem a lagrangian point distant satellite could also be done in permanant distant monitoring of the daylight side. While far more expensive than a low orbit one, never mind an $80k miniobservatory, a mission like that could be multifaceted such as another solar satellite. Seems like there’s some rather important monitoring work still being done by some rather obsolete hardware out there at present.

    It’s probably a great idea offering lots of bang for very few bucks, but I’m just having problems thinking about all sorts of design problems that have to be overcome to ever get good data out of something like that along with the nature of the data that would seem to be far more desirable (high spacial resolution) to acquire that cannot be done so by the approach. My appologies for sounding like such a negative naybob.

  189. cba
    Posted Feb 2, 2008 at 10:48 PM | Permalink

    Jim,

    It’s even slightly worse than that. Absorption and emission are related and for the simplistic Stefan approach, would effectively have an increase in the emissivity factor as there is a change in co2 concentration. The whole burden of temperature readjustment would not appear to be necesssary as the emissivity factor is increasing as well. More understandably, emissivity is a function of wavelength, not a simple single valued engineering constant. This change in emissivity would not be present for a variation in TSI, making the earth more sensitive to solar variations than to ghg concentration changes. At least this is what my understanding is at present.

    The furtherest I got was to calculate a doubling of the 1976 std atm co2 to compare with the 1976 std atm. It took all day and I can still smell some burnt heatsink grease from my laptop. I think I came rather close to frying it and at present, I’ve got nothing close in speed to switch over to in order to continue. It is a fairly high precision one dimensional model that still has to be verified a bit better also.

  190. _Jim
    Posted Feb 2, 2008 at 11:51 PM | Permalink

    Thanks Charles; I’m glad you went through the trouble of running your 1-D ‘model’ varying OCO concentration levels and sacrificing a laptop in the process for the furtherance of science.

    The one question I had was: How true to life were the plots/graphs that were generated, in cases where in actuality there were very many fine lines indicating absorption and/or transmission (IOW did the graphing functions undercount or average these many fine lines for display on a discrete point graph)? I recall seeing several plots that had fine detail owing to a scale setting that observed a reduced width of spectrum (higher magnification if you will) knowing, of course, all along that the displayed ‘lines’ would not affect internal housekeeping of the spectral response (and energy) that was taking place in the calculations in the computer.

    PS: We might be asked shortly to conduct this exchange on another blog but I have only that one question.

  191. Philip Mulholland
    Posted Feb 3, 2008 at 1:11 AM | Permalink

    cba #188

    It takes more than two geostationary satellites to fully cover the earth’s globe. I like your idea of using a Lagrange point, I assume you would choose L1 as that location gets us back to needing only one satellite. :smile:

    As an erratum to #184. From the perspective of the moon it takes a bit less than 25 hours for the moon to “scan” the earth’s face because of the moon’s eastward progression in its orbit (but you all knew that anyway). :smile:

  192. Pat Keating
    Posted Feb 3, 2008 at 10:01 AM | Permalink

    189 cba
    Interesting.
    How did you handle the overlap of the CO2 and H2O absorption in the 560-600 cm-1 region?
    Because that is in the low-frequency wing of the CO2 15u band, it is quite important when calculating doubling.

  193. cba
    Posted Feb 3, 2008 at 10:11 AM | Permalink

    Jim,

    That laptop is essentially my main computer now so it’s not dead and I can’t afford to let it cook.

    The accuracy of my effort is at present anybody’s guess because I don’t know. I did a great deal of adjustment to it and was using hitran – which is line by line, perhaps accurate to 1% or better – at least before the line broadening stuff goes in. The resolution was done with a 1nm spectral bin from 200nm out to 65000 nm which covers a fair amount. The charts I graphed were simply reduced in resolution by a factor of 65 or 100 so do not accurately reflect the data even though they still appear quite serious in resolution. The goal was to compare the results with TOA satellite measurement graphs to get a feel, but those are always done versus Freq. (in a /cm number) rather than by wavelength which I chose because I’m much more comfortable working in that realm. My workload and fear of frying the laptop have so far conspired against continuing so far and I haven’t even managed to convert the data into a /cm frequency graph for a direct comparison yet.

    If the big bugs are gone, my toy might be good to a few percent. About the only comparison numbers I have are the total outgoing power absorption and the delta power for the doubling from 1976. Total power is perhaps somewhere in the general ballpark and the delta power is probably about right but a bit low when compared to a doubling from 1750 co2 levels – but then from what I’ve seen, I would expect it to be fairly close to that as each doubling seems to provide a decreased increment rather than the same increment value. I’m using fairly stock radiative transfer calculations and I need to work on that area more as I suspect a possible problem there, probably with assumptions/approximations.

    It’s also still clear sky only and radiative only with no scattering. With nearly 50 layers, it’s probably a bit higher resolution than would likely be found in time based models (gcms)

    Even for clear sky only, it’s boiling down to a function of lapse rates and upper temperatures which are not done in the model but rather supplied to it and my guess is that the lapse rate is substantially affected by conduction/convection for most of the matter in the column and the little above that is affected by space weather.

  194. cba
    Posted Feb 3, 2008 at 10:33 AM | Permalink

    Pat,

    very simply. I constructed a wavelength array of bins 1nm wide each. Then, the attentuation factor for the broadened line was summed into each bin at the average value for that bin for each line. That is the line strength at the edge of the bin was averaged with the line strength at the other edge or prorated for each side if the peak was inside the bin. I think for most lines the width is substantially greater than the bin width but it didn’t have to be to get the approximate average absorption over the bin.

    I used a custom program for this where the line database values were read in, corrected for line width and intensity due to T and p and then accumulated the results by wavelength for all molecules of interest based upon their concentrations and did so by layer. The output was an array of optical path thickness by row wavelength for each column layer.

    It then became a matter of feeding in the appropriate spectral intensities for the initial (surface) using the planck equation by unit wavelength and doing the exponential decay function. Of course the planck equation has to be applied to each layer to determine the emissions from that layer as well as determining the extinction of the energy passing through from the surface and lower layers.

  195. cba
    Posted Feb 3, 2008 at 10:54 AM | Permalink

    Phillip,

    I would have guessed closer to 45 or 50 minutes than an hour for lunar orbit time increase.

    The albedo is going to be the reflected/incoming so really only matters for the sunlit side. Hence a full coverage really isn’t that necessasry. Yes, the L-1 point would be the preferred with the moon getting in the way occaisionally.

    Such a satellite could perhaps provide a late early warning for solar events as well as another solar research platform and could provide its TSI information of the incoming as well as measuring the outgoing.

    Of course there’s virtually no comparison between an 800 million $ satellite versus a $ 80,000 telescope when it comes to cost and there’s definitely some benefit to the cheaper versus nothing even though the data from the sat. could provide far more benefit.

  196. pochas
    Posted Feb 3, 2008 at 11:12 AM | Permalink

    cba:

    How many lines of code? What language? What is the data source for emissivities, etc?

    If you could annotate the code and archive it somewhere, perhaps even on this site, it would go a long way
    toward educating many of us on the mechanics of radiation calculations. It would even serve as a partial
    answer to Steve McIntyre’s quest for an exposition of the 2.5 oC CO2 doubling paradigm.

    Hope you have the time for this.

  197. cba
    Posted Feb 3, 2008 at 12:07 PM | Permalink

    Pochas,

    it’s done in visual c++ as a window’s console program. The rest is just excel spreadsheet.

    Line data comes from the Hitran database and dealing with line width and temp. corrections from papers referenced there. Atmospheric data T,P, concentrations by altitude, comes from the 1976 std atmosphere.

    Unfortunately, while it provides power absorption, it does not translate the results into temperature variation. It might be possible to do that by assuming the std atm lapse rate is a given at 1976 levels and then finding the delta power distribution needed to balance the new absorption rates from added ghgs and then doing an energy balance by altitude to adjust for the new temperatures. However, I’m not there yet.

    Depending on what we’re talking about, spreadsheets for excel that can be several hundred mb as well as the raw hitran output file of similar size. Source code is of course more like a few 10s of thousands of bytes and hundreds of lines long but for other than desk checking code, probably isn’t that much of value without having the database. Being written in c implemented on a visual c++ window’s core.

  198. Pat Keating
    Posted Feb 3, 2008 at 12:11 PM | Permalink

    194 cba
    Very interesting.
    I have been doing something similar, but doing each molecular species separately and using fairly coarse layers. What layer thickness are you using?

    What kind of correction did you use for T and pressure? Do you have a reference for that?

  199. cba
    Posted Feb 3, 2008 at 1:11 PM | Permalink

    Pat,

    I can specify in the input files which molecules to use of the 38 or so in the database. It’s part of the input file configuration definitions to my code. Thickness is also an input. In general, I believe it’s 1km thickness 0-25k and 2.5k above out to around 50km and then I think it’s 5k increments above that. It is actually the values used for the 1976 std atm data which I got from GATs on the web as I recall.

    Data from hitran was extracted for std conditions using their javahawks routine and then corrected for T and P of the layer. The run configuration files contain T P and thickness for the shell along with concentrations of the various gases in that thickness. The data is then corrected for the P, T and line broadening. Seems there is also a partition function file from the hitran info that must be used.

    I believe the P and T corrections are done according to the 1996 hitran paper appendix.

  200. Jordan
    Posted Feb 3, 2008 at 2:21 PM | Permalink

    Pat (#159). You say that your equation (Vout = g.Vin+a.Vout) is not a difference equation. But it has no derivative terms, so it is not a differential equation either. It seems to be an elaboration which reduces to nothing more than Vout = K.Vin (K is a constant).

    Rich (#164 and #166).

    I’m afraid that I don’t get your difference equation:

    T(n)- T(n-1) = F + (a-1)T(n-1)

    Try these, they are the same equation in different guises (wish I could present in a neater format):

    T(n-1) = F.(1-a^n)/(1-a)

    or T(n) – T(n-1) = F.a^n

    OK, so question mark the difference equation, but you press on to say:

    To turn this into an infinitesimal from which to derive a differential equation, write
    dT(t) = dt (F+(a-1)T(t)),

    LHS semms fine to me – units of Temperature (T(n) becomes dT). Not so on the RHS A mysterious ‘dt’ pops up unannounced. That changes everything else on the RHS from units of Temperature to Temperature per unit time. Something not quite adding up.

    Without looking for answers here, please take another look at my points in #155. For example, the question (which relates to modelling a simple passive system) :

    If I claimed to have “adjusted the feedback” (parameter ‘b’) by doing no more than altering the time step T, would you be convinced if I also claimed that my RC filter will now amplify the input signal?

    On similar grounds, what about assertions like “if the feedback factor is ‘a’, then amplification is 1/(1-a)”? I see no reason to accept this proposition. It is not needed in the case of my simple RC filter (being in the happy position of knowing enough about my RC filter to complete the model).

    A modelling exercise should be just a matter of expressing the things we know about the process in terms of mathematical equations. If we are not sure what the steady-state climate response is to a feedback or forcing, why should we believe 1/(1-a) as the steady state response? The only thing that this particular result has going for it is that it drops out of an incomplete mathematical model. IOW nothing.

    Anyway, I have already gone further than I had intended on this thread. Apologies for that.

  201. Pat Keating
    Posted Feb 3, 2008 at 4:27 PM | Permalink

    200 Jordan.
    No, you are right, it’s not a differential equation, but it doesn’t need to be. It’s a simple feedback equation (which is what we are talking about), for a general system. If you want, you can use vin(f), vout(f’), g(f,f’), a(f’,f) with complex frequencies f, and you can handle all the more-complex, time-dependent behavior you want.

    Draw the following picture:
    – a block labeled “g”, which represents the system
    – an input to its left side, which is labeled “vin”
    – an output from its right side labeled “vout”
    – now draw a line from where the output line comes out of the system block down a little ways and then across back to left, to the right side of a second block labeled “a” under the system block. Now, connect the left side of this feedback block to the input line at the left side of the system block. That’s it.

    That is the block diagram for the feedback I was talking about. It’s certainly not trivial. (Even for a purely resistive network, the simplest case, there are twice as many independent parameters, and the behavior is quite different for a0, and even more different if a ->1).

  202. Pat Keating
    Posted Feb 3, 2008 at 4:32 PM | Permalink

    201
    The last line was scrambled by the LT symbol — it should read “for a less than 0 and a greater than 0, and even more different if a approaches +1)”

  203. Pat Keating
    Posted Feb 3, 2008 at 4:37 PM | Permalink

    cba
    You and I are doing some similar things. If you are interested in more interchange, let me know and I will let you have my email address, to save Steve’s bandwidth.

  204. cba
    Posted Feb 3, 2008 at 5:57 PM | Permalink

    pat,

    i’m somewhat interested, but for the time being i’m significantly loaded down and i don’t know how long it’s going to continue that way.

  205. Pat Keating
    Posted Feb 3, 2008 at 7:22 PM | Permalink

    cba
    OK, perhaps we’ll talk again at some later date. Good luck with your work.

  206. Andrew
    Posted Feb 3, 2008 at 8:45 PM | Permalink

    Raven:

    maksimovich says:

    Faint sun paradox

    Thanks. Those keywords allowed me to find answer. The 30% weaker sun is an expectation that comes from nature of stars and would not be affected by the new understanding of short term TSI variations.

    But this raises another puzzler: if the climate is extremely sensitive to TSI variations then how can we reconcile the much hotter climates in the past with much lower TSI levels? GHGs? That sounds like a weak explanation there is no clear correlation between GHGs and temps in the historical record. Comic rays? A weaker sun would allow more cosmic rays and therefore cause cooling so that does not work well either. How about different effects for different types of radiation? Did the sun in the past produce different types of radiation?

    Two possibilities, that I know of. Although the sun was weaker then, it also was rotating faster and therefore had a stronger solar wind, as I recall, which means fewer cosmic rays, not more. The second possibility is GHG’s, but the problem with those are that they aren’t correlated with temperatures going back that far.

  207. Leif Svalgaard
    Posted Feb 3, 2008 at 9:50 PM | Permalink

    206 (Andrew):
    And the atmosphere CO2 content was 10-100 times what it is today.

  208. Posted Feb 3, 2008 at 10:01 PM | Permalink

    JEG has an interesting hypothesis on solar/ENSO/climate connections here . As I read it, the main hypothesis is that solar activity affects stratospheric ozone which affects stratospheric characteristics which affects ENSO behavior which affects (mainly Northern Hemisphere) climate, with a lag. An alternate path involves differential heating of the equatorial Pacific when irradiance changes.

  209. bender
    Posted Feb 3, 2008 at 10:27 PM | Permalink

    #208 JEG’s been following bender’s lead I see.

    But hark, what enlightenement through younder consensus breaks:

    Interestingly, the model internal variability does exhibit a weak peak in the millennial band, though it knows of no numerical constant or physical process that could have introduced such a timescale. It is an illustration that the nonlinear atmosphere/ocean feedbacks it embodies can generate variability at unexpectedly long periods.

    It is Robock (1978).

  210. bender
    Posted Feb 3, 2008 at 10:31 PM | Permalink

    Re #208
    Ouch. Hank Roberts won’t like this:

    We conclude that in our model, solar perturbations of mid-range amplitude are sufficient to
    generate persistent SST anomalies of a fraction of a degree on millennial timescales, via the
    thermostat mechanism, reflected in the likelihood of strong El Niño events. Hence, ENSO may
    have acted as a recipient and transmitter of solar influence.

    Richard Sycamore, where are you?

  211. bender
    Posted Feb 3, 2008 at 10:32 PM | Permalink

    I always did like that JEG.

  212. Raven
    Posted Feb 3, 2008 at 11:23 PM | Permalink

    Looks like some of the AGW advocates have finally woken up to the possibility that a PDO shift could result in cooler temperatures for many years. However, I suspect the spin will be that 100% of the warming is CO2 induced and the consensus can’t be wrong on the CO2 sensitivity but any cooling can be blamed on ENSO is masking the GHG warming. The UK Met Office started this repositioning last year by predicting a cool 2008 followed by a rapid warming in 2009.

  213. maksimovich
    Posted Feb 3, 2008 at 11:53 PM | Permalink

    re(Leif)

    Not quite correct CH4,Then a substantial increase in biological Nox production,following the “peak nitrogen crisis”.

  214. Andrew
    Posted Feb 4, 2008 at 7:29 AM | Permalink

    Leif, yes, I forgot to mention that, I suppose!

    Speaking of an ENSO/solar connection, didn’t Kristen Byrnes have graphic which showed that solar minimums almost always coincide with Negative ENSO conditions? Can someone find this? She doesn’t seem to have that anymore.

    maksimovich, peak Nitrogen, eh? Not sure how you can run out of something that makes up most of the atmospere, but okay!

  215. Suzanne
    Posted Feb 4, 2008 at 2:39 PM | Permalink

    I have a copy of it but don’t know how to post it here. What’s your email address? I’ll forward it to you.

  216. Jim Arndt
    Posted Feb 4, 2008 at 2:44 PM | Permalink

    #206 Andrew,

    What time scales are we talking about, millions or 100s of millions. If it is the latter you have to take into account that the Earth was a little closer and rotating faster. The orbit of the Earth was different. Water Vapor and most other GHGs where higher.

  217. D. Patterson
    Posted Feb 4, 2008 at 2:52 PM | Permalink

    207 Leif Svalgaard says:

    February 3rd, 2008 at 9:50 pm
    206 (Andrew):
    And the atmosphere CO2 content was 10-100 times what it is today.

    Except, for one example, part of the Carboniferous Age, when the CO2 was dangerously low like the present conditions and close to the minimums required by the plant kingdom. On the geological time scales, there has never been a persistent and consistent correlation between major changes in the amount of carbon dioxide in the atmosphere and the air temperatures.

  218. Dennis Wingo
    Posted Feb 4, 2008 at 3:06 PM | Permalink

    (#154) Leif

    146 (Dennis): Yes, I have also lamented that there is no money for research, but such is it. The albedo measurements are very important and should be continued. Palle is working on setting up a global network for this.

    I am thinking about putting together a conceptual design of a spacecraft system to improve our knowledge of the non TSI factors related to solar influences. We already have a lot of good spacecraft out there doing good work but we seem to be missing important data. NASA has just announced several new climate instruments for NPOESS and for the NPOESS preparatory mission. Would you be willing to make some suggestions for instruments for such a mission? I have access to some very sophisticated tools for spacecraft development and a white paper would result.

    (I have a good publication record in the space biz, just google my name).

  219. Leif Svalgaard
    Posted Feb 4, 2008 at 3:21 PM | Permalink

    218 (Dennis): I would be gald to help, if and where I can.

  220. Leif Svalgaard
    Posted Feb 4, 2008 at 3:22 PM | Permalink

    219 (me): in case you don’t know what “gald” is, it is Leif fat-finger talk for “glad”.

  221. See - owe to Rich
    Posted Feb 4, 2008 at 3:45 PM | Permalink

    Jordan #200, you wrote

    I’m afraid that I don’t get your difference equation:
    T(n)- T(n-1) = F + (a-1)T(n-1)

    Try these, they are the same equation in different guises (wish I could present in a neater format):
    T(n-1) = F.(1-a^n)/(1-a)
    or T(n) – T(n-1) = F.a^n

    My formulation is algebraically equivalent, but avoids using a^n. In turning it into a differential equation, I am creating a new function whose tangential slope is approximately the slope of the chord of the old function. I’m going to rename the old function N, and use T for the new function, and make a small change to enforce N(0)=T(0)=0.

    Then we have
    N(n) = F(1+…+a^(n-1)) = F(1-a^n)/(1-a), and
    T(t) = (1-exp(-At))F/A

    Both functions start at 0 and go to F/A = F/(1-a), but what about in between?
    Consider the n and t which give N(n) = T(t) = F/(2A).
    That is, n = log(1/2)/log(1-A) and t = log(1/2)/(-A).
    Since log(1-A) ~ -A if A is small, we see that the two functions agree well if A is small, i.e. a is just below 1.
    If A is not small, then the discrete and continuous models show different time evolution, which agrees with a remark you made earlier.

    HTH,
    Rich.

  222. Bill F
    Posted Feb 4, 2008 at 9:25 PM | Permalink

    From JEG’s paper linked above:

    It is noteworthy that although solar forcing has peak-to-peak variations of approx 2Wm-2 even in the strong scaling case, much smaller than the summer-winter insolation difference (peaking at approx 40Wm-2), their effect is disproportionately large

    Wouldn’t this finding be consistent with a higher sensitivity to TSI changes over large time scales than has previously been assumed?

    On a separate note, ENSO as a mediator for solar climate interaction makes perfect sense. The equatorial pacific is the largest open ocean area on the planet, quite far from most sources of dust, and has abundant warm water to produce DMSO. In such an environment, the limiting factor on cloud cover would likely be cloud condensation nuclei, and the equatorial region would be the one most subject to the effects of solar modulated cosmic rays. When cosmic rays make it to the equatorial pacific, they can readily provide cloud condensation nuclei with the DMSO and those clouds result in cooler waters underneath. When cosmic rays are not abundant, there are fewer clouds, warmer waters, etc. This would dovetail nicely with any pattern of la nina activity being more prevalent at solar minimums. And surprise! We currently have a prolonged period of low solar activity and a strong la nina in progress…and reports all from all over the world of cold conditions not seen in at least the last 3 decades…

  223. Dennis Wingo
    Posted Feb 4, 2008 at 11:45 PM | Permalink

    (#220) Leif

    Haha. Great. It is going to take me a while (a month) to get the time to do this. In the mean time there is a lot of activity at NASA in this area and there is a general level of renewed interest in Earth science missions.

    http://www.spaceref.com/news/viewpr.html?pid=24684

  224. See - owe to Rich
    Posted Feb 5, 2008 at 1:30 AM | Permalink

    #181 Knight: you are probably, like me with many a posting, disappointed at the lack of response. So here is one.

    You strike me as a “solar denialist”, someone who is trying to minimize the assessment of the impact of the Sun’s variability on climate. Still, I am always willing to listen to reason and analysis. The first thing you have to do to persuade me that volcanoes contributed significantly to cooling during the Maunder and Dalton minima, is to provide statistics to show that there were more and/or bigger volcanoes during thes periods. A mere list of volcanoes does not do much for me if there is no comparison with other epochs.

    Then, suppose that there really are data to support an assertion that those periods had extra volcanic activity. What would that really mean?

    You propose two effects: first that the volcanoes would cool the climate, and second that the volcanic dust would cause under-reporting of sunspots. I am with you on the first, but on the second I think it serves us ill to demean the observational powers and tenacity of our forebears. It is true that without our equipment they might have found it difficult to notice a sunspot as small as the recent spot 10982. But spotting small spots is not the essence of measuring solar cycles. (Aside: large spots are hard to miss – I saw one group with my naked eye on two successive evenings in 1968.)

    The essence, in my view, is the length of the cycles and the sunspot number around solar maximum. There exist so many observational times and places for this purpose, that it is not credible that volcanic dust (and clouds therefrom) can prevent such studies from being properly conducted.

    Your article does raise one further, intriguing, possibility. Suppose again that your hypothesis is broadly correct. Were those volcanoes an unhappy coincidence, or was there a causal link? If it was causal, what was the nature of causation? The fairly obvious possibility is that changes in geomagnetics from prolonged solar minima affect the Earth’s crust and vulcanism. A second possibility is that orbital effects with respect to the solar barycentre, which some people assert are the main cause of solar grand minima, also affect tidal stresses on Earth. This seems less likely, as they would be a second order effect compared with the Moon.

    Whatever the nature of the causation, if it exists then it provides a fascinating feedback effect to reinforce solar inactivity. And we should start betting on increased vulcanism in the next decade. And we should bet on increased vulcanism being blamed on AGW – Anthropogenic Global Whatever :-)

    Rich.

  225. Jordan
    Posted Feb 5, 2008 at 3:00 PM | Permalink

    Pat – I think I understand your point about functions of complex variables. Perhaps the most familiar examples are the Laplace Operator (‘s’) and its cousin (‘z’) in the discrete world . I agree that transformations can mask the fact that differential or difference equations underly the maths.
    You said in post #121:

    You can, of course, look at it as an iteration — after all
    S= 1+a+a*a+a*a*a+…= 1 + a*S,
    so that S = 1/(1 – a). You can do the same thing with any system with feedback.

    I would like to disagree Pat. We cannot expand Laplace Transformed Functions into power series. This step invalidates the maths and the Inverse Laplace Transform is then no longer defined. That led me to believe you were referring to a difference equation as power series in ‘z’ are fine.
    Further:

    If there is internal feedback within the system to a forcing (vin in my post) then you will get the 1-a factor.

    Similarly, I don’t agree. The steady-state response of a transfer function is determined by the final value theorem. For example, the final value of a/(s+a) = 1 in response to a unit step. There is no (1-a) factor involved.

    Rich – oops (blushing). That first step in your difference equation is rather straightforward.
    That said, I still believe there are issues with the dimensions when the mysterious ‘dt’ appears in a later step in that post. I wonder if the same issue is creeping into your post #221 when you introduce the following equation as a solution:

    T(t) = (1-exp(-At))F/A

    With a little transposition, this can be expressed as the following differential equation:

    dT/dt = F – A.T(t)

    But look at the dimensions of this equation …. temperature-per-unit-time.

    F is not forcing temperature. It is forcing the rate of change of temperature . It is true to conclude (as can be seen in the solution above) that ‘A’ appears on the denominator. Something like ‘A’ absolutely has to appear in the solution in order to resolve into units of temperature.

    If I can be allowed to re-define A, take a look at the following differential equation. It is expressed in temperature:

    A.(dT/dt) = F – T(t)

    This has unity “gain” in response to a temperature forcing F. There is no factor such as 1/A or 1/(1-a) in the solution. If we believe there is amplification (K) to give a scaled-up response to F, the equation can be modified as follows:

    A.(dT/dt) = (K. F) – T(t)

    In this equation, amplification (‘K’) has to be accounted for on its own merits. Nothing to do with feedback.

  226. Jim Arndt
    Posted Feb 5, 2008 at 3:48 PM | Permalink

    Lief,

    Don’t know if this is significant but it reads well. I think it shows a stronger solar correlation to temperature than you do.

    http://cc.oulu.fi/~usoskin/personal/aa7704-07.pdf

    http://sait.oat.ts.astro.it/MSAIt760405/PDF/2005MmSAI..76..969G.pdf

  227. maksimovich
    Posted Feb 5, 2008 at 6:39 PM | Permalink

    re 214 Andrew,yes i meant to say NOX crisis.

  228. Erl Happ
    Posted Feb 5, 2008 at 6:49 PM | Permalink

    Dennis Wingo,
    Need to get a paper to you that, by deduction, links heating in the upper atmosphere to the 1998 El Nino. Interestingly, prior to the El Nino event cloud cover was light and therefore lots of blue sky over the oceans between 45°N and 45°South. As the tropical heating episode developed cloud cover gradually increased at these latitudes over the oceans. A year after the decline in temperatures began, cloud cover increased quite dramatically and the area of relatively blue sky fell to less than half. In the middle of the subsequent La Nina two individual months stand out where clouds anomalously disappeared yielding more blue sky than at any prior point. Changes in cloud cover are similar in all oceans. erlathapps.com.au

  229. Erl Happ
    Posted Feb 5, 2008 at 7:08 PM | Permalink

    re 214 Andrew

    Speaking of an ENSO/solar connection, didn’t Kristen Byrnes have graphic which showed that solar minimums almost always coincide with Negative ENSO conditions? Can someone find this? She doesn’t seem to have that anymore.

    If you do a seach on Erl Happ in Svalgard 1 and 2 you will see lots of references to this idea. You will also see a graph that indicates a swing to La Nina dominance in the coming solar cycle. I too would like to see what Kristen has to say.

  230. Pat Keating
    Posted Feb 5, 2008 at 7:38 PM | Permalink

    225 Jordan

    1. Regarding the summation/expansion issue, you are wrong. Read up on Perturbation Theory in QFT.

    2. Regarding your equation a/(s+a)=1, I’ve no idea what you’re talking about.

    This has become tiresome, and I have lost interest. You are welcome to keep your views, which seem to be too strongly held to listen to reason.

  231. Suzanne
    Posted Feb 5, 2008 at 9:45 PM | Permalink

    re 229 Erl Happ

    Erl, like I mentioned to Andrew-214, I have a copy of Kristen’s graph (removed from her site for unknown reasons) where she said:

    [10-20-07
    Take a look at the graph below. Did anyone at the IPCC even know that minimums in solar magnetic activity are always accompanied by ENSO negative conditions?]

    The graph depicted the simulated magnetic flux over Solar Cycles 18-23 above the Multivariate ENSO Index (1950-2010).

  232. Andrew
    Posted Feb 5, 2008 at 10:02 PM | Permalink

    Suzanne:
    linkcrazy543@hotmail.com
    Thanks so much!
    Jim Arndt, I haven’t actually done the calculations, but that would need to be taken into account, to. The paper I’m refering to is here:

    http://xxx.lanl.gov/abs/astro-ph/0306477

    I haven’t gone over it, so I don’t know if it is taken into account.

    Rich, from what I can tell, Leif is a solar scientist with some interesting and potentially very troubling (for all of climate science) results. As 2 and some other posts would seem to suggest, Leif isn’t to sure what the climate implications of his work are, though other posts indicate to me he has some ideas what it could mean. I’m chewing the information over.

    BTW, Leif, I’ll try to answer some questions regarding direct solar forcing tomorrow, if I can. As for indirect forcings, the literature is very obscure to me, so quantification is tricky!

  233. Andrew
    Posted Feb 6, 2008 at 7:44 AM | Permalink

    Okay, so as I understand it, your reconstruction:

    http://www.leif.org/research/TSI-LEIF.pdf

    http://www.leif.org/research/TSI%20(Reconstructions).xls

    From the caption of figure 5 ISPM (Independent Summary For Policy Makers) Direct Solar Forcing is irradiance divided by 4 times .7 (for albedo). Thus between 1750 and 2000, your reconstruction means just .03 W/m2 direct forcing. If CS is between 1.5 and 4 for doubling CO2 (IPCC), then 1.5/3.8 W/m2=~.4 and 4/3.8 W/m2=~1, which means that using your reconstruction, direct forcing is responsible for between .012 and .03 degrees of warming (actually, since it takes time to reach equilibrium, the real response is 80% of this). By contrast, The IPCC’s numbers here:

    http://www.grida.no/climate/ipcc_tar/wg1/251.htm

    indicate solar effects ten times greater, with 67% uncertainty.

    Note that if we assume that the effects of the eleven year cycle are damped compared to the long term trend (ie, a moving average would be better, becuase 1750-2000 might just indicate a low in the solar cycle versus a high in the solar cycle) then 2000 was actually .02 W/m2 above the 1750 value, corresponding to even less direct forcing.

    It seems to me we have a problem. If the Sun was thought to be able to pull us out of the LIA, how, if the direct forcing is so weak? Perhaps some amplifiers? (Note that feedbacks are implicit in the sensitivities of 1.5-4 so that isn’t the answer, unless, as you already mentioned, CS is even higher than we suppose.)

    Note! The highest values above 1750 are .1 (unsmoothed, in 1837) and .05 (smoothed, a tie between 1783 and 1954). 2007 value is -.11 W/m2 and most recent smoothed value .01 W/m2.

  234. Erl Happ
    Posted Feb 6, 2008 at 8:00 AM | Permalink

    231 Suzanne, I would appreciate a copy and also the website address where it appeared. The coincidence of La Nina cooling events with the minimum in the solar cycle, a minimum in geomagnetic activity and also possibly a minimim in irradiance of a wave length shorter than the visible spectrum can not be accidental. It must be connected with upper atmosphere conditions that influence the degree of penetration of solar radiation through the atmosphere and therefore the level of solar radiation incident at the surface. email is erl@happs.com.au

    Have you seen todays report at http://www.lowell.edu/blog/?p=89 SORCE meeting

    the 0.1% variation in solar brightness over the course of its activity cycle does not impart sufficient radiative forcing to account for climate change since about the mid 20th century; another, presumably anthropogenic, signal must be present. Here’s the critical twist: the spectral distribution of the Sun’s cyclic variability is quite different depending on where you look in the spectrum. In particular, the ultraviolet variability is much, much larger than the total variability — a few tens of percent — and the UV radiation significantly affects the stratosphere, in ways that are poorly understood. The next avenue for real progress in Sun-climate relationships lies in breaking down the total irradiance variation into its spectral components, and beginning to understand what effects they have on climate. We are just now starting to assemble the observations, which must be done from space, to address this issue.

    In fact, you can find Claus’s results in Figure 4 of an article on our Solar-Stellar Spectrograph web site. Claus finds that right now, near the dead minimum between solar cycle 23 and 24, the Sun is fainter than it was between the 1996 minimum of cycles 22 and 23. This is hugely important, since identifying changes in the solar brightness at minimum, when the Sun is mostly or entirely free of “active” features like sunspots, is perhaps the critical observation for determining whether the Sun is growing steadily brighter or fainter on the long term. Even more interesting, my preliminary analysis of my 2007 spectroscopic observations of the Sun show the same thing…it looks like the current minimum is lower than the last one.

  235. Pat Keating
    Posted Feb 6, 2008 at 8:30 AM | Permalink

    233
    If I understand Leif correctly, he believes that orbital effects on the insolation of Earth are much larger than TSI effects, and that they are large enough to explain observed temperature changes.
    That is my understanding, but of course Leif can speak for himself…..

  236. Jim Arndt
    Posted Feb 6, 2008 at 10:21 AM | Permalink

    Hi,

    Andrew another influence to take into account is the cosmic ray flux. It seems that as solar activity lowers it allows more cosmic ray to penetrate the atmosphere an cause more clouds thus making albedo higher. See here.

    http://spacecenter.dk/research/sun-climate/

  237. Jim Arndt
    Posted Feb 6, 2008 at 10:29 AM | Permalink

    Hi,

    Andrew another thing about the early Earth is the are many factors that are vastly different from today. You must take into account the composition of the atmosphere, orbit, rotation, asteroid / comet bombardment and distance from the sun.

  238. Andrew
    Posted Feb 6, 2008 at 10:41 AM | Permalink

    Thanks Jim, but as I said (I think), I’m not sure how to actually quantify these indirect effects. The literature is pretty obscure to me. Let me know if you have a formula for coverting Cosmic Ray Counts into W/m2 radiative forcing! The other thing is, of course, that my post assumes both that Leif’s Recon is better than the alternatives (I’m sure Leif thinks so) and that climate responds the same way to all forcings (not necessarily so, but a good approximation, I think).

  239. Jim Arndt
    Posted Feb 6, 2008 at 11:04 AM | Permalink

    Hi,

    Andrew, Here is the experimental data for the comics ray flux and clouds.

    http://spacecenter.dk/research/sun-climate/Scientific%20work%20and%20publications/resolveuid/1f09645cb14320a460fc697de6be993b

  240. GW
    Posted Feb 6, 2008 at 11:36 AM | Permalink

    Perhaps one of you who received the graph from Suzanne would be able to and kind enough to repost it here for the rest of us ?

    Thanks

  241. Andrew
    Posted Feb 6, 2008 at 11:45 AM | Permalink

    Thanks again, Jim. I’ll see what I can do from there.

    GW, sorry, haven’t checked if I have it yet!

  242. See - owe to Rich
    Posted Feb 6, 2008 at 1:38 PM | Permalink

    Andrew #238

    Thanks Jim, but as I said (I think), I’m not sure how to actually quantify these indirect effects. The literature is pretty obscure to me. Let me know if you have a formula for converting Cosmic Ray Counts into W/m2 radiative forcing! The other thing is, of course, that my post assumes both that Leif’s Recon is better than the alternatives (I’m sure Leif thinks so) and that climate responds the same way to all forcings (not necessarily so, but a good approximation, I think).

    The point about Cosmic Ray counts is that they (supposedly) affect albedo, so if one knew the albedo effect then one could easily calculate the implied forcing. But if you are going to call this a forcing, then “climate responds the same way to all forcings” is clearly incorrect, unless you believe that increased CO2 has the same effect on clouds as sun-modulated cosmic rays. And I don’t.

    Rich.

  243. Andrew
    Posted Feb 6, 2008 at 1:52 PM | Permalink

    Rich, I meant their is approximately the same temperature response for W/m2 of radiative forcing (theoretically) but as you correctly point out, some of those sensitivities contain cloud feedbacks. Presumably cloud feedback on CR induced warming would be the same as a equal amount of cloud feedback from CO2 enhanced warming of similar magnitude (which would make it tricky to separate CO2 and CR effects on clouds). So my question was really, what is the magnitude of the albedo effect, in W/m2? As I understand it, its quite large.

  244. See - owe to Rich
    Posted Feb 6, 2008 at 2:36 PM | Permalink

    Andrew, yes I agree that it would be nice to know the magnitude of the albedo effect. In fact, you may have noticed Leif and Dennis Wingo discussing new instruments to measure this. That is, to measure the change in albedo – the W/m^2 falls straight out of that.

    But no, I do not agree with “cloud feedback on CR induced warming would be the same as a equal amount of cloud feedback from CO2 enhanced warming of similar magnitude”. As far as we know, cosmic rays do not induce a warming at all – they induce nuclei around which vapour can condense. So the mechanism is completely different from CO2, whose warming may either increase or decrease cloudiness – and the jury still seems to be out on that one.

    Hope this explanation, from a mathematician rather than a physicist, is both accurate and helpful.

    Rich.

  245. See - owe to Rich
    Posted Feb 6, 2008 at 2:50 PM | Permalink

    Happ 234, Happ 234, by the left Happ 234…sorry I couldn’t resist it ;-)

    Claus finds that right now, near the dead minimum between solar cycle 23 and 24, the Sun is fainter than it was between the 1996 minimum of cycles 22 and 23.

    That is fascinating. Of course, as you can read on CA Forum->Surface Record, I am heavily into solar cycle lengths, so I am not surprised that cycle 22/23 minimum, following the crazily short 22, was hotter than the current (still probably a good year to go) 23/24 minimum, following the crazily long 23. However, this doesn’t square with Leif #1’s early postings on “all minima give the same TSI”.

    I await Leif’s comments with interest – possibly it is a mere misunderstanding about the magnitude of diffences between solar minima.

    Rich.

  246. cba
    Posted Feb 6, 2008 at 6:57 PM | Permalink

    244, 243 Andrew Rich,

    To gain a handle on w/m^2 for albedo, recognize that it is currently accepted and measured that earth’s albedo over the solar insolation wavelength range is right at 30% or 0.30. TSI average value at the top of the atmosphere is 1367 W/m^2 (give or take a few W). That puts the total albedo effect at 1367 * .3 = 410 W/m^2 of energy being reflected out into space instead of being absorbed in some fashion. Averaging this over the surface area comes out to about 102 W/m^2 less energy coming in than the 342 W/m^2 averaged value over the whole surface.

    That means a change in the albedo is going to have a significant change in the total energy budget. Most of the planet surface has significantly less albedo than 30% with oceans being perhaps less than 5% and that’s 3/4 of the globe. Snow is about the only thing that can contribute greater than 30%. Clouds can contribute more than 30% and are undoubtedly responsible for much of that 30% albedo.

    Clouds also are a double edged sword. They reflect sunlight in the day but they also seriously absorb IR at night. They can emit as well but at a much lower temperature than the surface. Net result is contributions to both increase and decrease.

    If one were to consider just cloud cover, there are numerous types of clouds to deal with that are going to have different parameters such as transparency or opaqueness and altitude. I think clouds can have as high as 80% reflectivity. Coverage of clouds tends to be around 40-50% of the globe. This could account for as much as 26 of the 30% of our reflectivity (note i’m working totally from memory and half of that is based on my own simple analysis and may be somewhat off but I think it’s in the ballpark).

    A simple albedo change will be responsible directly for the percent of W/m^2 coming in. With an average of 342 and a 1% change in albedo – you’re taking out of the system 1% of that 342 (3.42 watts/m^2). A change of 1% in cloud cover (assuming clouds that are highly reflective and not very translucent) is going to have significantly more impact than that as the albedo for those are much higher than the surface.

    Concerning if a forcing of 1w/m^2 is a forcing and is identical, I think not. If you increase ghg concentrations, there is more power absorption, but there is also more emission because of increased emissivity. If you change the tsi, there is an increase of power with no increase of emissivity or emissions. If you change the makeup of the TSI, the location of where that power gets absorbed could change. Pushing into the IR means more absorption higher up. Pushing into the uV could affect higher up areas with molecular changes that affect absorption at other wavelengths and also more power into the ocean beneath the surface by many feet. Hopefully, these is somewhat persuasive even if it’s not quantitative or a full logical proof.

  247. Suzanne
    Posted Feb 6, 2008 at 8:22 PM | Permalink

    Andrew 232 – I’ve emailed Kristen’s graph to you.
    Erl-234, same to you as well. Btw, thanks for today’s report of SORCE meeting. Much appreciated.
    Also, if either of you can figure out how to repost the graph here for others, like GW-240, by all means.

  248. Sean
    Posted Feb 8, 2008 at 4:22 AM | Permalink

    Along with many on this list, I find it intuitive that the greatest nearby source of energy in our lives – the Sun – should play a role in climatic variations here on Earth.
    Similarly, it seems obvious that, on a planet dominated by a thermally active fluid – the ocean – many of the observed changes in air/surface temperature should be closely interlinked with the status of the great currents, such as the AMO, the PDO, El Nino/La Nina, etc.
    The crux is, of course, the ability to conceive of a scientifically testable mechanism which provides a link between the two presumptions, a requirement which is at the root of much of the debate on this fascinating blog.
    I may have missed any discussion of them, but would someone like to comment on the following two papers which did try to explain how the solar cycle could directly influence atmospheric temperature and hence climate?
    Shindell, et al
    Labitzke

  249. solarphile
    Posted Feb 8, 2008 at 6:00 AM | Permalink

    Towards a better representation of the solar cycle in general circulation models
    The Influence of the Solar Cycle and QBO on the Late Winter Stratospheric Polar Vortex

  250. Andrew
    Posted Feb 8, 2008 at 6:19 AM | Permalink

    Sean, UV can account for only for only twenty percent of the variance statistically (which is why the CR hypothesis is generally more likely than the Ozone/stratospheric chemistry one. Still interesting stuff, as 20% of the variance is more than currently placed in the models. From the Jan Veizer paper I posted above:

    Considering that statistical evaluation
    of 20th century data shows that
    solar UV radiation may account for only
    about 20% of the variance in surface
    temperature data (Foukal, 2002),

    (Note, however, that this is the surface record we are talking about.)

    BTW everyone, I had the image already to go and…Climate Audit seems to have gone down, or at least I couldn’t get here. No access to my photobucket account from here in my school library.

  251. solarphile
    Posted Feb 8, 2008 at 6:25 AM | Permalink

    @250 Does your 20% include all the appropriate amplification mechanisms?

  252. Andrew
    Posted Feb 8, 2008 at 6:29 AM | Permalink

    solarphile, Just UV, but not Cosmic Rays. TSI can acount for as much as 80% of the variance statistically, I think (though Leif’s results probably change that). This should not be taken to mean that UV actually accounts for 20% of the warming, just that that is the amount of the variance it can account for.

  253. maksimovich
    Posted Feb 8, 2008 at 1:52 PM | Permalink

    re 252 Disregard thinking of the earth sun climate connections as a response to the “heat engine” regime,and consider the causal mechanisms that move the terrestrial climate to a cooler state.

    Then the UV-GCR-OZONE complex and its response becomes more apparent.

  254. Posted Feb 8, 2008 at 1:56 PM | Permalink

    Leif, Is it a hard fast rule that cornal holes of one cycle have to end before the next cycle can start??? Or is this some rule of thumb?

  255. Gunnar
    Posted Feb 8, 2008 at 3:01 PM | Permalink

    Back on Dec 21, in response to a request from Leif, I predicted

    climate: the global spatial average temp anomaly will drop to at least -.5 deg by the end of 2009

    http://www.climateaudit.org/?p=2470#comment-181154

    It has now dropped to -0.05. 0.76 was the maximum since 1979, which occured in April of 1998.

    I say that contrary to AGW, there is no linear trend upwards, but only occasional step changes that correspond to measured solar output changes. The last 10 years have been above average. Since the peak in 4/98, there have only been 9 months below average: 3 in 99, 3 in 2000, once in 04, once in 06, and now jan 08. 9/120 = 7.5%. I suspect that over the next 10 years, we’ll spend a lot more time below average.

  256. JimP
    Posted Feb 8, 2008 at 3:39 PM | Permalink

    I’d like to make a guess as to when the next sunspot cycle is going to start. It is based on Solaemon’s spotless day’s data. This is just for fun. Most astrophysicists have been wrong on predicting the start of the next cycle. So, what the heck, I’ll give it a try. My excel spreadsheet is here.

    The first assumption is that the sun has shifted its pattern of the start of the sunspot cycles from a delta of 33 +/-5 months to a delta of 66 +/- 4 months. Cycle 23 had a delta of 25 and did not follow either pattern. If cycle 24 is as short as cycle 23 then it could have a start date of delta 58 or November 2008. If cycle 24 follows the traditional 66 +/-4 the start date would be between March of 2009 and November 2009.

    I separated the number of spotless days by cycle and compared each one of them to the upcoming cycle 24. Cycles 10, 11, and 14 have much slower starts and an average delta of 68. Cycle 13 and 15 appear to be the most similar and have an average delta of 62. This indicates cycle 24 have a delta of 62 which is at the low end of Solaemon’s 66 +/- 4 estimated delta.

    Cycle 24 has a lower number of total spotless days at this point, delta 49, compared to 13 and 15. This might indicate that cycle 24 will have a longer delta than 13 and 15. Cycles 10, 11 and 14 had very low totals at this point.

    So I think that cycle 24 will start in May of 2009 and will have a delta of 64.

  257. Jim Arndt
    Posted Feb 8, 2008 at 5:09 PM | Permalink

    Hi,

    JimP #256
    I think your close to it. I also think it will have 50 to 60 spots total and not the low end of 90 like NASA has predicted, but we will see if they change their number in March when they are to meet again. Who knows it could also be a grant minimum.

  258. Suzanne
    Posted Feb 8, 2008 at 5:14 PM | Permalink

    Andrew-250, was unable to post yesterday at CA as well. Site claimed it was “busy” – probably a tech glitch or some such thing. Anyway, If you can’t post the graph here, you can probably recreate it by placing the simulated magnetic flux image for solar cycles 18-23 above the Multivariate ENSO Index and simply draw in the lines that correlate with minimums in solar activity and ENSO negative conditions between the two graphs.

    Btw, the following paper and article may be of interest:
    MODELLING SOLARMAGNETIC FLUX AND IRRADIANCE DURING AND SINCE THE MAUNDER MINIMUM
    K.F. Tapping1, D. Boteler1,2,3, A. Crouch1, P. Charbonneau1, A. Manson1,
    H. Paquette2

    Abstract.
    Using sunspot number as input, we construct a model for the evolution of magnetic
    flux from strong elements in active regions to weak remnants during the solar cycle and
    thence estimate the historical record of irradiance from the Maunder Minimum to the
    present. The magnetic flux model is a fragmentation cascade starting with strong-field
    elements, which fragment into weak-field elements and then into a background field. The
    model indicates the mean total irradiance during the Maunder Minimum was between 1
    and 1.5 Wm−2 lower than it is at present.

    Also here’s a very timely Popular Mechanics article: Sun Dimming in Fight for Anti-Ice Age Solar Telescopes – Canada National Research Council – Popular Mechanics

  259. JimP
    Posted Feb 8, 2008 at 6:04 PM | Permalink

    Another interesting aspect of my prediction is that Timo Niroma has predicted possible minimums to be between 2008.5-2008.6 and 2009.4-2009.6 based on Jupiter’s cycles around the sun. The 2009.4-2009.6 lines up very well with the spotless days theory.

  260. Andrew
    Posted Feb 8, 2008 at 6:15 PM | Permalink

    Okay here it is:

    I attempted to replicate it using the Southern Oscillation Index (can’t seem to find ENSO!) But no luck. Frequently, Solar minimums do coincide with positive SOI (La Nina) but not always:

    Still, I’m not sure how well ENSO is represented by SOI.

  261. Dennis Wingo
    Posted Feb 8, 2008 at 7:58 PM | Permalink

    To the folks doing the albedo studies with the global telescope network.

    Please contact me via my email at wingod at earthlink.net. I have been talking to the folks at the U.S. Air Force academy and they are putting together a new 1.4 meter telescope for Near Earth Object studies. However, due to the limited time needed for the albedo study it appears that they would be more than willing to support the albedo network. I just need you to contact me and I can put you in touch with the USAF academy folks doing this new scope.

    Thanks

    Dennis

  262. JimP
    Posted Feb 8, 2008 at 8:01 PM | Permalink

    Andrew,
    I looked at it from a little bit different perspective. 1st I compared the ONI index to the cosmic rays using a 3 year average to match the ONI index. The El Nino’s and La Nina’s seem to lag the cosmic rays by 20 months but not always.

    The RSS temps almost seem to be controlled by the ONI. I was reading the other day how some estimate the AMO has contributed up to 30% of the current temperature increase of the last 10 years. If that is true, the oceans drive the RSS temps and the cosmic rays seem to regulate the ocean cycles.

  263. Suzanne
    Posted Feb 8, 2008 at 8:12 PM | Permalink

    Andrew-260, o.k.- looks good.

    Still, I’m not sure how well ENSO is represented by SOI.

    Hmmm… now do I dare take you to the forbidden land? :) http://www.john-daly.com/sun-enso/sun-enso.htm

  264. EW
    Posted Feb 9, 2008 at 2:46 PM | Permalink

    Hey, where did you guys get the SOI or other ENSO indices? I got 12 monthly values plus a yearly one from Hadley center from 1866 to 2004 (I wondered why their year has 13 months ;-) ) and then there are some from approx, 1955 to present from Australia, also monthly, but the values aren’t exactly identical to those of Hadley, so what to do? Anyway it is rather a pain in the * to convert these to Excel. I did that manually and hated every second of it.

  265. Posted Feb 9, 2008 at 3:41 PM | Permalink

    254 (dscott): just a rule of thumb. There are no hard rules when it comes to solar cycle behavior.

    260 (Andrew): what are we supposed to see in the Figures you show? I can’t see any relationship. My ‘rule’ of thumb here is that if I cannot see it by eye, then it ain’t there to the degree that I need to worry about it.

  266. Posted Feb 9, 2008 at 3:57 PM | Permalink

    I kinda promised to report on the SORCE meeting, but the Flagstaff crew have alreeady done a good job [as was mentioned in this blog]. See
    http://www.lowell.edu/blog/?p=91 and previous postings there.

  267. Andrew
    Posted Feb 9, 2008 at 4:04 PM | Permalink

    EW, mine’s from the Australian Bureau of Meteorology:

    http://www.bom.gov.au/climate/current/soihtm1.shtml

    I just made them annual averages. As for ENSO, I have no idea where it comes from.
    Leif, the idea is that during solar minimums, there are almost always La Ninas (at least, in the first graphic). This may or may not be a massive coincidence.
    Suzzane, thanks, it looks to me like SOI is an Okay but not perfect stand in.
    JimP, interesting graphs. Thank you. Where did you get the cosmic rays? As a warning, moving averages sometimes shift data forward in time (in my experience) putting peaks after they actually happened.

  268. Posted Feb 9, 2008 at 6:42 PM | Permalink

    Ref 265 Leif Said:

    260 (Andrew): what are we supposed to see in the Figures you show? I can’t see any relationship. My ‘rule’ of thumb here is that if I cannot see it by eye, then it ain’t there to the degree that I need to worry about it.

    In a chaotic system eye balling can be misleading. Sea surface temperatures should follow changes with a predictable lag, but decadal oscillations have a different time frame that complicate things beyond eye balling. This is where A. A. Tsonis has a good part of the puzzle solver started. Just comparing a couple of curves looking for something that stands out isn’t going to cut it.

  269. Erl Happ
    Posted Feb 9, 2008 at 7:09 PM | Permalink

    Leif,
    A few questions that may be worthy of your consideration:
    I see at the SORCE meeting some mention of the possibility that the changing spectrum of solar radiation across the cycle may influence earthly climate. Does this argument appear to have legs? Amongst the references the following seemed to be the most direct.

    Mean Circulation
    Terrence R. Nathan [trnathan@ucdavis.edu] and John Albers, University of California,
    Davis; and Eugene C. Cordero, San Jose State University, CA.
    An ever-increasing body of evidence shows that changes in solar spectral irradiance(SSI) over the 11-year solar cycle (SC) can produce changes in stratospheric ozone. Changes in stratospheric ozone can in turn produce changes in planetary wave drag (PWD) via longitudinal variations in ozone heating, which was recently expounded upon in a paper by Nathan and Cordero (2007, JGR-Atmospheres). Because SSI-induced changes in PWD may have potentially far-reaching consequences for the global circulation, including the zonal mean flow, the Brewer-Dobson circulation and stratosphere-troposphere communication, it is important to understand the connection between SSI and PWD. In this study we employ analytical and numerical models of the extratropical atmosphere to examine the connection between SSI and PWD. The models couple radiation, ozone and dynamics and provide in a relatively simple but self-consistent way the means to explicitly identify the pathways that connect changes in SSI to the wave-driven zonal-mean circulation. The sensitivity of the stratospheric circulation, particularly stratospheric sudden warmings, to changes in SSI associated with the SC is addressed.

    Question 2: Why should the interception of short wave radiation stop at the stratosphere?

    Question 3: What wavelengths are capable of directly heating the atmosphere and has the differential heating effect been quantified?

    Question 4: Is changing humidity in the tropics (as observably happens when cloud cover increases) likely to attenuate infrared radiation received at the surface?

  270. maksimovich
    Posted Feb 9, 2008 at 8:07 PM | Permalink

    Erl First one must understand the causal mechanisms from observations ,over extending to fast adds unnecessary complications that tend to promote confusion.

    That said the abstract you pose in your post is in the relevant area of scientific topics for further research. Jack Eddy in the Nasa LWS programme suggests the following.

    RECOMMENDED TOPICS FOR EMPHASIS
    IN A SERIES OF INFORMAL SUN-CLIMATE INSTITUTES
    • Influence of solar variability on the oceans and ocean circulation
    • Influence of solar variability on the dynamics of the lower atmosphere,
    including possible effects on Hadley cell and Walker cell circulation
    • Interpreting cosmic nuclide data in terms of solar activity and irradiance
    • Solar-driven changes in stratospheric ozone and their climatic impacts
    • Solar influence on internal variability modes in the atmosphere, such as
    ENSO, NAO and the Pacific Decadal Oscillation
    • Comparisons of solar variability, stratospheric ozone variations, and
    climate in the 25-year period for which direct solar measurements are
    available

    We can understand this more easily for example with say the interesting paper Rozema et al science 2002.

    Measurements in the tropics suggest a change of ~6% of the total ozone column
    during the 11-year solar cycle from solar low to solar high (9, 10), whereas estimates from satellite data indicate a column variation of ~2 to 3% (11, 12). Continuous monitoring of spectral UV irradiances at Earth’s surface has begun only in the last decade. The collected data are still too noisy to confirm the 11-year variations, but they do show the predicted anticorrelation of UV-B irradiance with ozone column amounts(13,14)Historical reconstructions of solar spectral irradiance based on analyses of solar cycles (7) suggest that long-term UV-C variations exceed the 11-year solar cycle amplitudes by a factor of ~2 (see the second figure). During the 17th-century Maunder Minimum (1645 to 1715), UV-C irradiance may have been 3 to 4% lower than today (7). Reduced solar UV-C radiation during the Maunder Minimum epoch of low solar activity suggests that ozone concentrations may also have been lower and surface UV-B radiation higher.

    The jury is undecided with the last point(uv c ) ,however increased GCR will have the same residual effect on ozone and indeed this is evident in the 14 c proxies.

    GCR of course as an activator in catalytic reaction-diffusion chemical mechanisms and reducing SAT being well understood.

  271. ejmohr
    Posted Feb 10, 2008 at 12:28 AM | Permalink

    This is a very interesting thread. After a few days of reading and thinking about the arguments presented here I have a few comments.

    Not that long ago, when I was at university solar irradiance was known as the solar constant. Suggesting the sun had any effect on climate would usually get you a few laughs but that was about it. The mainstream viewpoint was that past climate changes had to have another cause and the main suspect was insolation changes via Milankovitch Cycles.

    The first book I ever read that suggested the sun might have an influence on climate was Hoyt and Schatten’s “ The Role of the Sun in Climate Change.” This book carefully reviews the data and leaves the reader with no firm conclusion since it paints a complex picture of the arguments for and against solar influence.

    So, if Leif is correct, we are looking at what is pretty much a solar constant and, as he has said, either climate is very sensitive to small changes in irradiance, or there are other mechanisms that effect climate.

    Having said that I found Hoyt and Schatten’s graphs on solar like stars to be quite interesting.

    Perhaps this information is now outdated, but the book shows the Calcium II index for solar like stars varies much more than what we have observed on our local star. If this is true, the sun has only been in the active part of the observed range for similar stars.

    In addition to this, the cycle length on solar like stars shows an inverse relationship between cycle length and activity.

    Hoyt and Schatten also mention the the greater variance in shorter wavelengths over the solar cycle and how this may complicate any climate connection. My thought was always that since UV photons are more energetic if somehow that energy was not absorbed entirely in the higher atmosphere but transferred to the lower atmosphere this could have some effect, but that is just a guess.

    I suppose if the present cycle continues its slow wind down we may have a chance to observe and learn what this means, and perhaps how it effects climate.

  272. cba
    Posted Feb 10, 2008 at 6:17 AM | Permalink

    Erl 269,

    I’ll try to provide some thoughts on your questions based on observations over my modeling efforts.

    Question 2: Why should the interception of short wave radiation stop at the stratosphere?

    It depends on the concentrations of absorbing molecules at the altitude. uV a and uV b have little in the way of absorbers lower down. Also, scattering of light is wavelength dependent (blue skies) so that shorter wavelengths are preferred. The demand for suntan oil is indicative of some levels of uV making it to the surface. Shorter uV is absorbed rather well though. At the surface, shorter wavelengths do much better at getting through below the ocean’s top skin as red tends to be filtered out quickly.

    From my early very basic observations of effects on absorption, it seems that total solar insolation tends to deposit its absorbed energy a bit more uniformly through the atmosphere as compared to the IR – due to the IR being absorbed extremely rapidly at some wavelengths.

    Question 3: What wavelengths are capable of directly heating the atmosphere and has the differential heating effect been quantified?

    All wavelengths coming in convey energy and those which are absorbed will most likely be converted to heat unless reradiated out prior to being converted into heat (kinetic energy of the molecule). Also, emissions of radiation depend on the temperature where the emissions are occurring and from the emissivity (same as the absorption) of the material. What must happen at any point is that a balance must be reached for the absorbed power (incoming energy rate) and the emitted power. The incoming includes absorption, convection, etc, all combined. For a given temperature and molecular concentrations that create the emissivity, there will be a certain amount of radiation. If there is not enough incoming power being absorbed or too much, the temperature must adjust. If there is an increase in absorbing molecules (rise in co2) then there is also an increase in emissivity and so there must be an adjustment to temperature as energy is mostly coming from lower down and reradiation emits the same amount outboud as back down so an increase in emissivity for a small section will be cause more to be radiated (at the same temperature) than is absorbed due to the increase in absorption.

    The lapse rate for the atmosphere must be the result of conservation of energy for all mechanisms of transfer as there isn’t much heat capacity in air, especially at lower pressures. There is also considerations of the ideal gas law with pressure and volume affecting temperature that adds more complexity to the understanding but that is related to the adjustments reached and isn’t part of the radiative balance once reached.

    To make it more fun, at least in the lower parts of the atmosphere, convenction is quite significant and minor variations in radiative can be compensated for by variations in convective.

    Question 4: Is changing humidity in the tropics (as observably happens when cloud cover increases) likely to attenuate infrared radiation received at the surface?

    It is likely to. Over half the incoming solar energy is IR and uV, mostly IR. Clouds are going to reflect significant amounts of visible as well, I think allow through maybe 10% of the energy. At IR wavelengths, water vapor is a super absorber in the bands of interest and even solid and liquid h2o tends to absorb IR fairly well over many wavelengths. Since clouds tend to be fairly cool, that which is absorbed tends to radiate at a lower rate, bringing about the joys of convection for rising energy. Fortunately, coulds are highly reflective in the visible where much of the incoming energy is occuring. I think of the 30% albedo, clouds are between 20 and 26%, leaving the surface to provide only the remaining 4 – 10%.

  273. MGW
    Posted Feb 10, 2008 at 9:40 AM | Permalink

    Re: 260 and related:

    The van Loon et al. ’06 paper, describing the coincidence of peak QD solar output w/ “cold events” in the east tropical Pacific presents a view that might be of interest. The paper does not find correlation with the opposite – i.e. no evidence for solar minima w/ El Nino.

    On the other hand, recent work by White and Liu (two papers in ’08) build on the White ’06 work (#182). They find resonant excitation of the ENSO and QBO signals related to the QDO(quasi-decadal oscillation) of the solar signal – due to constructive interference of harmonics. Modeled dynamics show the timing of El Nino/La Nina pairs and La Nina/El Nino pairs to be largely governed by the quasi-decadal external forcing. The conclusion is not unlike the van Loon conclusion – a point emphasized in the paper; nor is it unlike the suggestion made in # 260 (and others) in this thread.

    van Loon et al ’06 Coupled air-sea response to solar forcing in the Pacific region during northern winter, JGR 112 D02108, doi: 10.1029/2006JD007378

    White and Liu ’08a: Resonant excitation of the quasi-decadal oscillation by the 11-year signal in the Sun’s irradiance. JGR 113, C01002, doi:10.1029/2006JC004057

    White and Liu ’08b, Exciting ENSO with the quasi-decadal wave packet response to the 11-year signal in the Sun’s irradiance, to be submitted to the JGR.

  274. Richard Sharpe
    Posted Feb 10, 2008 at 12:41 PM | Permalink

    The demand for suntan oil is indicative of some levels of uV making it to the surface. Shorter uV is absorbed rather well though.

    Actually, skin color is a better proxy for this.

    It is very clear from human skin color that much more UV energy gets through around the equator than at higher latitudes.

  275. cba
    Posted Feb 10, 2008 at 2:01 PM | Permalink

    Could be a problem with proxies there. It might be that the cold weather is affecting both the sale of suntan oil and the ability of the natives to get out and enjoy the sunshine and beach scene. After all, nose, cheeks, and forehead don’t require much compared to legs, arms, backs and bellies. Then again, the comments weren’t specific to time of year or lattitude.

  276. kim
    Posted Feb 10, 2008 at 7:15 PM | Permalink

    The sun is very sultry and we must avoid its ultry-violet rays, as the song says.

    H/t No. Coward and Plum.
    ================

  277. Andrew
    Posted Feb 10, 2008 at 10:23 PM | Permalink

    Richard, cba don’t forget to correct for Ozone depletion ;).

    I know that given that it is based on reconstructions that Leif is obviously disputing here, but I recall this breakdown of the make up of TSI over time:

    I’m wondering if Leif could estimate what portion of the TSI in his reconstruction is of each type? Is there some constant ratio? Is the graph actually completely bogus?

  278. Dennis Wingo
    Posted Feb 10, 2008 at 11:01 PM | Permalink

    Leif

    Here is an extremely interesting graph that someone here (I forgot who) linked to a site that talks about spotless days vs solar cycle length. At the very bottom there is a hugely interesting graph of spotless days vs smoothed sunspot number. There seems to be an inverse correlation between (as one would expect), the smoothed sunspot number at height of the cycle. What interests me is that this graph would tend to refute your statement that cycles before 1950 have not been properly accounting for sunspots: Comment?

  279. Raven
    Posted Feb 11, 2008 at 12:02 AM | Permalink

    Dennis,

    I don’t see any coorelation between spotless days and wolf number. Some weakcycles are preceded by a large number of spotless days – others are preceded by a small number.

    Also, if they missed sunspots in the past one would expect to see a larger number of spotless days as well. This appears to be the case which supports Dr. Svalgaard’s claim.

  280. Leif Svalgaard
    Posted Feb 11, 2008 at 8:10 AM | Permalink

    277 (Andrew): The percentage makeup of TSI as shown is about right. The curves shown are for illustration only and I can live with them. You could put my reconstruction instead and it would not change the relative makeup.

    279 (Raven): Exactly. I think they undercounted sunspots in the past, and the smallest spots wouldbe undercounted the most. This would tend to give you more spotless days.

  281. Andrew
    Posted Feb 11, 2008 at 8:20 AM | Permalink

    Missing small spots kind of reminds me of the “Tiny Tim Storms” probably missed in the past that are biasing current cyclone numbers upward. I wonder, is there a good record of individual spots? If there is, would it be possible to just remove small spots from recent years, rather than try to guess how many were missed in the past? Would it be better to know the sunspot number including or excluding small sunspots?

    Thanks for answering my question, Leif.

  282. Leif Svalgaard
    Posted Feb 11, 2008 at 8:39 AM | Permalink

    281 (Andrew): the modern record is quite good, and one can play games with the small spots. My claim that spots were undercounted in the past has a more solid basis, though. Higher solar activity results in more ultraviolet and xray emission which in turn results in more ionization of the Earth’s upper atmosphere, which in turn results in stronger electric currents causing a larger magnetic effect that can be easily measured on the ground. Such measurements go back more than 200 years and it is from those that we infer solar activity.

  283. Dennis Wingo
    Posted Feb 11, 2008 at 9:03 AM | Permalink

    (#282)

    Leif

    From the graph it would seem that if spots were missed, they were small ones as indicated by the number of spotless days. Would the small, fairly transient spots make that big of an impact on the magnetic field? I can see if spots were undercounted during active times which would tend to suppress the Wolf number in the more active parts of a cycle. I would like to ask you to think about this one a bit. I tend to rely on your call on this but this does seem to be an anomaly in your proposition.

  284. Leif Svalgaard
    Posted Feb 11, 2008 at 9:18 AM | Permalink

    283 (Dennis): I spend a good fraction of my life thinking about these things :-)
    It is the count at solar maximum that mostly determines the magnetic effects. At solar minimum, we don’t get much. But because even a single tiny spot results in a sunspot number of 11, they have a big impact at solar minimum, and certainly on the number of spotless days. My contention about solar activity is mainly about the maxima as the effect on the minima is so small [except when you count spotless days]. But I’m not quite sure what your point is.

  285. Andrew
    Posted Feb 11, 2008 at 10:37 AM | Permalink

    Thanks again, Leif. 11 for a single, small spot? Who came up with that? I’m beginning to wonder exactly how this is all calculated. It probably takes to long to explain, do you have a link to an explanation of how sunspot numbers are determined?

  286. Jim Arndt
    Posted Feb 11, 2008 at 10:58 AM | Permalink

    Leif,

    what is your estimation of the beginning of the next cycle (SC24). We are still seeing spots from SC23 i.e. 982. We have a running theory that it will be May 2009 and I predict it will be weak and far below NASA’s low estimation of 90. Have you seen the latest numbers, I know weather is not climate but you can’t ignore the numbers at this solar minima. For January 2008 RSS -.629, UAH -.588 PDO -1.56 and snow coverage is the highest since 1966.

  287. Pete
    Posted Feb 11, 2008 at 11:05 AM | Permalink

    #286(Jim) To add that the ‘Warmers’ GISS Series just posted +0.12C. 13 Year Low.

  288. Jim Arndt
    Posted Feb 11, 2008 at 11:15 AM | Permalink

    Hi,

    Yea Pete for some reason GISS is way out of sync with the other methods. Don’t know why but considering the director well you catch my meaning.

  289. Raven
    Posted Feb 11, 2008 at 11:47 AM | Permalink

    The explaination given by the warmers is that GISS includes the poles but the other series do not. However, it is not clear to me how they get the data for the poles and why they think it has any connection with the actual temperatures at the poles.

  290. Jim Arndt
    Posted Feb 11, 2008 at 12:11 PM | Permalink

    Hi,

    My original post was SC24 and solar relationship to current temperatures and not debating temperatures. Please try to stay OT.

  291. Posted Feb 11, 2008 at 12:49 PM | Permalink

    285 (Andrew): The sunspot number is SSN = 10 G + S, where G is the number of ‘groups’ and S is the number of spots. So, if you have only one single spot on the Sun, G = 1, S = 1, and SSN = 11. This is not so silly as it sounds because the resulting number correlates very well with all knids of solar parameters.

    286 (Jim): My guess for minimum is summer of 2008. My estimate for SC24 Rmax is 75, thus the smallest cycle in a 100 years. I’m not hot on the idea of using Earth’s temperature as a proxy and predictor for the sunspot cycle.

  292. Andrew
    Posted Feb 11, 2008 at 2:07 PM | Permalink

    Okay, thanks again. I might just be learning more than I want to know about the sun! ;)

  293. Jim Arndt
    Posted Feb 11, 2008 at 2:51 PM | Permalink

    Hi,

    Leif,

    Sorry for the mis-understanding I was looking at it the other way around. I think using the spotless days proxy works pretty good but we will see. I was stating that temperature to TSI for the second part of my post seeing we are in a low activity stage and temperature seems to be following.
    I have another question, I thought it was a little weird but here it goes. I have heard a theory floating that the solar wind may be interfering with the east west air currents. Is this possible considering the possible effect on satellites.

  294. Posted Feb 11, 2008 at 4:03 PM | Permalink

    293 (Jim): “the solar wind may be interfering with the east west air currents. Is this possible considering the possible effect on satellites”.
    I would not consider this to be possible as stated. Now, many people think there is a connection between solar activity and the weather, and they may have a different opinion. And since you mention satellites, you may be thinking of the ionosphere rather than surface winds, in which case such a connection is much more likely, except that there are no “east west air currents” there that I am aware of.

  295. Jim Arndt
    Posted Feb 11, 2008 at 4:25 PM | Permalink

    Hi,

    Leif,

    Thank you,
    I was referring to the east to west weather patterns and that the currents (jet stream) get sent off course causing them to send the tropical warm air toward the northern latitudes and polar regions. What I was referring to was that the solar wind partials that penetrate the magnetic field cause a drag on the atmosphere. That is why it sounded weird to me. So please forgive me since my terminology is off at times. It was a theory being floated and I’m not sure of the merits.

  296. Posted Feb 11, 2008 at 4:49 PM | Permalink

    295 (Jim): The solar wind does not drag on the atmosphere, so you can stop worrying about that.

  297. Jim Arndt
    Posted Feb 11, 2008 at 5:22 PM | Permalink

    Hi,

    Leif

    Thank you. Just need to get that off my chest. It sounded a bit odd so just need some input.

  298. Jim Arndt
    Posted Feb 11, 2008 at 5:32 PM | Permalink

    Hi,

    Leif,

    On the start of SC24 you think it will be this summer. I think JimP is on to using spotless days to determine the next cycle. JimP says the following:

    The first assumption is that the sun has shifted its pattern of the start of the sunspot cycles from a delta of 33 +/-5 months to a delta of 66 +/- 4 months. Cycle 23 had a delta of 25 and did not follow either pattern. If cycle 24 is as short as cycle 23 then it could have a start date of delta 58 or November 2008. If cycle 24 follows the traditional 66 +/-4 the start date would be between March of 2009 and November 2009.

    I separated the number of spotless days by cycle and compared each one of them to the upcoming cycle 24. Cycles 10, 11, and 14 have much slower starts and an average delta of 68. Cycle 13 and 15 appear to be the most similar and have an average delta of 62. This indicates cycle 24 have a delta of 62 which is at the low end of Solaemon’s 66 +/- 4 estimated delta.

    Cycle 24 has a lower number of total spotless days at this point, delta 49, compared to 13 and 15. This might indicate that cycle 24 will have a longer delta than 13 and 15. Cycles 10, 11 and 14 had very low totals at this point. So I think that cycle 24 will start in May of 2009 and will have a delta of 64.

    This is what I think but taking into consideration the the length of SC23. If SC23 continues to drag out then it points to a very low SC24. You said Rmax 75. How long will SC23 have to drag out to change your estimate?

  299. Leif Svalgaard
    Posted Feb 11, 2008 at 11:07 PM | Permalink

    298 (Jim): go back to #110 and check my analysis there. The size of SC24 is only very weakly correlated with the length of SC23, so the length as such is a poor predictor. No matter how long SC23 is, SC24 will still be about the same (namely Rmax = 75), unless, of course, we are heading into a deep Grand Minimum, which I don’t think [but also don't know].

  300. Erl Happ
    Posted Feb 12, 2008 at 2:27 AM | Permalink

    Leif,
    Any comment on #269?
    Does ultraviolet vary with geomagnetic activity? Is there an available, user friendly statistical series that summarises strength of ultraviolet and FUV on a monthly basis?

    I am seeing some evidence that the troposphere is heated independently of the surface. Cloud cover may vary directly with the extent of radiant heating of the troposphere.

  301. JimP
    Posted Feb 12, 2008 at 6:49 AM | Permalink

    Here’s a way of looking at the total number of spotless days per cycle compared to the average number of sunspots per year per cycle. Note cycle 23 is not over yet and so the total number of sunspots will continue to grow until cycle 24 begins. The spotless days will continue to be counted until they stop. This could be tomorrow, 2 years from now or 4 years from now depending on how the sun is feeling. So expect the “Total Spotless Days” curve to start dropping at the end as time goes on.


  302. JimP
    Posted Feb 12, 2008 at 6:57 AM | Permalink
  303. Leif Svalgaard
    Posted Feb 12, 2008 at 8:26 AM | Permalink

    300 (Erl): #272 was a good answer to #269. The sunspot number is a good measure of the UV. Geomagnetic activity and UV are not directly correlated.

    301 (JimP): In 302 you plot sunspot number per year. As SC23 drones on that number will fall and the number of spotless days will rise.

  304. Ulric Lyons.
    Posted Feb 12, 2008 at 10:48 AM | Permalink

    Once it is understood which variable output of the Sun, (not being EMR), drives and controls climatic variations on Earth, one will still have to refer to what is causing the variations in order to be able to predict weekly/monthly/yearly temperature change. So when it shown that the relative heliocentric positions of the Planets correlate perfectly to any change, or any extremes in temperature, we can predict climatic change over short, or very long periods, and we are able to also predict cyclones, and damaging solar flares.
    I would like to give a forecast for 2008, not based on a description of a full set of `rules` of relative planet positions, but by simply looking back 179.05 sidereal years, to when most of the planets were in the same relative positions. For the UK I am using this data; MONTHLY MEAN CENTRAL ENGLAND TEMPERATURE (DEGREES C) http://www.metoffice.gov.uk/research/hadleycentre/CR_data/Daily/HadCET_act.txt
    Looking back 179y and 1 month, we can get a relative, but not totally absolute idea of variation from the average temp. So for example, for Jan 2008, look to Dec 1828(warm), for Feb 2008 look to Jan 1829 (cold). The outlook for 2008 average looks low; 8.16C.

  305. Leif Svalgaard
    Posted Feb 12, 2008 at 10:52 AM | Permalink

    304 (Ulric): No perfect predictions, please.

  306. Ulric Lyons.
    Posted Feb 12, 2008 at 11:19 AM | Permalink

    305(Leif)Why not? I can guarantee when extremes in temperature will occur to the month, do you have a problem with that?

  307. See - owe to Rich
    Posted Feb 12, 2008 at 11:30 AM | Permalink

    Leif #291, #299

    I did find your #110 very interesting, and it just shows how hard it is to predict from matching graphs. You seem to think we may be only 5 or 6 months away from solar minimum.

    However, there is one other statistic which David Archibald uses, which is interval between first spot of new cycle and minimum. He says this is typically 12 to 20 months, which would put minimum at January 2009 at the earliest. Do you have any opinion (or indeed data) on this other statistic?

    Thanks,
    Rich.

  308. Leif Svalgaard
    Posted Feb 12, 2008 at 12:16 PM | Permalink

    306 (Ulrich): I have a problem with unsubstantiated claims. It is also not specified what an ‘extreme’ is. I too can guarantee that sumemer will be warmer than winter. And your premise that ‘most’ of the planets are in the same relative position every 179.05 years is simply wrong.

  309. Bob B
    Posted Feb 12, 2008 at 12:25 PM | Permalink

    Leif, when are you going to update your research page with the latest plots of TSI, sunspots etc? I really like the information you post :)

  310. Leif Svalgaard
    Posted Feb 12, 2008 at 12:34 PM | Permalink

    309 (Bob B): The problem is that the Sun changes so slowly, but maybe it is time for an update. For things where I plot yearly values, I update once a year. For the sector structure I update every day. Thanks for visiting my page.

  311. Earle Williams
    Posted Feb 12, 2008 at 12:41 PM | Permalink

    Leif,

    Thanks heaps for your continued participation in this thread. I’m hoping you’ve got the patience to answer a question or two from someone who isn’t an astro (my geophysics background focuses on looking down, not up) and isn’t up on the literature. Firstly, are sunspots currently well-enough observed to arrive at area measurements and perhaps intentisty as well? If so, what is the utility of integrating across area and time to arrive at the km^2-days for sunspot extent and duration? If there is utility in that, how about adding in intensity (if it exists)?

    Thanks again for being so accessible and patient.

  312. Earle Williams
    Posted Feb 12, 2008 at 12:42 PM | Permalink

    Fat-fingered above. Intentisty my eye!

  313. Ulric Lyons
    Posted Feb 12, 2008 at 4:22 PM | Permalink

    Jim #314 Hi, I am not saying what the physics is, it could be about tidal effect by the planets to the surface of the Sun as there is a strong polarity between squares and syzygies, synthesis of barycenters as well? I am mostly concerned with correlating positions with events and changes, it works out very well without even knowing the means and mechanisms, but could help in suggesting them, or eliminating false theories.

  314. Leif Svalgaard
    Posted Feb 12, 2008 at 8:06 PM | Permalink

    307 (Rich): The problem is ‘which’ minimum. Different definitions give different times. If we define minimum as the lowest value of the smoothed sunspot number, the time from first spot to minimum at the entry to SC23 was short (~6 months, see the plot on the last page of http://www.leif.org/research/Most%20Recent%20IMF,%20SW,%20and%20Solar%20Data.pdf. Hanging everything on a single spot is not very good ad there is a lot of randomness in the system. Better would be to base it on the, say, third spot.

    311 (Earle): measurements of sunspots (including their magnetic field and area) are now precise, but that does not mean that we can use those numbers for equally precise predictions. It is like predicting the weather from the precise and exact size and shape of individual clouds.

    313 (Ulric): the positions do not repeat precisely enough to maintain phase over thousands of years (I think you quoted 11,400 years) and do not line up on a time scale of months, but, hey, if it keeps you happy I’ll not spoil it. I’m just saying that none of the planetary stuff has ever seemed compelling to me.

  315. Leif Svalgaard
    Posted Feb 12, 2008 at 8:34 PM | Permalink

    316 (me): remove the last dot after the pdf in the url.

  316. Ulric Lyons
    Posted Feb 13, 2008 at 5:55 AM | Permalink

    [snip- SM - please no discussion of planets. The purpose of this site is to discuss mainstream science not to offer a platform for personal theories.]

  317. Ulric Lyons
    Posted Feb 13, 2008 at 8:05 AM | Permalink

    [snip- SM- Ulric, please. ]

  318. JohnB
    Posted Feb 13, 2008 at 8:07 AM | Permalink

    Ulric, Mike, last week I was looking through a book on weather published back in 1982. I came across an interesting reference to work by Ren Zhenqiu of the Peking Academy of Meteorological sciences and Li Zhisen of Peking Astronomical Observatory looking at Chinese weather data for the past 1000 years. The Chinese have more detailed and more accurate historical records than any Western country.

    (I think this is probably the relevant paper by Zhisen and Zhenqiu:

    http://www.springerlink.com/content/u56425j172359461/)

    According to Zhenqiu and Zhisen, the book said, whenever all the planets grouped on the same side of the Sun, China experienced cooler weather, in a 179 year cycle.

    Since 1000 AD, the article says, China has had five intervals of low temperature; in the first half of the twelfth century, in the fourteenth century, at the end of the fifteenth century, in the seventeenth century, and in the nineteenth century. This would seem to match the Wolf, Sporer, Maunder and Dalton minimums. According to Zhenqiu and Zhisen, these appear to correspond with the planetary alignment they identified.

    We are in a similar position now, although I think the closest alignment for Jupiter, Uranus and Neptune was sometime in the late ’90s.

    This sounds to me like a practical observation of the Barycentric motion referenced by Landscheit and Fairbridge, both of whom suggested we are now due for a “cold spell”.

    I also found this useful, a paper which looks at Barycentric motion and also points to a match with the Wolf, Sporer, Maunder and Dalton minimums, as well as forecasting an imminent cold spell for us now.

    http://www.ann-geophys.net/18/399/2000/angeo-18-399-2000.pdf

  319. Ulric Lyons
    Posted Feb 13, 2008 at 9:25 AM | Permalink

    [snip - Ulric, not here.]

  320. bender
    Posted Feb 13, 2008 at 7:33 PM | Permalink

    Speaking of spurious correlation … any signal can be represented as a sum of sinusoids of varying frequencies. That is the problem with non-mechanistic correlative pattern description. It too easily leads to junk science.

  321. Jim Arndt
    Posted Feb 13, 2008 at 7:50 PM | Permalink

    Hi,

    Leif I know you don’t like correlations but Joe D’Aleo at Icecap did a graph of TSI to temperature and PDO. It is very telling. I think if you change the smooth to 7 or 5 years you may get a different result but it is hard not to see the correlation. Here is the link.

    http://icecap.us/images/uploads/US_Temperatures_and_Climate_Factors_since_1895.pdf

  322. Leif Svalgaard
    Posted Feb 13, 2008 at 8:30 PM | Permalink

    323 (Jim): It would have been better science [if any] had he used a modern reconstruction of TSI, instead of the obsolete Hoyt and Schatten one. Smoothing is always a tricky business as it reduces the ‘number of degrees’ of freedom and therefore makes any correlation ‘better’.

  323. Posted Feb 13, 2008 at 8:43 PM | Permalink

    Ref 323 Jim Arndt

    That uses the Hoyt reconstruction and US mean temps so the only the choir will cheer. Hoyt has been superseded and the US is only a small fraction of the globe.

    Solar is a factor because it is the input. The impact of solar is lost in the chaos that is our atmosphere. So there is no defendable obvious correlation because of shifts in climate (oscillations) reduce the solar impact to noise.

    There a couple of things I would like to see proven. First is that Northern Hemisphere temperature changes have a greater impact on global temperatures than Southern Hemisphere temperature changes. This to me is obvious since the Northern Hemisphere has a greater percentage of land mass, it’s temperature change would be more volatile. It’s higher temperatures would have greater impact on global averages because it is more volatile.

    If that is done “regional” anomalies, in the Northern hemisphere could be used with better ethos to model climate trends. The Northern Hemisphere temperatures respond more quickly to change because of percentage land mass. It makes a more sensitive indicator of change than global averages. It is not a regional thing it is a physic’s thing.

    The second thing I would like to see is more focus on the tropical oceans. They absorb the most solar radiance. The tropics provide the energy that drives all the oscillations that control climate. Very minor variations in the tropical oceans can produce huge changes in global climate. Since there have been huge fluctuations in the pre-industrial past, there are probably natural climate drivers that are not that well understood. Subtle variations in the tropical oceans are likely the cause.

    now that I have bored everyone with my vague comment, back to lurking.

  324. bender
    Posted Feb 13, 2008 at 9:56 PM | Permalink

    #327

    The impact of solar is lost in the chaos that is our atmosphere.

    RC disputes this.

  325. EW
    Posted Feb 14, 2008 at 7:16 AM | Permalink

    Nice NOAA page for playing with graphs and correlations.

  326. Jim Arndt
    Posted Feb 14, 2008 at 11:02 AM | Permalink

    Hi,

    I don’t want to harp on this too long but here is temperature reconstruction using Wang and Lean TSI. It shows a correlation and better explains the results. It is not a perfect fit but it also leaves out the effects such as UV to ozone changes and of course PDO and AMO.

    http://www.fel.duke.edu/~scafetta/pdf/2007JD008437.pdf

  327. Leif Svalgaard
    Posted Feb 14, 2008 at 11:07 AM | Permalink

    SORCE 2008: We had a very successful meeting.
    Among the talks was this one:
    Stratospheric /Tropospheric Response to Solar Forcing
    D. Rind, J. Lean and C. McLinden
    Their first slide showed this:
    Mechanisms for solar influence on the troposphere
    • Influence on the stratosphere, with dynamical
    effects ‘extending’ downward (UV)
    • Influence on subtropical surface temperatures
    (land and sea), with subsequent atmospheric
    dynamic and moisture effects (TSI)
    • Influence on sea surface temperatures and
    atmospheric winds with subsequent ocean
    dynamical effects (TSI) (perhaps amplifying
    natural modes of air-sea interaction)

    after the talk, I asked why cosmic rays/clouds were not mentioned and the interesting answer was that “the mechansim was not specified well enough to be modeled, so was not considered”.

  328. Leif Svalgaard
    Posted Feb 14, 2008 at 11:17 AM | Permalink

    326 (Jim): About leaving out UV: This seems to be a favorite and false strawman. UV is closely tracked by the sunspot number as are TSI reconstructions, so the latter have UV variation built-in from the start. The actual magnitude of the variation (0.1%, 10%, 100%) doesn’t matter for the correlation, only the relative variation.

  329. Jim Arndt
    Posted Feb 14, 2008 at 11:25 AM | Permalink

    Hi,

    Here is some of the work by Henric Svensmark on CR and clouds.

    http://spacecenter.dk/research/sun-climate/Scientific%20work%20and%20publications/resolveuid/1f09645cb14320a460fc697de6be993b

    http://spacecenter.dk/research/sun-climate/Scientific%20work%20and%20publications/resolveuid/7be65644ef9a2888fd33813d5fd05a21

    http://spacecenter.dk/research/sun-climate/Scientific%20work%20and%20publications/resolveuid/e3a965fb4c422d284418e34c4a2f2cde

  330. Jim Arndt
    Posted Feb 14, 2008 at 11:28 AM | Permalink

    Hi,

    Leif isn’t it possible that UV interacts with the with the atmosphere to create more ozone which then has a warming effect on the upper atmosphere. This can add to the total warming seen on the surface.

  331. kim
    Posted Feb 14, 2008 at 11:30 AM | Permalink

    Does anyone know what Jasper Kirkby, of CERN, presented to the Royal Scientific Academy of Sweden, yesterday?
    =======================================

  332. Andrew
    Posted Feb 14, 2008 at 11:32 AM | Permalink

    Leif, hi, I think you (and Jim) are misunderstanding the UV thing. First off, Jim, Scaffeta and West doesn’t include or exclude such effects, because they weren’t analyzing the “how” of the influence, but the “amount” of possible influence seemingly implied by the data. Leif, the influence of UV is hypothetically different from TSI because in influences Ozone chemistry, which is thought to produce atmospheric heat. This is separate from its direct forcing effect.

  333. Leif Svalgaard
    Posted Feb 14, 2008 at 11:38 AM | Permalink

    330 (Jim): many things are ‘possible’. The effect of UV has been discussed in many, many places. [easy to find]. A general note of flow of heat: Warm air goes up, cold air goes down. A warmer stratosphere or [even higher] thermosphere does not heat the surface layers. My point was that UV is already incorporated in the TSI.

  334. Jim Arndt
    Posted Feb 14, 2008 at 11:39 AM | Permalink

    Hi,

    Andrew thank you, That is what I was saying that the Scaffeta/West paper is on the relationship of TSI and doesn’t include UV, PDO and AMO.

  335. Bob B
    Posted Feb 14, 2008 at 11:50 AM | Permalink

    Leif, I see that Hathaway has again reiterated he expects cycle 24 to be larger then 23.

    http://www.earthfiles.com/news.php?ID=1371&category=Science

    My question is there a drop-dead date (March08,Sept08?–etc)where his predictions can be more or less proven untrue?

  336. Jim Arndt
    Posted Feb 14, 2008 at 12:14 PM | Permalink

    Hi,

    Leif you are correct that warm air goes up a cold air goes down. But the atmosphere has thermals that mix the warm and cold air. Also some IR is radiated back to the surface and I believe that ozone acts as a thermal layer preventing some IR from being emitted back into space.

  337. See - owe to Rich
    Posted Feb 14, 2008 at 1:02 PM | Permalink

    #314 Leif

    If I wanted to audit Archibald’s claims on interval between first spot and minimum, I would need data on first spots (data on minima are easier to find, e.g. Lassen). Unfortunately I do not see how to use NOAA’s DSD.txt data to work out which spots are on which cycle. Is there a way to do this, or is there some other resource you know of (perhaps a data file at your own site rather than a graph)?

    Also, you said “depends on which minimum” [definition]. How many definitions are you aware of, and with your favourite, where did you place the minimum at start of SC23?

    TIA,
    Rich.

  338. Leif Svalgaard
    Posted Feb 14, 2008 at 1:21 PM | Permalink

    332 (Andrew): that was not my point. It was that UV goes up and down with TSI so any correlation with one automatically includes the other.

  339. jack mosevich
    Posted Feb 14, 2008 at 1:26 PM | Permalink

    I hope for an active cycle. Then I can get out my telescope and hydrogen alpha filter!
    See this link for 1 frame from a video I made in 2001 of a loop prominence:

    http://www.fortunecity.com/victorian/canterbury/222/mosevich.htm

    Note; my e-mail address as posted there is no lond=ger valid.

  340. Edward
    Posted Feb 14, 2008 at 2:22 PM | Permalink

    “I see that Hathaway has again reiterated he expects cycle 24 to be larger then 23.”

    This will not be the first time that a method designed to match the past failed miserably the first time it was used to predict the future. Think of stock markets.

  341. Leif Svalgaard
    Posted Feb 14, 2008 at 3:03 PM | Permalink

    337 (Rich): Hathawat’s site has a complete record of every active region since 1874: http://solarscience.msfc.nasa.gov/greenwch.shtml
    Use the latitude to distinguish between old and new cycle spots at minimum.

    My favorite definition of solar minimum is when the heliomagnetic current sheet is the flattest. I’m explaining this in the forthcoming ‘notes’ [coming real soon now]. Minimum is at least half a year away, maybe more. This ‘prediction’ will, of course, change as time goes on and we ain’t got the minimum yet :-)

  342. cba
    Posted Feb 14, 2008 at 3:58 PM | Permalink

    326

    I just read most of it. From what I thought I understood was that Lean & Wang’s TSI values were compared and that the greater one came fairly close to explaining the T increase. Leif’s calculation graph places his estimates just above Wang’s which would imply that perhaps his values might come extremely close to explaining the T rise – so far as that referenced thermodynamic model approach goes.

    Did I get that right ????

  343. Leif Svalgaard
    Posted Feb 14, 2008 at 4:19 PM | Permalink

    336 (Jim): The density of the atmosphere falls by a factor of 1000 for each 50 km height, so the amount of heat in the stratosphere in minute. There is very little termal mixing across the tropopause. The radiation budget has been rehashed ad nauseam in this and other blogs, so no need to go there agin.

    342 (cba): my TSI does not have a secular rise, so cannot explain a secular increase in Temperature, but the various papers that try to invoke a solar contribution look very simplistic to me, but then, again, I don’t really know enough about them and the atmospheric response to criticize them too severely. Lots of other people do that for me :-)

  344. cba
    Posted Feb 14, 2008 at 4:48 PM | Permalink

    Leif,

    This paper was a simple model of T vs TSI. It’d be interesting to see what your data would do under the same circumstance. Due to the nature of what they were doing, I’m not sure your’s would not provide an interesting result also, despite it being rather flat.

  345. Jim Arndt
    Posted Feb 14, 2008 at 4:59 PM | Permalink

    Hi,

    Leif thank you, I agree that it is not a complete correlation, because if it was it would be a “no brainer”. There are many other factors at work such as ENSO, AMO, PDO, Volcanic Eruptions, Other Aerosols and then you get to GHGs. But the research and money goes to the squeaky wheel which is now focused on GHGs and not other factors that can influence climate. I think there should be more focus on solar influences when you consider climate history. If we do have a grand minimum, which is entirely possible, and we enter a cooling phase then we are not prepared. I will not go into particulars about policy since it is way OT. But I think it is wrong to dismiss solar influence on the climate which is what is happening in the main stream at the moment.

  346. Leif Svalgaard
    Posted Feb 14, 2008 at 5:48 PM | Permalink

    344 (cba): Since my TSI is flatter than Wang et al., if you want to explain the same T-effect, you will have to crank up the climate sensitivity to solar forcing. That is fine with me, but some people think that is not good. My main purpose of joining this blog was to get an answer to that issue. So far I haven’t gotten any…

  347. kim
    Posted Feb 14, 2008 at 5:52 PM | Permalink

    A Flamsteed drawing suggested slowing of solar rotation during the Maunder Minimum. Do you believe that and is there any such suggestion today?

    H/t Spain.
    ======

  348. cba
    Posted Feb 14, 2008 at 6:15 PM | Permalink

    Leif,

    The paper in 326 is seeing that Wang – with flatter TSI is providing both a higher sensitivity but also providing some time response in the ballpark of what would be expected from a real system whereas the higher sensitivity one offers virtually no reasonable time reflex.

    I don’t have any feel for what would happen if you plug in the numbers to their approach due to the time reflex factor and the requirement for higher sensitivity. Perhaps the flat curve would dominate or perhaps the time factor would have a more interesting effect.

    sorry for the disjointed statements but after 2 x 12 hr days I’ve got to head back to the salt mine again tonight – either for a couple of hours of PR tonight or to close down the announced public viewing due to too many clouds and I’ve got to run now!

  349. Andrew
    Posted Feb 14, 2008 at 6:18 PM | Permalink

    Leif, that was kind of my point to, that Scafetta and West technically do include UV, so Jims comments about amplifiers don’t hold. By the way, although less of a rise in TSI might imply less impact on actual warming, if the Cosmic Ray theory were correct, there would still be greater solar forcing over the solar cycle than is currently accounted for, which actually might imply a lower climate sensitivity. On the other hand, if it is not, then, assuming that sensitivity doesn’t vary for different factors, climate sensitivity is underestimated (or rather, the empirical derivation are). Actually, that first assumption is needed for the latter statement, to.

  350. Leif Svalgaard
    Posted Feb 14, 2008 at 6:51 PM | Permalink

    347 (kim):
    I’m coauthor of a recent paper devoted to that question:

    On the solar rotation and activity
    Brajsa, R.; Wöhl, H.; Ruzdjak, D.; Vrsnak, B.; Verbanac, G.; Svalgaard, L.; Hochedez, J.-F.

    Astronomische Nachrichten, Vol.328, Issue 10, p.1013, 2007
    DOI:10.1002/asna.200710867

    Abstract:
    The interaction between differential rotation and magnetic fields in the solar convection zone was recently modelled by Brun (2004). One consequence of that model is that the Maxwell stresses can oppose the Reynolds stresses, and thus contribute to the transport of the angular momentum towards the solar poles, leading to a reduced differential rotation. So, when magnetic fields are weaker, a more pronounced differential rotation can be expected, yielding a higher rotation velocity at low latitudes taken on the average. This hypothesis is consistent with the behaviour of the solar rotation during the Maunder minimum. In this work we search for similar signatures of the relationship between the solar activity and rotation determined tracing sunspot groups and coronal bright points. We use the extended Greenwich data set (1878-1981) and a series of full-disc solar images taken at 28.4 nm with the EIT instrument on the SOHO spacecraft (1998-2000). We investigate the dependence of the solar rotation on the solar activity (described by the relative sunspot number) and the interplanetary magnetic field (calculated from the interdiurnal variability index). Possible rotational signatures of two weak solar activity cycles at the beginning of the 20th century (Gleissberg minimum) are discussed.

    Unfortunately, the main author did not understand that an Abstract should not describe what we did, but what we found, so here is that missing vital piece:
    “The linear fit shows a decreasing of the angular rotation velocity by about 0.1 degrees per day for an increase of the sunspot relative number by 100 units.”

    So it seems that higher activity is associated with slower solar rotation. Kind off the opposite of what Flamsteed found, so his result must stand alone. But, the issue is complex and not well understood.

  351. Posted Feb 14, 2008 at 6:55 PM | Permalink

    Ref 324 Bender said:
    #327

    The impact of solar is lost in the chaos that is our atmosphere.

    RC disputes this.

    Of course they do. They also assume that the majority of the apparent warming is anthropogenic. With Tsonis’ chaotic analysis it seems reasonable that there is significant natural influence on current climate. That doesn’t rule out anthropogenic, it just indicates that less biased research is required to find the clues to subtle but significant climate influences.

  352. Leif Svalgaard
    Posted Feb 14, 2008 at 6:59 PM | Permalink

    350 (me): Kim, are you sure you quote your info about Flamsteed correctly. If memory serves, he found a more pronounced differential rotation, but I could remember wrong. You go and check it out.

  353. kim
    Posted Feb 14, 2008 at 7:27 PM | Permalink

    It’s a paper from 2002 by Vaquero, Sanchez-bajo, and Gallego, and it is certainly possible that I misunderstand it.
    ================================================

  354. Edward
    Posted Feb 15, 2008 at 6:37 AM | Permalink

    I found this on another blog. Can anyone explain this?

    “Could not agree more on Dr. Landscheidt. Even though he may have been a largely self-taught Heliophysicist he had phenomenal instincts unlike some academicians, e.g, Leif Svalgaard, who are his steadfast detractors.
    Ran across a paper by Tsagas(2006) that might point to the second derivative zeroes in solar angular momentum due to barycentric orbit causing a relativistic perturbration by means of the Lorentz force. The transit through the gravitaional well would induce a repulsive force at right angles to the field, i.e., the polar axis, suffcient to stop the gravitational collapse of a black hole.
    The field’s hystereisis would rigidly oppose the imposed deformation due to the transit. DeJager, Versteegh(2004) criticised Landscheidt’s proposal believing it to imply a tidal force as the cause (a change in direction of the angular momentum), but the Dr. did not specifiy the source. I believe the Dr. will be found correct and has more surprises in store.”

  355. Leif Svalgaard
    Posted Feb 15, 2008 at 7:25 AM | Permalink

    354 (Edward):
    Another application of relativity is this abstract from SORCE 2008. Not many [if any] could take it seriously but its very presense shows that we are not blind dogmatists:
    Applying Relativity to Earth Climate Data The Damhsa Theory Signs of the Inflationary Universe
    Sheila A. Lynch [slynch@navc.org], Northeast Advanced Vehicle Consortium (NAVC),
    Boston, MA.
    Multi-million year climate data through proxies have recently been published that
    point to a clear trend of temperature differentials through time. In the past theories have
    been used to explain the broad ice ages by orbital forcing using Newtonian physics. The
    Damhsa (Gaelic for Dance) Theory is formulated by analyzing climate data and applying
    the General Relativity Theory and orbital forcing to the time series and proposes a
    solution to the variable data. This solution is gravitational waves. New theories on the
    inflation of the universe predict gravitational waves also. This change is extremely slow
    and not perceptible to human scale time but can explain the complex interactions of
    large-scale climate change and time. The climate fluctuations in time can be explained
    by gravitational waves of the expanding universe. The Earth’s position in space changes
    as the effects of gravitational waves as predicted by Einstein. Recent climate data shows
    wave patterns of a non-linear nature, which would correspond with a large mass in space,
    such as Earth, exhibiting the effects of gravitational waves by slight changes in the
    position of Earth to the Sun, which would slowly affect climate over large timescales.
    Oscillating gravitational waves are the signature of the universe expanding.

  356. Ulric Lyons
    Posted Feb 15, 2008 at 11:33 AM | Permalink

    #316(SM) I am hardly the first or only person to correlate 179yr planetary harmonic to solar/climatic variation, and the correlation can not be spurious (ref #320), as the exact points of conjunction, opposition and squares, coincide with key features on the records.

    SM - I try to keep a light editorial hand, but I’m not interested in providing a platform for discussion of this topic. It’s a big world and there are lots of usenet forums. Please do it elsewhere.

  357. cba
    Posted Feb 16, 2008 at 8:59 AM | Permalink

    is this no longer posting?

  358. Leif Svalgaard
    Posted Feb 16, 2008 at 9:16 AM | Permalink

    357 (cba): yes it is, but you must post something…. to keep it on the ‘new postings list’

  359. cba
    Posted Feb 16, 2008 at 11:31 AM | Permalink

    I must not be tuned in to the new postings list as I’ve just been checking for new postings, eagerly awaiting some more installments of basic info and comments and references to other papers and associated tidbits – like that S&W paper.

    Referencing 346, the S&W implication of a much higher sensisitivity to TSI (such as Wang) makes me a bit queezy as well. However, it does offer some explaination and provides a more reasonable relation for their time lag factor.

    Judging by Steve M’s expose’ of the Lampassas temperature measuring station for GISS and the apparent ‘correction’ to fix (exacerbate) the situation, the potential of overstatement of recent heating seems apparent.

    I do think though it’s becoming rather apparent that W/m^2 variations are far from being equal when it comes to actual temperature effects. That is a 1 W/m^2 change in solar insolation is not going to have the same effect as a 1 W/m^2 change in GHG absorption. Considering there are far more serious variations between summer NH and summer SH insolation as well as far more SH ocean absorption compared to NH, it’s hard to accomodate the notion of extreme sensitivity just to power variation (W/m^2).

    We’ve had reasonably constant conditions for quite some time now despite a 30% solar TSI increase, probably some albedo changes (like lush sahara vs pre sahara desert), orders of magnitude GHG concentrations, variations in obliquity, orbital eccentrity, etc., locations of land masses and changes in ocean current pathways and probably some significant variations in background cosmic ray levels and variations in amounts of vulcanism. All this suggests to me serious regulating feebacks and/or incredible insensitivy to just about everything (or both). Considering the long term ice age conditions that occur periodically, this suggests to me that it is a cloud / albedo type of mechanism which periodically becomes short circuited by an alternative surface based high albedo option – snow.

  360. Philip Mulholland
    Posted Feb 16, 2008 at 4:06 PM | Permalink

    cba #359

    Saharan albedo variations? It snowed in the Hoggar on 7th January 2008
    Froid et Neige sur (Tamanrasset)

  361. cba
    Posted Feb 16, 2008 at 5:04 PM | Permalink

    Philip,

    Rough numbers are sand is rather high albedo for surface stuff, up to maybe 40%. Snow might get up to 50 or 60 % maybe sorta. That’s not the big difference. If the sahara were a tropical rain forest (jungle) its albedo could be as low as maybe 12% or so, absorbing lots more than either snow or sand, not to mention lots more humidity being created – which does actually have some effect. Essentially, if the whole sahara were a jungle, there would be slightly lower total albedo for the earth unless compensasted for by additional cloud cover which is in the 50%-60% region, maybe higher.

    Evidently, at one time, the sahara was either heavily forested or was a jungle. Whatever the case, this area probably doubled or tripled its albedo when it became a desert. This is extra important considering that high angle incidence of sunlight on water produces a very low albedo – perhaps under 5% – and that’s 75% of the surface area, leaving only clouds and fairly reflective land areas to bring us up to the 30% region. While the poles are generally fairly reflective due to the snow, that’s not where all the strong incoming radiation is going. As for wintertime additions, one appears to have about 10% of the surface covered with snow at one time or another during the year, suggesting that this small amount might contribute something as well, for at least part of the time.

    Seems like I heard in the last week that some rare snowfall occurred around the middle east this year as well.

  362. Philip Mulholland
    Posted Feb 16, 2008 at 5:51 PM | Permalink

    cba

    I guess that this photo from Apollo 8 on 22nd Dec 1968 of South America (NASA: The Whole of the Western Hemisphere) shows what happens to the continental cloud cover when you grow a tropical rain forest. By the way the deliberate mistake is that this image is actually published laterally inverted but the text is correct. “A small portion of the bulge of west Africa shows along the sunset terminator”. Flip the image and the comment is true.

  363. Philip Mulholland
    Posted Feb 16, 2008 at 5:55 PM | Permalink

    cba

    I guess that this photo from Apollo 8 on 22nd Dec 1968 of South America (NASA: The Whole of the Western Hemisphere) shows what happens to the continental cloud cover when you grow a tropical rain forest. By the way, the deliberate mistake is that the image is actually published laterally inverted but this text comment is correct. “A small portion of the bulge of west Africa shows along the sunset terminator”. Flip the image and the comment is true.

  364. cba
    Posted Feb 16, 2008 at 6:23 PM | Permalink

    does that look like only 40-50% cloud cover on the daylight side?

  365. Dennis Wingo
    Posted Feb 16, 2008 at 10:59 PM | Permalink

    Leif

    I was looking at some of your recent presentations and I think that you got the X Axis off by a few years on this one?

    http://www.leif.org/research/Solar%20Polar%20Fields%20Normalized%20to%20Next%20Cycle.pdf

    Going through and doing some reading. This is important as we are doing some work with NASA right now on the reboost propellant requirements for the space station. NASA is using Hathaway’s numbers, which as you know are much higher than yours or most other researchers now.

  366. Leif Svalgaard
    Posted Feb 16, 2008 at 11:18 PM | Permalink

    365 (Dennis): A ‘polar field cycle’ goes from maximum [or rather polar field reversal] to maximum, so I plot the polar field for each 10 days interval or that time interval by the size [Rmax] of the NEXT cycle. So, the ‘cycles’ are not sunspot cycles. For example, cycle 21 peaked ~1980, so the polar fields at that time were reversing [i.e. zero], then they increased throught the minimum at ~1986 before reaching zero again at the peak of cycle 22 in ~1990. All of the polar fields values between 1980 and 1990 were then divided by the size of cycle 22. For cycle 24 we don’t know Rmax, so I divide by the ‘trial’ numbers 50, 75, and 165. The divisor that best matches the other ‘polar field cycles’ [normalized to the value of the NEXT sunspot cycle, light blue line is average cycle] seems to be 75, so that is our prediction. Hathaway’s prediction is premised on the high geomagnetic activity in 2003, which happened ~6 six before the minimum, although the geomagnetic precursor method specifies that the geomagnetic peak just before minimum should be used. Some people ‘stand by their results’ no matter what. This is not always a virtue.

  367. Dennis Wingo
    Posted Feb 17, 2008 at 1:30 AM | Permalink

    (#366) Thanks for the clarification. I like your recent presentations but I have one request. Could you please store them as landscape rather than portrait? I have to download them then save them then open up adobe reader and then turn them clockwise to read them. A pain.

  368. stargazer
    Posted Feb 17, 2008 at 3:15 AM | Permalink

    a guy on solar cycle 24 posted this

    http://news.bbc.co.uk/1/hi/sci/tech/7137905.stm

    any clue here ?

  369. Leif Svalgaard
    Posted Feb 17, 2008 at 7:01 AM | Permalink

    367 (Dennis): They are landscape because that is most convenient when showing them on a big screen duting presentations.

  370. Leif Svalgaard
    Posted Feb 17, 2008 at 7:05 AM | Permalink

    367 (Dennis): They are landscape because that is most convenient when showing them on a big screen during presentations.

    368 (stargazer): possibly, but we need to learn more about them [which we shall]

  371. kim
    Posted Feb 17, 2008 at 7:07 AM | Permalink

    They seem auroreal.
    =============

  372. Leif Svalgaard
    Posted Feb 17, 2008 at 7:09 AM | Permalink

    366 (me): “plot the polar field for each 10 days interval or that time interval by the size [Rmax] of the NEXT”
    should have been:
    “plot the polar field every 10 days during that time interval divided by the size [Rmax] of the NEXT”

  373. kim
    Posted Feb 17, 2008 at 7:24 AM | Permalink

    I wonder if the cloud effect from cosmic rays may be mostly invisible. By that I mean that massed clouds won’t be the effect, rather that there be brief, diffused, and dispersed phenomena which will reflect energy, but not be seen, en masse.
    ============================================

  374. cba
    Posted Feb 17, 2008 at 8:46 AM | Permalink

    More likely that we see small fractional increases in the average % of daylight cloud cover. The overall effect of more clouds is cooling which means changing the earth’s albedo. If the effect wasn’t visible, it’s not doing much for this and if it’s invisible – blocking IR more – then the effect should be warming, not cooling.

    It would seem the problem with CR cloud formation is that it’s not the only way clouds can form. It’s also likely that numerous conditions have to be met, such that not all areas, perhaps even relatively few areas significantly benefit from this effect.

    You can figure that the atmosphere is full of events going on all the time, including ionization of molecules. As such, there could be influences in an indirect fashion such as ozone formation blocking some light wavelengths. This sort of thing obviously could cause far more effect than the actual power absorption would, but even this is far less than the effect of very small additions of cloud cover.

    Just how close Svensmark and the rest are to unraveling this tangled ball of twine and identifying the actual effects remains to be seen. Hopefully, they’ll get a handle on it soon.

  375. pochas
    Posted Feb 17, 2008 at 10:18 AM | Permalink

    cba #374:

    I also wonder whether the effect of CR’s may be warming and not cooling. I took the Boulder CR count record and lagged it approximating a time constant of 20 years, removed the 1979-1998 mean, and plotted it against the UAH LT anomalies. The slopes are similar, and both positive! (.0206 for the CRs vs 0.0139 for the temps). Just a coincidence I suppose, since the paleoclimate records support an inverse relationship.

  376. cba
    Posted Feb 17, 2008 at 11:12 AM | Permalink

    clouds have both effects, warming and cooling. It all depends on the details of when and where. Outside of snowball earth, clouds provide most of the albedo. Clouds at night are going to absorb most of the surface radiation, blocking it from escaping. What’s more, their temperatures are quite low compared to the surface so they aren’t going to be radiating much of it from the top either.

    At night the thick cloud absorption I think is going to be total relative to bandwidth rather than limited to absorption lines and will extend from visible into the far IR. Emission at the top is much lower due to T being rather low (and it’s a T^4 relation). I think there too will be some continuum as well as line emission but that will perhaps be limited by particle size on the longer wavelength side. Note I’m speculating here as I don’t know the details of what’s been measured and just using common rule of thumb concepts for guidance. Net result is probably going to be a bit of warming at night due to the IR trapping and downward reradiation and perhaps reflection but h2o tends to be rather low emissivity at longer wavelengths so it’s mostly absorbing. Essentially, a cloud is going to absorb more IR than all the co2 could possibly do.

    Correlation does not mean causality. What’s more, effects of CRs on the atmosphere in the form of cloud formation couldn’t possibly last for days, much less for years. To get years, you’ve got to invoke surface energy storage because the atmosphere doesn’t have the heat capacity for that either. Also, that little exponential decay concept comes into play as well and it can invoke quite a bit of limitation as well.

    Relating that correlation to something real could get interesting, if it’s possible.

  377. maksimovich
    Posted Feb 17, 2008 at 11:37 AM | Permalink

    Harrison and Stephenson(2005)found sensitivity in changes to diffuse radiation from cloud cover.

    Changes in DF and the frequency of overcast days represent changes in the weather and the atmospheric energy balance. The decrease in the proportion of direct solar radiation associated with an increase in DF will lead to a local reduction in daytime surface temperature. Further, because the net global effect of cloud is cooling (Hartman 1993), any widespread increase in the overcast days could also reduce temperature. At Reading, the measured sensitivity of dailyaverage temperatures to DF for overcast days is K0.2 K per 0.01 change in DF(for 1997–2004). Consequently the inverse relationship between GCR and solar activity will lead to cooling at solar minimum. This might amplify the effect of the small solar cycle variation in total solar irradiance, believed to be underestimated by climate models (Stott et al. 2003), which neglect a cosmic ray effect.

    There are also to other mechanisms to be taken into account with solar minimum.

    1)Less ozone production due to decrease in UV (and increased ozone attenuation from increasing GCR Nox)

    2)Increased No2 from GCR.

  378. Dennis Wingo
    Posted Feb 17, 2008 at 7:07 PM | Permalink

    (#370)

    Leif I understand landscape for presentations. What I was talking about is that when you download them, they download in portrait mode, which makes the presentation sideways.

  379. Dennis Wingo
    Posted Feb 17, 2008 at 7:19 PM | Permalink

    (#377)

    I install solar power systems and have some pretty good data on the reduction in broad spectrum solar energy on the ground from clouds. A good number for an overcast day is a reduction of 80-90% of the irradiance reaching the ground. That is about how much the drop in power from the solar arrays are.

  380. Leif Svalgaard
    Posted Feb 17, 2008 at 7:49 PM | Permalink

    378 (Dennis): Landscape vs. portrait: Yes, it is a pity that the rendering software can’t turn and scale automatically as appropriate. Nothing I want to do about it. Perhaps the latest version of Adobe reader can do this. Who knows.

  381. cba
    Posted Feb 17, 2008 at 8:35 PM | Permalink

    Dennis,

    The number I’ve seen most often associated with solar power in overcast is about 10% of clear sky power to panels. Panels should peak right near the red/IR area and cut off below about 1000 nm and fall off on the short wavelength end usually in the blue as I recall the wavelength chart. The IR cutoff is rather sharp as it’s where the eV energy drops below the threshold. Cloud reflection could be around 80% + not including atmospheric absorption but I think cloud albedo (reflection and absorption) drops more towards 50%. I would expect the clouds to radiate rather little in the IR due to low T leaving roughly the 10% translucence in the visible.

  382. Leif Svalgaard
    Posted Feb 19, 2008 at 11:55 PM | Permalink

    For what it is worth:

    In a symposium on the potential role of solar variability — increases in heat coming from the sun — held in Boston at the annual meeting of the American Association for the Advancement of Science, experts in solar science, climate modeling, and atmospheric science explored the issues surrounding who or what is to blame for the rapid rate of change.

    There are several possibilities, but the most likely answer is that human industry — that is, heating, cooling, automobile exhaust, manufacturing, and power-generation — is the fundamental culprit. Such activities rely heavily on burning gas, oil, and coal on a massive scale, and the end result includes carbon dioxide, a so-called greenhouse gas that traps the heat radiating from the ground, keeping it from escaping back into space.

    “I’m looking for the millennial scale of solar variability,” said astronomer Sallie Baliunas, a researcher at the Harvard-Smithsonian Center for Astrophysics in Cambridge. She added that “the records do show variability,” such as changes in radioactive carbon-14 abundance and a beryllium isotope in sediment that suggest changes in solar output. “Did the sun cause what we see on the ground?” she asked. “It doesn’t seem so. But there is some fuzziness in the data, which suggests it could go either way. The answer isn’t known at this time.”

    What is becoming known, especially from computer models of global climate, is quite gloomy. Warming that was first noticed in the 1960s has increased steadily, and is probably directly linked to human activities.

    Scientists suspect the changes in the amount of beryllium-10 and carbon-14 found in various layers of sediment reflect solar activity, because the magnetic disturbances associated with sunspots tend to block the normal flow of cosmic rays reaching the Earth from space. The cosmic rays collide with atoms in the Earth’s atmosphere, creating the unusual isotopes; beryllium and carbon thus serve as a “signature” for cosmic-ray and solar activity.

    “Our Sun is a variable star,” said David H. Hathaway, a sunspot specialist from NASA’s Marshall Space Flight Center in Hunstville, Alabama. “It varies by about one-tenth of one percent” in energy output. But “there are suggestions the sun” varies “more than that, because we see it has gone through some periods, such as the Maunder minimum.” During the Maunder minimum, which lasted from 1645 to 1715 and is also known as the Little Ice Age, there was an absence or near-absence of sunpots and northern Europe experienced especially cold winters.

    Baliunas has also based her research on studying surface activity that is detectable on distant stars that are reminiscent of the sun. There is considerable variability in the 60 sunlike stars she has examined, she said, depending on how fast each rotates and other factors.Unfortunately, she added, “there is no model to explain [solar surface activity] on the century-to-millennium time scale,” and long-term changes in solar output need further study.

    According to Casper M. Ammann, a climate modeler at the National Center for Atmospheric Research in Boulder, Colorado, in the years since 1950, “there is no observed trend” in solar radiation. The 11-year sunspot cycle has not been significantly abnormal. This is just part of the reason for the difficulty of determining the sun’s influence on Earth’s climate. Ammann explained that “for the past 150 years people have tried to see whether the monsoons are linked to the 11-years solar cycle,” but without success.

    In essence, he added, it’s now very clear that the atmospheric changes being seen now — global warming — “have nothing to do with changes in solar activity. It’s greenhouse gases. It’s not the sun that is causing this [climate] trend.”
    The Earth’s atmosphere — and its relationship to the sun’s energy output — is so complex that even as warming began, “up until 1960 we couldn’t see it.” But now, he said, since warming has been confirmed, the world’s climate scientists “are probably not overestimating the problem. It’s probably worse than the estimates.”

    In fact, he said, global warming is occurring at an incredibly rapid rate, faster than any previous episodes of climate change known from the paleo-climate data.
    Ammann did add, however, that there is reason to hope that the most dire consequences can be avoided. Although it’s clearly too late to avoid the heating of the earth’s atmosphere, “we can substantially cut [it]” by severely reducing the amounts of carbon dioxide going into the air. “It is absolutely achievable,” he said — if by mid-century societies can generate enough will to make the necessary changes.

  383. Leif Svalgaard
    Posted Feb 19, 2008 at 11:58 PM | Permalink

    For what it is worth:

    “In a symposium on the potential role of solar variability — increases in heat coming from the sun — held in Boston at the annual meeting of the American Association for the Advancement of Science, experts in solar science, climate modeling, and atmospheric science explored the issues surrounding who or what is to blame for the rapid rate of change.

    There are several possibilities, but the most likely answer is that human industry — that is, heating, cooling, automobile exhaust, manufacturing, and power-generation — is the fundamental culprit. Such activities rely heavily on burning gas, oil, and coal on a massive scale, and the end result includes carbon dioxide, a so-called greenhouse gas that traps the heat radiating from the ground, keeping it from escaping back into space.

    “I’m looking for the millennial scale of solar variability,” said astronomer Sallie Baliunas, a researcher at the Harvard-Smithsonian Center for Astrophysics in Cambridge. She added that “the records do show variability,” such as changes in radioactive carbon-14 abundance and a beryllium isotope in sediment that suggest changes in solar output. “Did the sun cause what we see on the ground?” she asked. “It doesn’t seem so. But there is some fuzziness in the data, which suggests it could go either way. The answer isn’t known at this time.”

    What is becoming known, especially from computer models of global climate, is quite gloomy. Warming that was first noticed in the 1960s has increased steadily, and is probably directly linked to human activities.

    Scientists suspect the changes in the amount of beryllium-10 and carbon-14 found in various layers of sediment reflect solar activity, because the magnetic disturbances associated with sunspots tend to block the normal flow of cosmic rays reaching the Earth from space. The cosmic rays collide with atoms in the Earth’s atmosphere, creating the unusual isotopes; beryllium and carbon thus serve as a “signature” for cosmic-ray and solar activity.

    “Our star, the sun, is a variable star,” said David H. Hathaway, a sunspot specialist from NASA’s Marshall Space Flight Center in Hunstville, Alabama. “It varies by about one-tenth of one percent” in energy output. But “there are suggestions the sun” varies “more than that, because we see it has gone through some periods, such as the Maunder minimum.” During the Maunder minimum, which lasted from 1645 to 1715 and is also known as the Little Ice Age, there was an absence or near-absence of sunpots and northern Europe experienced especially cold winters.

    Baliunas has also based her research on studying surface activity that is detectable on distant stars that are reminiscent of the sun. There is considerable variability in the 60 sunlike stars she has examined, she said, depending on how fast each rotates and other factors.Unfortunately, she added, “there is no model to explain [solar surface activity] on the century-to-millennium time scale,” and long-term changes in solar output need further study.

    According to Casper M. Ammann, a climate modeler at the National Center for Atmospheric Research in Boulder, Colorado, in the years since 1950, “there is no observed trend” in solar radiation. The 11-year sunspot cycle has not been significantly abnormal. This is just part of the reason for the difficulty of determining the sun’s influence on Earth’s climate. Ammann explained that “for the past 150 years people have tried to see whether the monsoons are linked to the 11-years solar cycle,” but without success.

    In essence, he added, it’s now very clear that the atmospheric changes being seen now — global warming — “have nothing to do with changes in solar activity. It’s greenhouse gases. It’s not the sun that is causing this [climate] trend.”
    The Earth’s atmosphere — and its relationship to the sun’s energy output — is so complex that even as warming began, “up until 1960 we couldn’t see it.” But now, he said, since warming has been confirmed, the world’s climate scientists “are probably not overestimating the problem. It’s probably worse than the estimates.”

    In fact, he said, global warming is occurring at an incredibly rapid rate, faster than any previous episodes of climate change known from the paleo-climate data.
    Ammann did add, however, that there is reason to hope that the most dire consequences can be avoided. Although it’s clearly too late to avoid the heating of the earth’s atmosphere, “we can substantially cut [it]” by severely reducing the amounts of carbon dioxide going into the air. “It is absolutely achievable,” he said — if by mid-century societies can generate enough will to make the necessary changes.

  384. Dennis Wingo
    Posted Feb 20, 2008 at 12:24 AM | Permalink

    (#380)

    Leif

    I am assuming that these are power points that you are converting to pdf. You can set the direction of the print in your preferences control panel for printing. Just set to landscape and it will print out and store correctly. It is annoying the way it is now.

    Also, in 382/383 Postulate based upon the following proposition.

    You are correct in both your prediction for the minimum, the smoothed sunspot number, and the duration of CY-24. This means that there will be little increase in the sunspot numbers until the end of 09-early 2010. If the southern winter coming up is as cold or even colder than last winter (which set records in many areas around the globe in the southern hemisphere) and if the following northern winter is as cold and or colder than the current one (which is also setting many records around the world for cold), then at that time would you be willing to entertain and discuss possible means whereby the reduction in solar activity has resulted in an influence on climate? It will also be assumed that you are correct about TSI, both cycle to cycle, and over the last 100 years.

    Just think about and maybe talk about possible linkages in the solar/terrestrial system that would amplify solar forces.

    I have my own opinion but would like to hear what you think. Hint: my postulate will be connected to some recent interesting results from Ulysses.

  385. Raven
    Posted Feb 20, 2008 at 3:24 AM | Permalink

    Lief,

    I found this interesting graph:

    More discussion on the source of the data can be found here

    As with most things going on it the atomosphere there is enough correlation to make the data interesting but not enough to make it compelling. In this case, all of the solar minimums appear to be IR emission peaks. Is there anything that could explain this phenomena?

  386. Stephen Richards
    Posted Feb 20, 2008 at 4:27 AM | Permalink

    Lief

    Ammaan is in the AGW/Hockey club. Be aware when quoteing him. :)

  387. cba
    Posted Feb 20, 2008 at 7:20 AM | Permalink

    Raven 385

    That’s a fascinating ref. I hadn’t seen yet and a clever approach.

    Emission peaks during SSC minima would suggest a possible tie in with total cloud cover and the solar cycle/CR/cloud formation connection and increased emission from cloudtops.

    The total radiated emissions shown in the graph are somewhat troubling as these should be in the 235 W/m^2 range for equilibrium. If the graph is accurate for these, then we’re heating up far beyond what any co2 forcing changes could do with a deficit of 20W/m^2 and co2 doubling causing a mere 3-4w/m^2 as well as cyclical variations several times that amount.

  388. Erl Happ
    Posted Feb 20, 2008 at 8:21 AM | Permalink

    Raven, cba
    My twopence worth is below as posted on the site referred to. Looks as if there is a solar driven mechanism here that requires little amplification. An atmosphere undisturbed by short wave radiation and the solar wind develops cloud cover, solar radiation dims. The oceans of the tropics cool. The ITCZ loses strength, warm moist tropical air is drawn towards the low pressure zones of mid latitudes and when the two air masses interact it can generate snow in China and other parts of the world that are normally warm.

    Dr Jim. Looks to me as if your outgoing long wave radiation data is a good fit to the behaviour of the Southern Oscillation Index. Your data shows that warming across the entire tropical region that we describe as an El Nino event occurs when long wave radiation falters. Cooling across the tropics occurs when long wave radiation increases. This observation contradicts the conventional wisdom that ENSO is an internal oscillation of the climate system. In my opinion this is a major breakthrough in climate science.

    Seems to me that some factor is moderating the ability of the atmosphere to retain long wave radiation emitted by the Earth.

    From an analysis of cloud area in the tropics (University of Oxford Global Retrieval of ATSR Cloud Parameters and Evaluation project) against University Alabama Huntsville temperatures in the troposphere I find that cloud area is driven by the fluctuation in temperature in the troposphere. The troposphere may be heated directly by wave lengths shorter than visible light as is apparent in the ionosphere and the stratosphere. Only very short wave lengths are capable of this direct interaction with molecular gases. Sudden stratospheric warmings are well documented. Also well documented are the increases in satellite drag that occur when the atmosphere is inflated as the magnetic influence of the solar wind is felt in the ionosphere. The link between the solar wind and UV radiation is sunspot/coronal hole and solar polar emission activity but the relationship is not a simple one.

    During the La Nina cooling event of year 2000 cloud area fell away dramatically for a single month on two separate occasions. Reduction was about 50% of the average area at that time in the zone 30°N to 30°S latitude. Within these two months (February and May 2000) temperatures in the troposphere as shown by the UAH data rose by 0.5°C and temperatures fell at the surface by 0.1°C. My interpretation is that surface temperatures fell due to reduced inhibition of long wave radiation due to reduced cloud cover. This points to strong activity in absorbing IR radiation by condensate rather than vapour.

    What this points to is a factor driving the “greenhouse effect” that is much more influential than carbon dioxide concentration in the atmosphere. Is it the amount of condensed water that is present in the troposphere as clouds? Is the amount of condensation conditional on temperatures in the troposphere responding directly to wave lengths shorter than the visible? I suggest that the thing to do is to correlate your IR data with a series showing the flux in ultraviolet. With your map reading software you can look at http://www.temis.nl/uvradiation/world_uvi.html Otherwise someone may be able to supply us with a series showing the flux in the strength of radiation shorter than visible wave lengths.

    Seems to me that you could evaluate the CO2 effect by looking for a seasonal signal in IR data. Co2 falls during northern hemisphere summer. So does cloud area. CO2 and cloud rise simultaneously in Southern Hemisphere summer. If you could quantify the cloud effect the remainder is due to the CO2.

    Yes “Experiment and observed data TRUMPS theory, every time.”

    Your breakthrough is to demonstrate a big variation in the flux of heat from Earth to Space. That’s very important. Have the IPCC models the capacity to replicate this? If not, let them re-examine their assumptions and re-evaluate the role of the sun in driving the Earth’s climate.

    Currently sunspot and solar wind activity is very low. The southern skies have been cloud filled all summer. The tropics has cooled, the winter hemisphere has been experiencing historically low temperatures. What then is the expectation for the northern hemisphere growing season for grains? When will the prairies warm?

    In nine of the last twelve solar cycles La Nina cooling has been experienced at solar minimum running through to geomagnetic activity minimum a year to 18 months later than solar minimum.

  389. Raven
    Posted Feb 20, 2008 at 8:48 AM | Permalink

    What I find interesting is that volcanos appears to reduce IR emissions which would imply a heating effect, however, the actual effect is cooling. I suspect this is because the aerosols reduce the incident radiation which results in lower IR emissions.

    This has some interesting implications for GHG theory because human aerosols would cause a similar reduction in IR emissions. This means reduced IR emissions over time cannot be attributed solely to heat trapping by GHGs.

  390. stargazer
    Posted Feb 20, 2008 at 9:58 AM | Permalink

    Lief #382 This to me is troubling. I find it difficult to accept that the sun is not factor in warming…(and ’cause’ of MWP, the fall ino and the ‘rise’
    out of the LIA, and that the world temp seems to have levelled off. etc. etc). Can I ask if all these posts
    have given you any ‘inspiration’ at all… as to where to look for a solar/warming connection. might Raven, cba & Erl be (or any of the others) be ‘on to something’

  391. Andrew
    Posted Feb 20, 2008 at 11:05 AM | Permalink

    “there is no observed trend”

    This statement is demonstrably false. If you were to say “we have conflicting observations, some show a trend and some don’t” you’d be right. ACRIM, for one, has a postive trend. Additionally, Cosmic Rays reach a lowest recorded value around 1991/92. Ammaan is full of it.

  392. Edward
    Posted Feb 20, 2008 at 12:22 PM | Permalink

    The following are email converstations I had with Henrik Svensmark about solar climate effects:

    It looks like Svalgaard and some of the other persons like Gavin Smith
    are mainly referring to solar irradiance and UV. Which might be because
    they refuse to consider the possibility of cosmic rays and clouds. It
    don’t know what to do except to continue to do good science. I any case
    the Sun don’t care what we believe it is doing, and it might now turn
    out that solar activity is decreasing sufficient to make a cooling
    impact. We will see.
    BW,
    Henrik

    —–Original Message—–
    From: Edward McCann [mailto:EVM@nrc.gov]
    Sent: 13. februar 2008 16:08
    To: Henrik Svensmark
    Subject: RE: CERN Concerns

    Please refer to discussions at climateaudit.org concerning Leif
    Svalgaard research. Leif says that the sun varies very little over time
    and, therefore, if the sun plays a part in climate then the earth would
    have to be extremely sensative to it’s small changes. Also refer to
    AGWers such as Mann, Hansen, Jones, etc. Main stream astrophyicists
    such as Leif downplay the sun and Leif is now saying we underestimated
    the strength of pre 1950 solar cycles.

    Leif gave this info at climateaudit about the SORCE meeting which
    downplayed solar effects on climate: kinda promised to report on the
    SORCE meeting, but the Flagstaff crew have alreeady done a good job [as
    was mentioned in this blog]. See
    http://www.lowell.edu/blog/?p=91 and previous postings there

    >>> “Henrik Svensmark” 2/13/2008 8:27 AM >>>
    Dear Edward McCann,

    There have been a large resistance to ideas we are working on over the
    years. If you think that there is an increased resistance, what are you
    referring to. The CERN project is not the only experiment, we have
    already made substantial progress in demonstrating a mechanism here in
    Copenhagen, and experiments are continuing. The CERN project will
    properly not start before 2010.

    Sincerely,
    Henrik

    ________________________________

    From: Edward McCann [mailto:EVM@nrc.gov]
    Sent: 13. februar 2008 14:20
    To: Henrik Svensmark
    Subject: CERN Concerns

    My name is Edward McCann and I am an engineer and phyicist. I am very
    interested in the sun to climate link. I am very concerned about the
    objectivity at CERN. I think you are part of the team at CERN for the
    cosmic ray experiments. Is this correct? How are things going?
    There
    is a huge effort to discredit solar effects on climate which is why I
    am
    so concerned.

  393. Larkin Lowrey
    Posted Feb 20, 2008 at 3:34 PM | Permalink

    I am very curious about historic GCR flux trends. Claims that the sun’s magnetosphere has increased significantly over the last century appear to have been discounted or refuted (at least by Dr. Svalgaard), yet I have seen a fair amount of isotopic data suggesting otherwise. I recently stumbled on the paper linked below. It uses extraterrestrial 44Ti and thus avoids many of the factors that can affect ice core and other terrestrial measurements.

    Long-term solar activity reconstructions: direct test by cosmogenic 44Ti in meteorites
    I. G. Usoskin1, S. K. Solanki2, C. Taricco3,4, N. Bhandari5, and G. A. Kovaltsov6

    PDF

    I suspect that this subject has been beaten to death already, but I would appreciate it if one of y’all could comment on why the data from the above paper does not support the notion that the solar magnetic field has increased over the last century.

  394. GW
    Posted Feb 20, 2008 at 6:19 PM | Permalink

    Leif, Regarding #382 :

    You have read Dr. Schwartz’s paper HEAT CAPACITY, TIME CONSTANT, AND SENSITIVITY OF EARTH’S CLIMATE SYSTEM; perhaps you even had the opportunity to discuss it with him at the AGU conference last fall. Are you personally convinced that whatever warming there has been can be attributed to CO2 to the extent the AGW proponents claim ?

    Have you been reading the other threads on Climate Audit and other related sites which have illustrated the problems with the GISS surface temp database, problems with the actual surface station measurement sites (Anthony Watts) and the problems with the GISS temperature adjustments which supposedly correct the surface station problems (but don’t) ? Add to that the fact that Dr. Hansen had steadfastly refused to release the sourcecode for the GISS temp adjustment algorithms (not sure – he may have finally done so after the tremendous outcry following SM’s discovery last year of the blatant error in the temp database,) and you can not help but get the sense that the rapid increase as stated by Ammann may be a lot less rapid and a lot less large than AGW proponents would like the public to believe.

    Now I know full well that the Southern Hemisphere’s very harsh winter last year, the Antarctic icepak currently being at all time highs, the incredible recovery of the Arctic sea ice following last summer’s meltdown (which was due to unusual air and sea currents according to JPL,) the harshly cold winter in Eastern Europe and Asia, and the current strong La Nina in the Pacific – is likely a pure coincidence with the fact that it is all occurring during the present lull in solar activity – despite Erl Happ’s statement (388) that 9 of 12 past La Nina’s occurred during/near past solar min’s.

    Nevertheless, it becomes difficult to not draw the conclusion, if not be at least skeptical, that the entire AGW movement is less about objective science and understanding, and all about controlling dirty energy and preventing the damage that occurrs in the process of extracting that energy from the crust and then consuming it.

    So, what do you think about all this Leif ? And Erl Happ (388) raised some some interesting points – what do you think of them ?

    Thanks, for your continued time and patience in participating in this blog, and at SS24 too.

    Regards,

    GW

  395. Leif Svalgaard
    Posted Feb 20, 2008 at 10:37 PM | Permalink

    Folks, I’m the road. Will be home Saturday and then have time to look at the comments.
    A general response would be that I never ‘refuse’ to look at other data and that the data cannot be ‘troubling’. Such emotional notions have no place in good science, although as humans we may be led astray by them from time to time.

  396. Gary
    Posted Feb 21, 2008 at 4:03 AM | Permalink

    I thought sensitivity to the sun was pretty obvious. Diurnal variations coupled with summer/winter conditions surely gives enough data to estimate sensitivity. The change in incident radiation at ground level from summer to winter and temperature effects should be enough to determine sensitivities. There seems to be a complete lack of observation in this discipline. I wonder where we would be without the computer?

  397. stargazer
    Posted Feb 21, 2008 at 4:38 AM | Permalink

    Leif #395

    I used the word ‘troubled’ in my post (#390)… and you said in #395 “and that the data cannot be ‘troubling’. Such emotional notions have no place in good science, although as humans we may be led astray by them from time to time.”

    let me explain why I used this word.
    I said I was troubled… but not by the data, if that is what the data says then that’s what it says. (I work building and designing measuring equipment for physics use.) Of course I will believe Solar Astronomers if that is the data. However I have been looking into this Climate stuff for a long time (because I am interested) and what I find ‘troubling’ is that on balance I also ‘accept’ that ‘even’ over the past 10000 years the temp. has been higher (and much lower) than ‘now’ (even with ‘our’ Co2 input) perhaps some 11 times… and on the face of it no-one seems to know why….but it cant be c02

    The two solar Astronomers mentioned in # 395 were (in my view), not totally dismissive….Sallie Baliunas saying “Did the sun cause what we see on the ground?” she asked. “It doesn’t seem so. But there is some fuzziness in the data, which suggests it could go either way. The answer isn’t known at this time.”

    And David Hathaway saying “there are suggestions the sun” varies “more than that, because we see it has gone through some periods, such as the Maunder minimum.” During the Maunder minimum, which lasted from 1645 to 1715 and is also known as the Little Ice Age,

    And again… I feel, (having read all these posts) that you too (please excuse me if I an wrong) also think that there is ‘something else…. you said “A general response would be that I never ‘refuse’ to look at other data” I find this true and all your posts to be enlightening , (and I thank you for being here) this comment is also why I and others are here.we are trying to reconcile ‘the data’ with what we think we know. And yes “although as humans we may be led astray by them from time to time.” might be the case…

    however…I would sooner the world be ‘warmer’…. we can learn to live and cope with that.

    But the real reason I used the word ‘troubled’ is that IF this is the end of the matter… and no-one looks any further… not knowing why the LIA and MWP (as examples) happened, we leave ourselves open to disaster on an unprecedented scale should we be heading toward a major ‘cooling’ of the world as I and others think may be happening. I am not a climatologist nor a solar astronomer which is why some of us are here viewing this, and your other threads. Because (and this is totally human) I am frightened for my family, wife and kids.should a new LIA be happening, because I can see the full ‘potentional’ of this! should it happen, and it makes global warming look good .

  398. Steve McIntyre
    Posted Feb 21, 2008 at 8:44 AM | Permalink

    Caspar Ammann says:

    global warming is occurring at an incredibly rapid rate, faster than any previous episodes of climate change known from the paleo-climate data.

    I wonder how he knows that.

  399. pochas
    Posted Feb 21, 2008 at 9:33 AM | Permalink

    #330 Jim Arndt:

    Leif isn’t it possible that UV interacts with the with the atmosphere to create more ozone which then has a warming effect on the upper atmosphere. This can add to the total warming seen on the surface.

    Just wanted to add to this that ozone’s strongest absorption band is at 9.6 microns, which is right inside the “clear window” for IR emission from the surface directly to space. It seems plausible to me that increased ozone concentrations and/or stimulated emissions from solar radiation would radiate in the clear window, potentially directly to the earth’s surface. Another way of saying this is that strong ozone emissions from the stratosphere will restrict the clear window for radiation to space, producing surface warming via downwelling IR, the same mechanism as for a “greenhouse gas”. Can radiation absorbed by ozone in the UV bands be reemitted in the IR?

  400. cba
    Posted Feb 21, 2008 at 2:29 PM | Permalink

    pochas,

    energy absorbed however is fair game for being emitted as IR or other wavelengths, depending on circumstances such as pressure and temperature.

    for stimulated emissions you really need a population inversion of states to be important.

    You’re not going to get a net energy flow from cold to hot. Also way up there it’s mostly vacuum and not much to absorb or emit even when the T is slightly higher so there’s not a lot of energy to come from there.

  401. pochas
    Posted Feb 21, 2008 at 2:50 PM | Permalink

    cba #400:

    Well, it is the thermosphere – 1500 C with an active sun so we’re not going from cold to hot. And apparently there is enough ozone present to absorb most of the incoming UV.

  402. DeWitt Payne
    Posted Feb 21, 2008 at 3:14 PM | Permalink

    pochas #399,

    Above the troposphere the atmosphere is optically thin for IR but thick for short wavelength UV (less than 0.3 micrometers). I’m pretty sure that absorption in terms of energy by ozone in the stratosphere at 9.6 micrometers of IR emitted by the surface is insignificant compared to the energy absorbed by ozone and oxygen in the UV. Because convection in the stratosphere and above can be neglected, all the energy absorbed must be emitted for the temperature to remain relatively constant in line with observation. If I understand correctly, the balance of emission in the upper atmosphere comes from the 9.6 micrometer band of ozone and the 15 micrometer band of CO2. Lower ozone and energy absorption decreases more than emission and the temperature goes down. Increase CO2 and emission increases but absorption doesn’t change significantly and the temperature goes down.

  403. cba
    Posted Feb 22, 2008 at 7:54 AM | Permalink

    I don’t think this area can even be said to be in LTE. It’s too much of a vacuum for an actual specific temperature to exist and subject to far more than radiative energy.

    There’s just so little of it that it’s just a small factor in satellite orbit corrections. Also, most of that really short uV is coming from the corona which is being heated well above the 6000 k of the photosphere of the sun and it’s incredibly tenuous as well.

    As I remember it, the absorption of a line is related to emission of that line and on the T of the material. Line absorption exists when the T of the absorbing gas is less than that of the continuum emitting body. As that T increases beyond that of the continuum emitting body, the absorption line turns into an emission line. Also, one would expect that under such conditions, the gas up there is likely to have properties of continuum emissions as well.

  404. Erl Happ
    Posted Feb 22, 2008 at 8:52 AM | Permalink

    Since posting 388 when I was excited by the discovery of the relationship betweeen Outgoing Long Wave Radiation (revealed in 385) and ENSO activity I have had time to do a bit more work and clarify things in my own mind. The following is the result.

    ENSO in the Pacific is the strongest manifestation of a phenomenon that occurs right across the tropics. At http://junkscience.com/blog_js/2008/01/12/processing-33-years-of-ir-longwave-data/ you will see an analysis of outgoing long wave radiation (OLR) over time. Take the graph and fit it to the Southern Oscillation Index and it becomes apparent that tropical, oceanic, surface warming is associated with low OLR whereas surface cooling is associated with enhanced OLR. This can also be confirmed with the OLR data at http://www.cpc.ncep.noaa.gov/data/indices/olr

    At: http://www.atm.ox.ac.uk/project/grape/images/images_con.html you can view images of cloud cover and confirm to yourself that periods of high OLR are associated with increased cloud cover in the tropics while El Nino heating events are associated with low OLR. Clouds reflect solar radiation.

    This is not just Pacific ocean peculiarity but a global phenomenon. Look at any current sea surface temperature anomaly map to verify the point.

    When there is no cloud to reflect radiation the tropical ocean acts as a heat sink. You can put a lot of heating into water, which has a high transparency, the ability to mix cool with warm and a high latent heat value without raising surface temperature very much. So, in periods of low cloud cover, surface temperatures will slowly rise as energy is comprehensively absorbed and consequently OLR is low.

    Going back to the cloud images, there are two months in early 2000 in the middle of the La Nina cooling event of that year when there was a lot of cloud across the tropics when, lo and behold, the cloud magically disappeared. These were February and May 2000. In these months temperatures in the troposphere, as shown by the UAH data, rose by 0.5°C. Simultaneously there was a fall in temperature at the surface as shown by the Hadley CRUT data for the tropics between 30°N and 30°S.

    The temperature increase in the troposphere had another source of energy driving it other than emissions by the Earth. There is only one way that mid to upper atmospheric molecules can be warmed other than by OLR and that is by the very short wave radiation that is responsible for ozone in the stratosphere and free ions in the ionosphere/thermosphere.

    The ability of air to hold moisture varies with temperature. The temperature of the air depends directly upon solar activity. Cloud cover depends upon solar activity. The extent of cloud cover determines whether the ocean will warm or cool. The oceans hold a hell of a lot more thermal energy than the atmosphere.

    The Atlantic is a triangle with a relatively large area of Tropical Ocean to garner warmth, a big variably cloud free area adjacent to the Sahara and a small area at high latitudes to focus all that warmth. There is a relatively fast water transit time between the two zones. The north Atlantic is where the so called global warming is focussed.

    In the last three solar cycles there has been a heavy dominance of El Nino heating events. In the previous three solar cycles the reverse was the case. Solar cycle 24 will be La Nina dominant. This cycling between SOI positive and negative has been going on for a long time. While there are no sunspots there will be less very short wave radiation and the total atmospheric column is more compact. Satellites can travel much closer to the Earth without suffering atmospheric drag and therefore require less frequent re-boosting to maintain their orbits.

    Solar minimum is a time for extensive cloud cover in the tropics as atmospheric moisture is condensed due to the gradual cooling of the troposphere. Geomagnetic activity and the solar wind which affect atmospheric density and temperature peak several times over the cycle and not necessarily at sunspot maximum. Subsidiary peaks of sunspot activity are associated with peaks in short wave radiation and the solar wind. For some reason yet to be explained the biggest El Nino event is often experienced soon after geomagnetic minimum. Personally, I think this has something to do with the disturbance of the dense low speed solar wind of polar origin that prevails at solar minimum, a wind that seems to affect the atmosphere to minimal degree while in its steady state. A little crack in the magnetosphere engineered by an atypical wind component emanating from the first disturbances of the sun in the new cycle opens the flood gates as it were.

    Why does dense slow solar wind affect the magnetosphere/ionosphere/thermosphere so little at solar minimum? Perhaps because there is so little short wave radiation at solar minimum to create the population of ionic material that will react to the solar wind.

    Hence, short wave radiation and the solar wind act in tandem. The wind displaces ionic material and neutrals. The tropical zone in daytime suffers a variable degree of inflation and displacement and this affects atmospheric temperature, cloud cover and the amount of solar radiation that makes it to the surface.

    Gradual warming, according to this scenario, would result from a secular increase in short wave radiation or the solar wind between solar cycles. Within cycles we have the ENSO swings to keep us amused. The causation is the same in each case.

  405. DeWitt Payne
    Posted Feb 22, 2008 at 10:43 AM | Permalink

    Charles #403

    I don’t think this area can even be said to be in LTE. It’s too much of a vacuum for an actual specific temperature to exist and subject to far more than radiative energy.

    I’ve lost the reference, but LTE applies for CO2 at least, up to 70 km. This seems to be because the inelastic collision cross section for CO2 and oxygen is unusually high. So the lower stratosphere is in LTE but the mesosphere and thermosphere (if I have them in the correct order) are not.

  406. cba
    Posted Feb 22, 2008 at 1:46 PM | Permalink

    Dewitt,

    Numbers I recall aare something like 45km up is the region where one starts to see the non LTE territory. It’s also lower ionosphere. Thermosphere tends to be identified as starting around 85km up to maybe 400km +. That doesn’t mean LTE can’t exist as high as your number there but it seems strange if it’s only for one or two molecule types up there. I read recently that the lack of collisions and interactions up there tended to start permitting segmentation by mass although what little I’ve looked at concentrations that high up, I’ve not noticed any.

    While I can’t discount offhand that ozone has significant influence, it just seems that we’re dealing with too small a factor to be of consequence.

    Erl’s expose’ just above looks really fascinating. It needs to be fleshed out and verfied.

    Erl,

  407. maksimovich
    Posted Feb 22, 2008 at 2:45 PM | Permalink

    While I can’t discount offhand that ozone has significant influence, it just seems that we’re dealing with too small a factor to be of consequence.

    No? Let us remove it then instantaneously.

    http://www.climateaudit.org/?p=2470#comment-183507

  408. DeWitt Payne
    Posted Feb 22, 2008 at 2:48 PM | Permalink

    Charles,

    Numbers I recall are something like 45km up is the region where one starts to see the non LTE territory. It’s also lower ionosphere.

    Not according to the entry in Wikipedia, which I have no reason to question in this case. The thermosphere (80km up) is where the components start to stratify by mass, i.e. where collision frequency is too low to keep the components mixed so LTE no longer applies. The ionosphere is the upper part of the thermosphere. At 45 km, you’re still in the stratosphere.

  409. cba
    Posted Feb 22, 2008 at 6:31 PM | Permalink

    No? Let us remove it then instantaneously.

    http://www.climateaudit.org/?p=2470#comment-183507

    Maksimovich,

    The link took me to the top of the page and not to a particular comment. A search on the page indicated references to ozone were all to stratospheric ozone, not mesopheric or thermospheric which is what I’m referring to.

    Dewitt,

    D layer ionosphere can be as low as 50km. 45km is the demarcation of mesosphere from stratosphere, up to around 85km or so to the thermosphere.

    D layer is the inner most layer and is responsble for absorption of low HF and MW.

    Lyman uV can deal with NO. Radiation

  410. cba
    Posted Feb 22, 2008 at 7:25 PM | Permalink

    oops, somthing missing from above post.

    Radiation

  411. DeWitt Payne
    Posted Feb 22, 2008 at 7:35 PM | Permalink

    still missing. Don’t use less than or greater than symbols. They are interpreted as defining the start or end of an HTML tag.

  412. Erl Happ
    Posted Feb 22, 2008 at 8:13 PM | Permalink

    cba #406

    Erl’s expose’ just above looks really fascinating. It needs to be fleshed out and verfied.

    With luck by mid next week.

  413. maksimovich
    Posted Feb 22, 2008 at 8:46 PM | Permalink

    re 409

    Sorry a bit obscure.The reference was to a Solar proton event(2006).During a SPE we see NOx(NO+NO2)increase by 40-50 ppb and decreases in Mesospheric ozone by 20-30%.Precipitation and transport sees NOx into the upper stratosphere with subsequent ozone loss of 20%.

  414. pochas
    Posted Feb 22, 2008 at 11:24 PM | Permalink

    Lots to look at on ozone. Here’s the total ozone map:

    ftp://toms.gsfc.nasa.gov/pub/omi/images/global/FULLDAY_GLOB.PNG

    Here’s the UAH temp anomaly map for comparison:

    http://climate.uah.edu/

    Which comes first, the chicken or the egg?

  415. cba
    Posted Feb 23, 2008 at 7:46 AM | Permalink

    Dewitt, 409

    evidently it was the less than symbol.

    radiation less than 1 nm also causes significant effects around d layer. Additional culprits there are micrometeors and galactic cosmic rays.

    All these apparently affect N2 and O2

    segregation by mass happens above 80km where it’s pretty much non LTE

    Somewhere I’ve seen claims of non LTE as low as 45 or 55 km but that may be only under some conditions. More common is it being up to the top of the mesosphere.

    I hope the 3rd time is the charm on this.

  416. JimP
    Posted Feb 23, 2008 at 9:27 AM | Permalink

    I have put together 3 charts comparing the RSS Global Temperatures, the Oceanic Nino Index, and the Outgoing Long Wave Radiation and Cosmic Rays.

    The Oulu Cosmic Ray Station has a long list of publications. The most recent is “Cosmic rays and climate of Earth: Possible Connection.”

  417. JimP
    Posted Feb 23, 2008 at 9:30 AM | Permalink
  418. JimP
    Posted Feb 23, 2008 at 9:32 AM | Permalink
  419. Posted Feb 23, 2008 at 10:32 AM | Permalink

    418 (JimP): how was the missing data in ~1979 handled in the calculation of the 3-month average?

  420. Posted Feb 23, 2008 at 10:45 AM | Permalink

    418 (JimP): what I see is that ONI and LOR are nicely correlated [as they should be - no surprise], and that none of them have any correlation with the cosmic ray intensity [which - to me - is no surprise either].

  421. pochas
    Posted Feb 23, 2008 at 11:27 AM | Permalink

    JimP:

    I tried to find a model for these variables of the form tempAnom = a * NINO + b * RAYS + slope * time

    and was able to get a good correlation coefficient (0.7) but only when the slope, which is essentially an unexplained extraneous trend, controlled the model. I concluded that Nino events and cosmic rays and volcanos do influence sub-decadal LT temps, but they are just bumps and wiggles on the graph. On my best try, slope was 0.0185 deg C per year. The time period I looked at was 1979 – 2005. Also, I was able to visually observe that El Chichon and Pinatubo did cause significant cooling for several years.

    In view of recent trends, “slope” may have changed.

    As a curiosity, I found that the correlation with cosmic rays was the inverse of what you would expect. Higher ray counts correlated with higher temperatures. This does not bother me too much since the earth’s ocean currents add reluctance to the system, and it is logical that the temperature response to cosmic rays will be out-of-phase with the cosmic ray counts.

  422. DeWitt Payne
    Posted Feb 23, 2008 at 1:48 PM | Permalink

    Charles #415,

    I wonder if the confusion over altitude is caused by latitudinal changes. The tropopause is at 10 km at high latitudes and closer to 20 km at the tropics. The temperature profiles in the Archer MODTRAN calculator, however, put the stratopause (defined as the change from warming to cooling again with altitude) at 50 km for both tropics and sub-Arctic. The temperature structure in the stratosphere changes a lot with latitude, though. How about ionization being higher at lower altitudes at the poles because the magnetic field directs energetic charged particles toward the poles?

  423. Erl Happ
    Posted Feb 23, 2008 at 4:00 PM | Permalink

    416 (JimP)
    The correlation between tropical surface temperatures or tropical lower troposphere temperatures on the one hand and ENSO indices on the other is even more compelling. The tropics have the heat surplus. The rest of the globe depends upon the tropical oceans for equable temperatures in winter. So, global temperature is a derivative.

  424. Erl Happ
    Posted Feb 23, 2008 at 4:12 PM | Permalink

    420 (Lief)
    Yes, it looks as if cosmic rays are about as correlated with outgoing long wave radiation, ENSO and surface temperatures as is Total Solar Irradiance.

    The real action lies in the varying proportion of solar energy that reaches the surface as mediated by cloud cover, as mediated in turn by the variable degree of direct heating of the troposphere by short wave radiation. There is no cut off at 30Km of altitude for short wave radiation. It’s a myth.

  425. Posted Feb 23, 2008 at 4:19 PM | Permalink

    424 (Erl):

    Yes, it looks as if cosmic rays are about as correlated with outgoing long wave radiation, ENSO and surface temperatures as is Total Solar Irradiance.

    Yes, which is ‘not at all’ correlated.

  426. maksimovich
    Posted Feb 23, 2008 at 5:34 PM | Permalink

    Re Jim 416

    You are using wrong hemisphere dataset for CR.Oulu should have some description on Stoermer Theory and the “moving” geometric qualities,

    Lief links to Bieber 2007 in the previous post(ls 2)or you can find the datasets here http://neutronm.bartol.udel.edu/~pyle/bri_table.html You may also want to use the South African stations to see the differences in the count rates.

    In addition LW is not the correct “sign” for CR modulation,The nitrogen oxides are surface radiation attenuators(they reduce IR reaching the surface) This is not controversial this is well understood.

  427. Posted Feb 23, 2008 at 6:48 PM | Permalink

    426 (Max…): For cosmic rays it doesn’t matter which station you use. The counting rates vary a lot, but the solar cycle modulation scales with the counting rate and any correlations with ONI, clouds, LOR or anything else would not be markedly affected.

  428. cba
    Posted Feb 23, 2008 at 7:52 PM | Permalink

    Dewitt,

    Since the boundaries are defined by the physical conditions, then the altitudes must vary by lattitude because of the differences – mag. field. I’m guessing here but I think there are variations in time for conditions that cause these conditions to change some by altitude as well. Hence I think you’re right that they vary by lattitude but I think they also vary from other external factors as well.

    BTW, there’s a post above with ref. and chart for outgoing longwave radiation that evidently was derived from ‘reverse engineering’ of some graphs to yield a total OLR by yr. I’m not sure it was done that way as I found a noaa site that permits the data to be displayed in tabular form. It seemed that there could even be a problem as surface area might have been treated uniformly by lattitude due to the graph approach – although that wasn’t indicated in my reading.

  429. cba
    Posted Feb 24, 2008 at 5:14 PM | Permalink

    #385 Raven

    The graph looks interesting but I think it is in significant error.

    I’ve downloaded some of the NOAA text file versions of the various graphs – monthly averages by lat/long.

    Weighing the w/m^2 averages by latitude and correcting for surface area, using a fairly good approximation for the earth being an oblate spheriod, I determined the total radiated power and divided by the total area.

    For the four time frames I’ve tried, the average emission is in the 235 W/m^2 range give or take less than 10 W/m^2. This is also in the ballpark of what should be expected from a static one dimensional or zero dimensional radiative model.

    What is rather interesting on the numbers too is that variations are mostly showing for the northern hemisphere rather than southern. Also, doing 2 Jun and 2 Jan number sets 30+ yrs appart, there was little variation in SH emissions and significant in the NH. An additional initial curiosity is that there seems to be more OLR in June than in Jan, where the eccentricity has the earth significantly closer to the sun and with higher TOA insolation despite the NH being in winter – away from the sun.

    Downloading the data is a bit of a pain. I’m scarcely 1/6th of the way through and haven’t really started processing it yet except for the 4 examples. I suspect that both the numbers are off absolutely on the supplied graph and that there may be some error in processing the values. It’s also possible that the total variations are incorrectly stated and even the graph line may look different once it is redone with correct data that is latitude weighted.

  430. stargazer
    Posted Feb 25, 2008 at 1:12 AM | Permalink

    another paper out…

    http://cc.oulu.fi/~usoskin/personal/CRAS2A_2712.pdf

  431. kim
    Posted Feb 25, 2008 at 6:06 AM | Permalink

    Thank you, J. Usoskina, for the clarity.
    =======================

  432. Erl Happ
    Posted Feb 26, 2008 at 2:45 PM | Permalink

    Leif,
    Is there data on the variation in short wave radiation over time that is easily accessed? If so where?

  433. Raven
    Posted Feb 26, 2008 at 3:33 PM | Permalink

    http://cc.oulu.fi/~usoskin/personal/CRAS2A_2712.pdf

    I don’t see this paper as a good paper for backers of the CR hypothesis because it points out how poorly CRs correlate with cloud cover over the short term. The only time range where CR climate correlations hold up is in the past where we have no cloud cover data that could refute it.

    That said, the correlation on the longer time scales is compelling – if these is a causal effect then it is likely hidden by other climate processes and may be impossible to isolate.

  434. Raven
    Posted Feb 26, 2008 at 3:46 PM | Permalink

    cba says:

    Weighing the w/m^2 averages by latitude and correcting for surface area, using a fairly good approximation for the earth being an oblate spheriod, I determined the total radiated power and divided by the total area.

    The person who produced the original graph many have forgot that the earth is a sphere. I did find it strange that I had never seen that data before. However, variations of 10 w/m2 are still interesting when compared to the theoretical forcing of CO2 which is around 4 w/m2 – especially if there is a correlation with solar minima. The differences in NH/SH are also interesting but may be simply a result of greater albedo caused by snow on land.

  435. Posted Feb 26, 2008 at 4:25 PM | Permalink

    http://xxx.lanl.gov/abs/0802.1584

    Magnetic cycles of the planet-hosting star tau Bootis
    Authors: J. F. Donati, C. Moutou, R. Fares, D. Bohlender, C. Catala, M. Deleuil, E. Shkolnik, A. C. Cameron, M. M. Jardine, G. A. H. Walker
    (Submitted on 12 Feb 2008)

    Abstract: We have obtained new spectropolarimetric observations of the planet-hosting star tau Bootis, using the ESPaDOnS and NARVAL spectropolarimeters at the Canada-France-Hawaii Telescope and Telescope Bernard-Lyot. With this data set, we are able to confirm the presence of a magnetic field at the surface of tau Boo and map its large-scale structure over the whole star. The overall polarity of the magnetic field has reversed with respect to our previous observation (obtained a year before), strongly suggesting that tau Boo is undergoing magnetic cycles similar to those of the Sun. This is the first time that a global magnetic polarity switch is observed in a star other than the Sun; we speculate that the magnetic cycle period of tau Boo is much shorter than that of the Sun.
    Our new data also allow us to confirm the presence of differential rotation from the latitudinal shearing that the magnetic structure is undergoing. The differential rotation surface shear that tau Boo experiences is found to be 6 to 10 times larger than that of the Sun. We propose that the short magnetic cycle period is due to the strong level of differential rotation. With a rotation period of 3.0 and 3.9 d at the equator and pole respectively, tau Boo appears as the first planet-hosting star whose rotation (at intermediate latitudes) is synchronised with the orbital motion of its giant planet (period 3.3 d). Assuming that this synchronisation is not coincidental, it suggests that the tidal effects induced by the giant planet can be strong enough to force the thin convective enveloppe (though not the whole star) into corotation and thus to play a role in the activity cycle of tau Boo.

    Comments: MNRAS, in press
    Subjects: Astrophysics (astro-ph)
    Cite as: arXiv:0802.1584v1 [astro-ph]

  436. Posted Feb 26, 2008 at 5:44 PM | Permalink

    435: Leif

    If true, can we conclude that our suns solar cycle is closely related to the effects of the Jovian year (10.6 years), and that papers to that effect are supported??

    Tom

  437. Posted Feb 26, 2008 at 6:41 PM | Permalink

    436 (tucker): not really, as the BIG difference is the size and closeness of the planet of tau Boo. Jupiter is just to small and too far away. Venus has a larger effect than Jupiter, but both are way too small to make any difference. You see, it is not that planets in principle cannot influence their star, but they have to be big enough and close enough, and our planets just aren’t.

  438. cba
    Posted Feb 26, 2008 at 6:42 PM | Permalink

    Raven,

    Looking at the data I’ve downloaded and processed which consists of 1 mo. averages quarterly, The data looks like a particular month varies by perhaps 5 W/m^2 over 10 years. It is highly cyclic with july being the highest value and Jan. being the lowest. This is for 1974-1986 so far. Typical values are around 235-239 for july and 225-229 for january. There is also almost that much variation going on over time for a given month and it looks like lower frequency variations drifting higher and lower. July reaches almost 240 W/m^2 one year and drops to near 234 W/m^2 at the low point.

    Considering there should be more energy coming in while the earth is closer to the sun in January, it’s strange to me that this is the lower relative emission time always.

    To get the rest of the data processed is going to take more time.

    There is clearly something going on with the data. Exactly what remains to be seen. However, with a monthly avg. it sees that variations of 5W/m^2 both during the year and beyond the year suggests that activities could well be self regulating and are not constrained to a fixed rate nor is there any runaway conditions occuring with that much variation in the OLR.

  439. cba
    Posted Feb 26, 2008 at 6:53 PM | Permalink

    Leif,

    Just a dumb question but what if the effect is magnetic field rather than gravity?

  440. Posted Feb 26, 2008 at 9:36 PM | Permalink

    438 (cba): Isn’t the global temperature highest in July (NH and all that)? So, emissivity goes with the temperature.

  441. Posted Feb 26, 2008 at 9:39 PM | Permalink

    439 (cba): be more specific. What effect? in general magnetic fields falls of with the cube of the distance so are even weaker than gravity.

  442. cba
    Posted Feb 26, 2008 at 10:00 PM | Permalink

    Leif,

    emissivity no not really, emission yes of course but…..

    This is the world wide average. Difference is going to be how much land vs how much ocean is in each hemisphere. Also, perihelion is currently in Jan. bringing in more solar insolation then.

    While I don’t know when the highest world wide temperature is, it sure is confusing to think it would not be in Jan since there should be higher insolation due to the reduced distance to the sun. Also, there should be more absorption due to more ocean – which has lower reflectivity than land – unless there’s more clouds in the way reducing the net insolation (albedo).

    Net result is a variation of around 5 W/m^2 between the seasons and variations of around 5 W/m^2 during the same season over a period of a few years.

    I would assume the variation is due mostly to a change in mean T. However, the possibility exists that there is also variation due to trapping of outgoing thermal radiation by clouds (perhaps in combination with reduced insolation absorption from higher albedo).

    Net result is emissions are not constant, even during the same time of year and the variations are larger than that expected from a co2 doubling, regardless of the actual cause. At least that is what seems to be the case at present.

  443. a scientist
    Posted Feb 26, 2008 at 10:02 PM | Permalink

    A list of all comments in the Svalgaard threads that relate to cosmogenic isotope variations and cite a literature source.

    Svalgaard #1


    Leif: “10Be and 14C back to ~1500 match quite well”


    D. Patterson: “climate modulation of the 10Be solar activity proxy”


    Leif: “similarity of IMF derived from 10Be, 14C, and Sunspot numbers”


    Leif: “to infer the strength of the heliomagnetic field from the cosmic ray intensity” held problematic


    Leif: “interpretation of the cosmogenic isotopes”: link to Comment on McCracken.pdf


    David Archbald: “Central England Temperature rose from … 1696 to … 1732” said “ties to C14 drop”


    Maksimovich: wheat price and 10Be


    Leif: Muscheler 2005; Solanki et al.; radiocarbon record


    Leif: Beiber et al.; assumption that “all the changes [in GCR] are due to the sun” should be re-considered

    Svalgaard #2


    Leif: Crowley & Hegerl 2008, Be10 and volcanism (including 1696)


    Leif: possible “volcanic contamination of the 10Be signal [from Greenland]” (McCracken)


    Leif: CR reconstructions: McCracken and Beer 2007; Usoskin et al. 2002


    Leif: Residence time of aerosols


    Maksimovich: GCR and C14

    Svalgaard #3


    Leif: CR proxies “difficult to calibrate”


    Leif: HMF as reconstructed from 10Be: strange dips, possibly volcanism


    Leif: Muscheler et al 2005 on calibration of C14 record


    Usoskin: Solanki et al 2005 reply to Muscheler et al 2005


    Usoskin: CA comment on McCracken & Beer 2007 and Muscheler et al 2007


    cross linked


    Leif: Sallie Baliunas interprets “carbon-14 abundance and a beryllium isotope in sediment”


    Larkin Lowrey: Usoskin et al 2006, GCR flux trends and extraterrestrial 44Ti

  444. Posted Feb 27, 2008 at 8:20 AM | Permalink

    snip

    Steve: Leif has said repeatedly that there is no discernible planetary influence on solar activity. It’s not a topic that I am prepared to discuss on this site.

  445. Erl Happ
    Posted Feb 27, 2008 at 8:38 AM | Permalink

    442 (cba)
    Is it not possible that emissions of Outgoing Long Wave radiation vary in the first instance according to changing albedo. With 100% albedo emission will equal incoming energy. Drop it to 30% and the ocean is a sink to absorb and store energy and emissions fall accordingly. As temperatures rise emission from the ocean via heat loss processes including evaporation and release of latent heat of condensation gradually increase to bring emissions up to an equilibrium figure. With an average albedo of 30% a swing in cloud cover brings the ocean in and out as a heat sink with emissions varying accordingly. Under this scenario a fall in emissions should be accompanied by a rise in ocean temperatures.

    This is the pattern of ENSO change. However, factor in a larger swing in albedo for this specific region (the tropics).
    ENSO is therefore an oscillation in the climate system that is linked to changes in albedo. The change in albedo is forced by an external agency which renders this oscillation not at all an ‘internal’ oscillation. The force involved is short wave energy from the sun. It creates the ionosphere, heats the stratosphere and penetrates to the surface.

    Climate science needs to revise it’s assumptions.

  446. Posted Feb 27, 2008 at 1:05 PM | Permalink

    445 (Erl):

    The change in albedo is forced by an external agency which renders this oscillation not at all an ‘internal’ oscillation. The force involved is short wave energy from the sun. It creates the ionosphere, heats the stratosphere and penetrates to the surface. Climate science needs to revise it’s assumptions.

    That FUV and shorter “creates the ionosphere and heats the stratosphere” is not in doubt, but it does not penetrate to the surface and there is no evidence that it forces a change in the albedo [what mechanism?] so it is not clear that “Climate science needs to revise it’s assumptions” based on that. That revision is needed may be true [or not] but the not because of this.

  447. cba
    Posted Feb 27, 2008 at 1:31 PM | Permalink

    Erl,

    The clouds forming a higher albedo are also likely to affect the OLR, blocking radiation from the surface, perhaps emitting more LR from further up. Overall, they cause more cooling than heating.

  448. Erl Happ
    Posted Feb 27, 2008 at 4:29 PM | Permalink

    Interesting post at http://icecap.us/images/uploads/This_La_Nina_Likely_To_Have_Legs.pdf re the course of the current La Nina and the relationship between ENSO and the PDO.

    Aa index still trending down.

    Cloud cover this southern summer has been very strong and Australia flooding.

    cba when you get the data for Jan Feb 2008 I think that you will discover that OLR is above the norm.

  449. pochas
    Posted Feb 27, 2008 at 5:27 PM | Permalink

    I was curious to see whether any of that 9.6 micron radiation from ozone actually shows up at the surface, and sure enough vis figure 1, wave number 1041.7, right in the middle of the “clear window”, link here.

  450. pochas
    Posted Feb 27, 2008 at 5:33 PM | Permalink

    #449 (me) So sorry – wrong link. right link

  451. Erl Happ
    Posted Feb 27, 2008 at 7:26 PM | Permalink

    Leif 446

    there is no evidence that it forces a change in the albedo [what mechanism?] so it is not clear that “Climate science needs to revise it’s assumptions” based on that. That revision is needed may be true [or not] but the not because of this.

    Supporting data and argument coming shortly. Need to post the paper to an accessible point.

  452. Raven
    Posted Feb 27, 2008 at 7:27 PM | Permalink

    Erl Happ says:

    cba when you get the data for Jan Feb 2008 I think that you will discover that OLR is above the norm.

    A falsifiable hypothesis – excellent! The Jan-Feb data will be interesting.

  453. cba
    Posted Feb 27, 2008 at 8:08 PM | Permalink

    sorry guys not at present time. The data currently avail. ends around mid to late 2007. That suggests Feb of 2008 may not happen for a year or two.

    I would expect that if T is falling, then possibly albedo is up and/or OLR is
    up (for reasons other than a T^4 increase).

    Since all the January values I’ve seen so far are much lower than july’s values.

  454. nevket240
    Posted Feb 27, 2008 at 10:33 PM | Permalink

    Leif:
    thanks for your time & good nature in this thread. Have you had the opportunity to read this article??

    http://www.warwickhughes.com/agri/Solar_Arch_NY_Mar2_08.pdf

    regards.

  455. Posted Feb 27, 2008 at 10:46 PM | Permalink

    454 (nev…): the article uses “the Total Solar Irradiance from the Hoyt and Schatten reconstruction”. This [old] reconstruction is not considered valid any longer by most workers in the field, so anything based on it is suspect.

  456. Posted Feb 27, 2008 at 10:56 PM | Permalink

    455 (nev…) David A’s article doesn’t actually use the TSI for anything [just piling on data numbing the reader], but more critical is the prediction [or assumption of a cosmic ray count significantly higher this minimum than previous minima. The latest count seems to indicate a slight decrease in the last month or so [although too short to be sure]. The run of cosmic rays in the next year or so seems to be a sharp test. We shall see.

  457. Posted Feb 27, 2008 at 11:15 PM | Permalink

    Folks, it is always nice to put some faces to names, so here is a link to some photos from the SORCE meeting in Santa Fe, a couple of weeks ago:

    http://lasp.colorado.edu/sorce/news/sns/2008/sns_feb_2008.pdf

  458. Erl Happ
    Posted Feb 28, 2008 at 5:51 AM | Permalink

    Leif,
    Thanks for that link to the SORCE papers. You are actually a pretty handsome guy in a nice Nordic sort of way and the papers offer lots of food for thought.

    Heres a few falsifiable hypotheses.
    Temperatures in the lower troposphere to spike upwards (simultaneously cloud cover diminishes and rainfall falls away) when short wave radiation (140-154 nm) increases, and also with any marked rise in geomagnetic aa index or during forbush events when cosmic rays disappear due to an increase in geomagnetic activity. This is conditional upon a decent floor to short wave radiation and geomagnetic activity as generally occurs when the SOI is below zero so that there is a good population of ions in the ionosphere, the atmosphere in the tropics is in an inflated condition and satellite drag therefore above average.

    Raven, cba
    I don’t mind stretching to accommodate the need for falsifiable hypotheses. Try this one that should be testable with data from any period: Outgoing long wave radiation to vary inversely with the Oceanic Nino index and directly with the Southern Oscillation index.

    Sorry, last minute glitches discovered in my expression so should post the promised paper tomorrow.

  459. Posted Feb 28, 2008 at 8:19 AM | Permalink

    458 (Erl): The papers are [of course] the most important and give a good view of the State Of The Art. Highly recommended.

  460. Dennis Wingo
    Posted Feb 28, 2008 at 12:00 PM | Permalink

    Leif

    Have you seen this paper?

    Comments?

    http://journals.cambridge.org/action/displayAbstract?aid=1739884

  461. Posted Feb 28, 2008 at 12:36 PM | Permalink

    460 (Dennis): yes; this is not the first paper that mentions this. Joan Feynman [e.g.] has talked about the Mayas [and others] in a similar context. For whatever it is worth, the heliomagnetic field was a bit [25% ?] lower [on par with the Dalton minimum ~1800] than the average for the last 40 years. I don’t know if there is a connection. Somebody [I'm sure] will jump on this and take it as confirmation of pet ideas, but the droughts seem to be a regional rather than global thing, so who knows….

  462. Erl Happ
    Posted Feb 28, 2008 at 2:16 PM | Permalink

    461 (Dennis) 461 (Leif)
    RE: Collapse of the Maya: In the northeastern Caribbean, sea-surface salinity (SSS) was lowered.

    NASA news item from last year

    During the first week and a half of October 2007, frequent heavy rains spawned by a persistent trough of low pressure stretched across the northwestern Caribbean, leading to floods and mudslides in several countries. The torrential rains were responsible for at least 47 fatalities in Haiti in villages north of the capital, said the AFP wire service on October 14. The poor island nation is especially vulnerable to mudslides due to deforestation in its mountainous regions. Cuba reported a 24-hour rainfall total as high as 306 millimeters (about 12 inches). The same weather pattern also impacted Jamaica and Central America. According to Reuters, a rain-induced landslide in Costa Rica buried 14 people, and at least 20 people had died across Central America as of October 14.

    This is the sort of rain that might reduce salinity in the ocean. It illustrates the migration of the convection zone in the Atlantic when the tropical ocean cools. East African rainfall responds to a similar migration in the Indian Ocean. We focus on the Pacific because of the stronger fall in sea surface temperatures but the phenomenon is global. A fall in temperatures in Greenland is to be expected.

  463. David Smith
    Posted Feb 28, 2008 at 5:31 PM | Permalink

    Pretty good recent OLR data is available here and here .

  464. Dennis Wingo
    Posted Feb 28, 2008 at 7:34 PM | Permalink

    Leif

    Thanks, anxiously awaiting your next update on your site! (just get them in landscape mode!)
    :)

  465. Posted Feb 28, 2008 at 8:24 PM | Permalink

    464 (Dennis):There is new stuff on my site. Papers will be in portrait, but presentations will (alas) be in landscape because that is the format of the screen on the wall and when giving a talk one does not want to waste time fiddling with turning the image.

  466. Dennis Wingo
    Posted Feb 28, 2008 at 10:12 PM | Permalink

    Leif

    Thank you very much. I am really intrigued with the decline/rise shape for the c23/24 transition. It does seem so far to support the low cycle prediction.

  467. Erl Happ
    Posted Feb 28, 2008 at 10:44 PM | Permalink

    406 (cba) I hope that this paper verifies and fleshes out the comment that I offered with considerable bravado in 404. The study is of a short period that encompasses the start of cycle 23, the La Nina at solar minimum, the El Nino of 1997-8 and the La Nina that followed in 1999-2000. A longer period would be better. But, as a pilot project to suggest areas worthy of investigation by others more skilful than I, it may suffice.

    For the data and the discussion: http://www.happs.com.au/downloaders/Cloud_temp_tropo.pdf

  468. Posted Feb 28, 2008 at 11:12 PM | Permalink

    466 (Dennis): Yes, it does look good for a low cycle [unless the Sun pulls a stunt on us]. The Solar cycle Prediction Panel is going to have a telecon on Monday to discuss this and other matters. The spotless day argument on the last page is also strong IMHO.

  469. Dennis Wingo
    Posted Feb 29, 2008 at 1:00 AM | Permalink

    Leif

    I particularly think that the ten day spotless charts are extremely interesting as I watch the daily reports and watch the number of 10+ spotless days increase. You know though I keep it in the back of my mind the minimum preceding cycle 19 that had a lot of spotless days including one stretch that I think was 43 days. It will be interesting to watch. I mourn for the loss of data from Ulysses.

  470. Posted Feb 29, 2008 at 6:24 AM | Permalink

    469 (Dennis): just to verify that my explanation of the graph was clear. What is shown is NOT how many intervals of 10 spotless days there was, but simply the accumulated number of single spotless days since the 10th spotless day after the maximum. The longest stretch of spotless days does not enter the picture as such. Ulysses will be missed, but then it was already operating past its nominal lifetime.

  471. cba
    Posted Feb 29, 2008 at 7:45 AM | Permalink

    Erl,

    I downloaded it and looked at it briefly this morning. I will try to go through it in a bit more detail.

  472. Posted Mar 1, 2008 at 8:28 AM | Permalink

    In the debate over solar influence, cosmic rays are sometimes ‘blamed’ for provided condensation nuclei. Here is a different twist:

    Bacteria – The Main Ingredient in Snowflakes, Scientists Say
    By Max Brenn
    15:44, February 29th 2008

    One might rethink playing with snow or walking in the rain as a new study by scientists from the Louisiana State University revealed that snow and rain might form mostly on bacteria in the clouds.

    Scientists have long known that the ice crystals in clouds, which become rain or snow, need to cling to some kind of particle, called ice nucleators, in order to form in temperatures above minus 40 degrees Celsius.

    Microbiologist Brent Christner at Louisiana State University sampled snow from Antarctica, France, and the Yukon and found that as much as 85 percent of the nuclei were bacteria, he said in a telephone interview with the Associated Press.

    “Every snow and ice sample we’ve looked at, we found biological ice nucleators. Here’s a component that has been completely ignored to date,” Christner said.

    The most common bacteria found was Psedomonas syringae, which can cause disease in several types of plants (tomatoes, green beams and other similar plants). The bacterium was found in 20 samples of snow from around the world and subsequent research has also found it in summer rainfall in Louisiana.

    Scientists have sought ways to eliminate this bacterium in time. Now they wonder whether this elimination would result in less rain or snow, or would soot and dust be the major generators of precipitation.

    “The question is, are they a good guy or a bad guy. And I don’t have the answer to that,” Christner said, quoted by the same source.

    One thing is for sure. Bacteria that infect plants may multiply on the plants’ leaves and drift into the atmosphere. These bacteria could then cause precipitation and land on another plant, where the life cycle could continue, Christner said.

    Virginia K.Walker, a biologist at Queen’s University in Kingston, Ontario, Canada said other studies have found bacteria serving as snow nuclei, but this is the first to identify it as Pseudomonas.

    “It’s one of those great bacteria…you can find them anywhere. They are really interesting,” Walker said.

    The study, supported by a Louisiana State University research grant and by the National Science Foundation and the Earth Institute at Columbia University, was published in today’s edition of the journal Science

  473. Francois Ouellette
    Posted Mar 1, 2008 at 8:43 AM | Permalink

    Leif,

    There is also an interesting link between phytoplankton, aerosols and clouds. Biological activity of phytoplankton leads to the production of DMS (Di-Methyl Sulfide), which in turn results in aerosols that are launched in the atmosphere. It has been shown to be directly correlated with low clouds. See the work of R.A. Cropp and A.J. Gabric for details.

  474. cba
    Posted Mar 1, 2008 at 2:24 PM | Permalink

    I don’t think I’m ready for that one.

    Also, I find it odd that such heavy material can get mixed in uniformly with the atmosphere and find its way up through the cloud to where water is forming – not to meniton it’s fairly cold up there. I’d prefer to believe that the bacterial is being picked up on the way to the ground – and even that is making me want to hold my breath indefinitely.

  475. Erl Happ
    Posted Mar 3, 2008 at 3:41 AM | Permalink

    cba Are you ready for this one?

    Image at : http://i249.photobucket.com/albums/gg220/erlandlong/Radiationandcloud.jpg

    Data from http://lasp.colorado.edu/sorce/news/2008ScienceMeeting/
    Comparison of Long-Term
    Solar Ultraviolet Irradiance Data
    Set and Proxy Model Data
    Matthew DeLand and Richard Cebula
    Science Systems and Applications, Inc. (SSAI)
    SORCE Science Team Meeting
    Santa Fe, NM 5-7 February 2008

    Three month averages of the SOI and the aa index are centred on the middle month.
    The small blue arrows record the upper turning points in the Southern Oscillation Index (sign reversed) as exhibited in the lower figure. It is apparent that these turning points occur at times of peak radiation intensity in the 150nm to 154 nm band of the Far Ultraviolet.
    The large black arrows record slope and duration of the major La Nina events that occur shortly after geomagnetic and short wave radiation minima.
    The black rectangles define the period where the aa index of geomagnetic activity is relatively low.
    The red rectangles in the upper diagram show the duration of the last three solar cycles in terms of sunspot activity.
    Orange rectangles exhibit an inverse relation between geomagnetic activity and the SOI. Elsewhere, outside the area of the black and red rectangles the relationship between the two appears to be direct.

    Background:
    From Wikipedia: THE TROPOSPHERE

    The word troposphere derives from the Greek “tropos” for “turning” or “mixing,” reflecting the fact that turbulent mixing plays an important role in the troposphere’s structure and behavior. Most of the phenomena we associate with day-to-day weather occur in the troposphere.

    The troposphere is the lowest portion of Earth’s atmosphere. It contains approximately 75% of the atmosphere’s mass and almost all of its water vapor and aerosols.

    The average depth of the troposphere is about 11 km (7 miles) in the middle latitudes. It is deeper in the tropical regions (up to 20 km (12 miles)) and shallower near the poles (about 7 km (4 miles) in summer, indistinct in winter).

    The region of the atmosphere where the lapse rate changes from positive (in the troposphere) to negative (in the stratosphere), is defined as the tropopause.

    Temperatures decrease at middle latitudes from an average of 15°C at sea level to about -55°C at the beginning of the tropopause. At the poles, the troposphere is thinner and the temperature only decreases to -45°C, while at the equator the temperature at the top of the troposphere can reach -75°C

    Interpretation:
    It is salutary to consider just how thin the troposphere is and to relate its height to the distance to a local landmark that can be reached by walking. As noted above it contains 75% of the atmosphere and almost all the water vapour. The presence of water vapour is a defining characteristic of the tropopause because it acts as a refrigerant gas absorbing heat at the surface and transferring it to altitudes where radiation to space is relatively unfettered. Secondly, condensed moisture forms clouds which absorb and reflect incoming and outgoing solar radiation. Only half the available solar radiation is geo-effective with clouds thought to contribute a loss of about 27% on the average. At times when the tropical oceans have little cloud cover outgoing long wave radiation falls as the oceans absorb heat and the Earth warms. (See paper referred to in post 467)

    The temperature at the tropopause is a function of two main factors, the first being the ease of loss of radiant energy emanating from beneath (fundamentally depending upon molecular density) and secondly the penetration of electromagnetic short wave radiation from above that imparts energy directly to atmospheric molecules.

    That short wave energy impacts the troposphere beneath the tropopause is evidenced by the presence of ozone within the troposphere. The temperature inversion that occurs at the tropopause is the result of energy imparted to atmospheric molecules sufficient to reverse the fall in temperature with altitude. Energy is imparted to air beneath the inversion point as well as above it. This should give rise to a gradual fall in the lapse rate of temperature with altitude in the upper troposphere.

    As can be seen in the figure above, the irradiance at 150-154 nm varied by about 30% between the trough and peak of the solar cycle over the period monitored. The SOI is strongly related to temperatures across the entire tropics between 30N Lat and 30S Latitude. It is the heat garnered in the tropics that determines how much can flow to those parts of the Earth cooler than it’s average temperature, by and large those parts at latitudes above 30°. The conjunction of peaks in the (sign reversed) Southern Oscillation Index with peaks in short wave radiation is direct evidence of an effect of short wave radiation on atmospheric transparency to solar radiation in the broad.

    It is not suggested that the particular wave lengths 150-154 nm are present in the middle troposphere and are the agents of change in this instance. It is presumed that most frequencies at these very variable short wave lengths change in a manner that is parallel.

    The mechanism whereby short wave radiation impacts temperatures in the tropics is probably very simple. It is likely via direct warming of the atmosphere at cloud level impacting relative humidity, condensation and therefore the contribution of cloud to the Earth’s albedo.

    The fast, sustained and relatively compact increase in radiation at the time of the 1997-98 El Nino warming event is noticeable. This warming is independent of geomagnetic activity relating simply to strong increases in short wave radiation.

    Peaks in FUV radiation intensity are closely related to many peaks in geomagnetic activity (data not shown). It is possible that the solar wind is contributing to the heating effect if there is a redistributive effect on neutrals over the tropics. However, the effect is likely to be weak. Some peaks in geomagnetic activity that occur at times other than peak FUV radiation intensity seem to be unrelated to the timing of changes in the SOI index. Despite this observation, it appears that the concurrence of geomagnetic activity and peaks in FUV radiation mean that geomagnetic indices are a guide to trends in the SOI, particularly between Solar Maximum and solar minimum. It is observed that many solar minima show low geomagnetic activity for a variable period. It is these cycles that are likely to exhibit prolonged minima in short wave radiation and consequent strong La Nina’s at solar minimum.

    Since ENSO phenomena are demonstrably driven by events external to the Earth itself climate science can now abandon the notion that these events are due to an internal oscillation in the climate system. Furthermore, we can look to change in the phenomena driving ENSO events for an explanation of the frequency of El Nino heating events over the last three solar cycles (since 1976).

    Is ultraviolet light the sole agent for change in temperature in the tropics? No, there are powerful factors that can accelerate and multiply change once it is underway. These include the dynamics associated with the intertropical-convergence zone, the Hadley cell, the flow of moisture into the stratosphere and the consequent effect upon reflective aerosols and the concentration of ozone in that layer. The prime dynamic affecting ozone persistence is probably its solubility in water.

    From Wikipedia: The electromagnetic spectrum of ultraviolet light can be subdivided in a number of ways. The draft ISO standard on determining solar irradiances (ISO-DIS-21348)[2] describes the following ranges:
    Name Abbreviation Wavelength range in nanometers
    Energy per photon
    Ultraviolet A, long wave, or black light
    UVA 400 nm – 315 nm 3.10 – 3.94 eV
    Near NUV 400 nm – 300 nm 3.10 – 4.13 eV
    Ultraviolet B or medium wave UVB 315 nm – 280 nm 3.94 – 4.43 eV
    Middle MUV 300 nm – 200 nm 4.13 – 6.20 eV
    Ultraviolet C, short wave, or germicidal
    UVC 280 nm – 100 nm 4.43 – 12.4 eV
    Far FUV 200 nm – 122 nm 6.20 – 10.2 eV
    Vacuum VUV 200 nm – 10 nm 6.20 – 124 eV
    Extreme EUV 121 nm – 10 nm 10.2 – 124 eV

  476. kim
    Posted Mar 3, 2008 at 4:31 AM | Permalink

    Do you mean ‘outside the area of the black and orange rectangles’ in your last sentence before background, Erl?

    The Sun if very sultry, and we must avoid its ultry-violet waves.

    H/t N. Coward.
    =============================

  477. Erl Happ
    Posted Mar 3, 2008 at 7:21 AM | Permalink

    Kim, 476
    Yes, black and orange. Dislexic translations.

  478. Erl Happ
    Posted Mar 3, 2008 at 7:26 AM | Permalink

    The large black arrows record slope and duration of the major La Nina events that occur shortly after geomagnetic and short wave radiation minima.

    Should read EL Nino events. Cringe.

    Looks like a spot the mistake exercise.

  479. Bill F
    Posted Mar 3, 2008 at 2:15 PM | Permalink

    468 Leif,

    If possible, can you give us a summary of the prediction panel telecon when it is over? Thanks!

  480. Posted Mar 3, 2008 at 6:08 PM | Permalink

    475 (Erl): Better to keep geomagnetic activity out of the equation. That both GA and FUV show a solar cycle variation does not mean that they are narrowly correlated [and certainly not on a time scale of months]. The correlation is of the same kind as that between reading ability and shoe size.

    479 (Bill): The detailed deliberations are not for the public :-) but the bottom line is clear: there has been little substantive change in the Sun; we are creeping towards the minimum and it is still too early to come down on one side or the other in the debate about the size of cycle 24.

  481. cba
    Posted Mar 3, 2008 at 6:57 PM | Permalink

    Erl,

    I went over some of your stuff this weekend. I’m having some problems making sense of the graphs and eyeballing what you’re referring to. I’ve definitely still have a bit of learning to do on the various details.

    I finally managed to get quarterly values of OLR from 74-07 into a spread sheet and processed such that I can view a single temperature for example. Since the data is monthly averages for Jan, Apr, Jul & Oct in raw form in the spreadsheet, I can manipulate it as desired – such as a global mean value or a SH / NH mean values or even a single 2.5 x 2.5 degree spot somewhere on the planet.

    I guess I need to next acquire the other factors like measured T, incoing solar irradiance, cloud cover, albedo, and the like. It’s quite interesting to see that the measured average OLR is 231.5 W/m^2 over the last 33 years compared to my clear sky standard 1976 atmosphere calculated value of 233.6 W/m^2 that I came up with in a 1 dimensional model. Comparing this with a nominal 1366W/m^2 (divide by 4 for average) solar insolation of about 341.5 W/m^2 which is reduced by about 31% shows about 235W/m^2 required emissions for complete radiative balance, ignoring chemistry and biology factors. The 1366 and albedo are rough values that need some refinement also to be more in line with the other values.

  482. Erl Happ
    Posted Mar 3, 2008 at 9:45 PM | Permalink

    481 (cba)
    Perhaps we can chat on google talk. I may be able to clarify what I am doing with the data. Queries encouraged.

    Its the variation in the input and the output that matters for the temperature of the body.

    I too need data on OLR for the tropics over time. It would be nice to demonstrate a symmmetry with changes in short wave radiation and temperatures in the tropics.

    Do the chemistry and biology factors really result in leakage of energy in the long term?

  483. Erl Happ
    Posted Mar 4, 2008 at 3:57 AM | Permalink

    480 (Leif)

    That both GA and FUV show a solar cycle variation does not mean that they are narrowly correlated [and certainly not on a time scale of months]. The correlation is of the same kind as that between reading ability and shoe size.

    I am interested in the timing of one in relation to the other. Have you any hard data on the timing of peaks in GA by comparison with FUV? From observation of the data produced at the Source meeting it seems that some of the GA fluctuations are in the troughs, some turning points slightly precede the FUV peaks and others closely follow.

    Agreed, the GA mechanism, if it exists, will be weak by comparison with the FUV effect. The nice thing about the FUV/UV effect is that it is instantaneous and shows up month by month. The cloud cover response is very difficult to assess but should be capable of being read in the data for outgoing long wave radiation.

  484. kim
    Posted Mar 4, 2008 at 6:39 AM | Permalink

    It’s interesting the manner in which consciousness can edge its way to understanding, here, in a way a grand data processing machine can’t. Perhaps if the machine could feel the sultry sun on its back?
    ===========================

  485. Posted Mar 4, 2008 at 7:08 AM | Permalink

    483 (Erl): Since you agree that the GA effect is weak, best to [at this stage] avoid it altogether so that you can concentrate on what you think is important.
    Here is a plot of 27-day averages of Kp [GA], Rz [FUV], and Dst [GA]:

    As you can see, there is very little detailed correlation, and no specific ‘lag’ between one and the other series. Sometimes [eg. 1980] GA is lower at solar max than at solar min [e.g. 1976, compared to 1980].

    You can make your own plots [or get hard data] at:

    http://omniweb.gsfc.nasa.gov/form/dx1.html

  486. cba
    Posted Mar 4, 2008 at 7:24 AM | Permalink

    Erl,

    I’m not familar with google talk? but the practical side of that is if it’s 2pm there is probably 2am here and vice versa. My guess is the written word would be superior and if not, would still be quicker than trying to setup the stuff from scratch and make it work.

    I still need to figure out how to set up a storage space for images that isn’t a mess like the one I have. It’s one of those freebie website locations that severly limits the ability – mostly to hook one into paying for minimally adaquate capabilities.

    Kim,

    Aside from the obvious problem that a computer program is limited to producing output based upon the preconceived notions of the programmer and tainted by imperfections by programmer, programming language designer/implementers and the constraints of the hardware, it has even more severe restrictions. Physics isn’t about just mathematics and applying some form of the scientific method. It’s driven by ultimately by the hunches, intuition, apparent correlations both real and imagined, and by the biases and expectations of those practicing it. And finally, it’s driven by sheer accident – the discovery of the unexpected. These are all things particularly unsuitable for computerization.

  487. Erl Happ
    Posted Mar 4, 2008 at 8:21 AM | Permalink

    483 (Leif)
    Thanks for the plot and the reference. It seems you have sunspots rather than FUV and two GA indices. But I can get the data using your guide. Thanks.

    I take your point in relation to the relevance of GA in driving atmospheric phenomena. The evidence for the influence of short wave radiation on the Earths albedo, given the coincidence of maxima in FUV and peaks in tropical heating as shown by the SOI (sign reversed) I find convincing. But, I am just a brief visitor here. You are in a position to be influential. Are you persuaded that short wave radiation is important in driving albedo and temperature response in the tropics?

    Or, do you see this connection as another case of a correlation similar in nature to the one between shoe size and reading performance.

    Robert Cahalan has an informative diagram showing the extent of interaction between short wave radiation and the troposphere in his presentation at the recent SORCE meeting.
    Richard Kleen at SORCE reported a dramatic drop in stratospheric aerosols that should add to solar forcing and short wave penetration of the troposphere.

    NASA has documented the humidification of the stratosphere over the last thirty years. That reduces ozone which, although pathetically small in concentration is supposed to eliminate some short wave radiation.

    Jay Mace presented a very interesting cross section of the tropical atmosphere showing a deficiency in zonal layer thickness of cloud between 30N and 30S latitude in the mid to upper troposphere. This conforms to the pattern that one might expect from the influence of short wave radiation.

    The swallows gather. Twill a summer make perhaps.

  488. Posted Mar 4, 2008 at 11:22 AM | Permalink

    487 (Erl): For your purposes there is no difference between sunspot number and FUV. You ask: “Are you persuaded that short wave radiation is important in driving albedo and temperature response in the tropics?”. No, I’m not, but I’m also not saying that it is not important. I don’t know, and I have not seen convincing arguments either way. I’m sure you can find many, many, many, many opinions, papers, posts, comments either way and there is no need to compile a list for me. Hopefully, some climate scientists are working on this [if not, there is no need for me to be interested] and at some point a resolution will be found.

  489. kim
    Posted Mar 4, 2008 at 11:29 AM | Permalink

    But sunspot number and FUV both have greater range than TSI.
    =====================================

  490. Posted Mar 4, 2008 at 11:42 AM | Permalink

    489 (kim):

    But sunspot number and FUV both have greater range than TSI.

    but their total ‘energy’ input is minuscule compared to TSI so it may not matter how much they vary.

  491. kim
    Posted Mar 4, 2008 at 1:06 PM | Permalink

    It only takes a little energy to move the valve than controls a great deal more flow.
    ==========================================

  492. kim
    Posted Mar 4, 2008 at 1:07 PM | Permalink

    ‘that controls’, not ‘than’.
    ==========

  493. Posted Mar 4, 2008 at 1:24 PM | Permalink

    491 (kim):

    It only takes a little energy to move the valve than controls a great deal more flow.

    this assumes that there is such a valve, and THAT is the question. With TSI you don’t need a valve. Whether there is a valve or what it is is unclear. I don’t know of any that all [or at least, most] agree upon.

  494. Erl Happ
    Posted Mar 4, 2008 at 3:51 PM | Permalink

    488 (Leif)

    Hopefully, some climate scientists are working on this [if not, there is no need for me to be interested] and at some point a resolution will be found.

    At some points in this blog you have mentioned friends who can be relied upon for an opinion. Might be time to ask them.

    The difficulty in sun/earth relations lies in the inability of specialists to bridge the gap. To do it in this case will require good data for short wave radiation over several cycles (just one paper so far trying to cobble together disparate series), good data on cloud cover (a field beset with measurement difficulties) or the extent of long wave radiation from the tropics (as an operative measure of the effect of cloud cover) and a willingness within the field of climate science to consider that ENSO phenomena might be global rather than local in their scope and might be the result of external forcing rather than internal oscillations of the climate system.

    I very much doubt that any work is being done. Currently the thinking is that greenhouse gas concentrations drive warming and that ENSO phenomena respond to this increasing concentration. This stack of cards will not yield without concerted effort by active and interested people.

  495. Posted Mar 4, 2008 at 4:50 PM | Permalink

    494 (Erl): you mention the lack of good data. Good scientists go where the data takes them, so without good data, one does not deviate too much [because one cannot] from the mainstream paradigm. It is not ‘unwillingness’ or bad will; just that “extraordinary claims require extraordinary evidence” and with without good data, it is hard the produce ‘extraordinary evidence’. I don’t think it is “inability” either. Most scientists consider themselves [yours truly included] as pretty able, even to bridge gaps to other fields, if there is a chance of making progress. What is often the real limiting factor is lack of funding for “outside of the box” work. But such is life.

  496. Erl Happ
    Posted Mar 5, 2008 at 3:27 AM | Permalink

    494 (Leif)
    Back at 420 you said in relation to 418 (JimP):

    what I see is that ONI and LOR are nicely correlated [as they should be - no surprise], and that none of them have any correlation with the cosmic ray intensity [which - to me - is no surprise either].

    What is your rationale as to why OLR follows the Oceanic Nino Index? If ENSO events are an internal oscillation of the climate system there should be no change in energy status…..or do you see the Earth as gaining or losing heat in the process? Is that not a bit heretical. Are you going where the data leads you and in the process have you strayed too far from orthodoxy?

    What causes the Earth to warm and at other times to cool?

    Is the change in OLR due to changing cloud status. If not, then what?

  497. kim
    Posted Mar 5, 2008 at 7:08 AM | Permalink

    Sure it’s clouds. What else has the range of effect?
    ============================

  498. Posted Mar 5, 2008 at 7:15 AM | Permalink

    496 (Erl):

    why OLR follows the Oceanic Nino Index

    Well, the warmer a body is, the more radiation it emits. El Nino is a warm situation, thus OLR should be higher. Now, where the Earth gets the surface heat from I don’t know, but the oceans would be a good guess. Heat could be stored at depth and well up to cause El Ninos. Pure speculation, of course, but if El Ninos are warmings then OLR must necessarily increase, so the correlation is just what I would expect. But, I’m not a specialist on this, so my expectation may not be worth much; suffice it to say that I, at least, was not surprised [maybe for the wrong reason - you tell me].

  499. cba
    Posted Mar 5, 2008 at 11:11 AM | Permalink

    Leif,

    I charted some NOAA satellite data for outgoing longwave emission. For fun I compared 1 month averages for Jan and Jul , 1974-2007. These values differ always by about 10W/m^2 for most of the graph – somewhat correlated – until about 2004when Jul dropped and Jan increased so that Jan is now 1-2 W/m^2 greater than Jul. At this time the equatorial band dropped significantly as well (equator +/- 1.25 deg).

  500. Posted Mar 5, 2008 at 1:05 PM | Permalink

    499 (cba): I don’t know what this all means. Why the cherry picking of bands? I was thinking globally. The details of the Earth’s radiation budget I’ll happily leave to others [and to other threads!]. My input is to try and determine what the Sun might be doing on short and [especially] long time scales.

  501. Erl Happ
    Posted Mar 5, 2008 at 8:04 PM | Permalink

    498 (Leif)
    At http://www.cpc.ncep.noaa.gov/data/indices/olr we have OLR for the equator 160E to 160W. I would prefer it to be for the entire tropics i.e. 30N to 30Slat. but it’s all I could find. However, I believe that the entire tropics move in the same direction at the same time in terms of temperature at least. Can’t think why OLR would be different.

    At http://www.bom.gov.au/climate/current/soihtm1.shtml there is data for the Southern Oscillation Index.

    Interpreting the SOI one must know that: quote From BOM:

    Sustained negative values of the SOI often indicate El Niño episodes. These negative values are usually accompanied by sustained warming of the central and eastern tropical Pacific Ocean, a decrease in the strength of the Pacific Trade Winds, and a reduction in rainfall over eastern and northern Australia.

    Hence, the sign of the SOI is the opposite of the Oceanic Nino index. Heating is shown by negative values of the SOI and positive values of the ONI.

    On page 9 of my paper at http://www.happs.com.au/downloaders/Cloud_temp_tropo.pdf
    figure 9 shows OLR and the SOI over the period 1996-2000.

    The relationship exhibited is exactly the reverse of what you described in #498. Warming of the ocean is accompanied by reduced OLR. If energy from solar radiation is going into heating the ocean it can not be emitted to space as OLR. A fall in the SOI (heating) is therefore accompanied by a negative anomaly in OLR and the two track each other nicely over the four year period.

    This contradicts the information provided by JimP in 418 but there you are.

    The inference is that EL Nino events occur when short wave radiation peaks warming the troposphere, reducing relative humidity and reducing cloud cover. That is what is indicated in my graph at http://i249.photobucket.com/albums/gg220/erlandlong/Radiationandcloud.jpg where the SOI is shown sign reversed for the convenience of those who like to see a line going up when temperature advances, therefore conforming to what they are used to with the ONI.

    In the Southern Hemisphere we obviously look at things differently! Little smile.

    More importantly, if the relationship between the SOI (or temperatures in the troposphere or across the tropics at the surface) and OLR is consistent we have a very strong link between the sun’s output of short wave radiation and terrestrial temperature. The Sun drives ENSO and that is a big part of long term temperature change. I say this because there is no doubt about the consistency of the relationship between temperatures in the troposphere (or the SOI as a proxy) and peaks in short wave radiation. The value of OLR is that it shows the effect of changing cloud cover on the amount of radiation getting through to the oceans. That is difficult to show by measuring cloud cover and transparency.

    So that we know what we are talking about here is the important relationship: Notice that the big variations are not at the surface but in the troposphere. There should be no doubt about what is driving what. If it doesn’t appear it is at http://i249.photobucket.com/albums/gg220/erlandlong/SWRandTroptemp.jpg

  502. Posted Mar 5, 2008 at 11:07 PM | Permalink

    501 (Erl):

    “If energy from solar radiation is going into heating the ocean it can not be emitted to space as OLR”

    I would think that if you heat something that it immediately begins to radiate the heat away, so somebody explain to me why not…

    I thought 1998 was a strong el Nino year and also a very warm year. The various switching of signs is lost on me. 416 looked clear to me. The ONI matching the temperature.

    About the Figure: I see no correlation or correspondence whatsoever between the Figure precisely to underscore the fact that there is none?

  503. Posted Mar 5, 2008 at 11:22 PM | Permalink

    502 (me): Last sentence should have read: “I see no correlation or correspondence whatsoever between the two time series. Did you post the Figure to underscore the fact that there is none? or what am I supposed to see [but can't]?”

  504. Erl Happ
    Posted Mar 5, 2008 at 11:51 PM | Permalink

    502 (Leif)

    I would think that if you heat something that it immediately begins to radiate the heat away, so somebody explain to me why not…

    Yes it does. But to the extent that the temperature rises in the body that is being heated, there is a storage phenomenon happening. Outgoing will then be less than incoming.

    If the water level in a tank is to rise, the input must exceed the output. Rain can fill the tank if the lid (the clouds) is removed. If the tank has the capacity to store water (no leaks) the level of the water will rise within the tank.

    Put the lid back on (clouds) and if there is leakage from the tank (outgoing long wave radiation) the water level (temperature) will fall no matter how hard it rains (incoming insolation).

  505. Erl Happ
    Posted Mar 6, 2008 at 12:11 AM | Permalink

    504 (Leif)

    what am I supposed to see

    1. Peaks in short wave radiation align with peak temperatures in the troposphere (UAH data).
    2. The troposphere exhibits much wider temperature variations than the surface (Hadley centre data).
    3. Deductively, the rise in temperatures in the troposphere at cloud level is not due to changes in temperature at the surface of the Earth but very likely changes in the quantum of short wave radiation from the sun.

    The suggested mechanism whereby short wave radiation changes cloud cover is via a fall in relative humidity at cloud level as air temperature increases and less condensation….hence less cloud.

  506. Posted Mar 6, 2008 at 6:40 AM | Permalink

    505 (Erl):
    1: peaks align? not that I can see. To really compare, compute the series to the same time resolution: 3 months. Plot the actual irradiance [not reduced to 1 AU] showing the 7% annual variation. Filter the data to remove variations longer than [say] two years [to remove solar cycle variation - why actually?]. Make scatter plot of result. Quantify the comparison.
    2: so what?
    3: the 150nm radiation does not penetrate to the cloud level, but is absorbed by ozone high in the Stratosphere.

  507. cba
    Posted Mar 6, 2008 at 7:51 AM | Permalink

    Leif, 500,

    What I was describing in #499 were overall global averages taken in two months, when the earth is at aphelion and perihelion and it was for the length of the dataset. After seeing the strange twist in the more recent data results, that is when I did a small equatorial band just to see if it provided a similar result – implying what was going on involved the equator region as opposed to not involving the equatorial region. Again, I was quite suprised at the 30 years worth of data which was strongly out of phase with the TSI at the TOA and was showing a significant and consistant difference of around 10W/m^2 between Jan and Jul for OLR for this entire duration which then shifted in 2004, despite the TSI being about 90W/m^2 higher in Jan than Jul due to perihelion. After all, I would have expected that the TSI would have a rather immediate effects on T which would then have an immediate effect on OLR.

    Rationalizing on this leads into differences of makeup in northern and southern hemisphere along with notions of what can cause delayed or reduced effects of TSI on T. The most obvious is the ocean/land percentage mix.

    In answer to your first question in #501, I think I know a reason – at least in theory. TSI is roughly 4x% visible, slightly more than that in IR and the balance in uV and further out wavelengths of decreasing significance. Also, from what I understand, variations of uV intensity are much greater than overall TSI variations. Each wavelenth is also subject to attenuation according to that wavelength. IR that isn’t absorbed in the atmosphere is going to be stopped at the ocean surface – right at the skin. Visible light will vary from red being absorbed at or near the surface while blue and green will penetrate many meters. The same goes for near or longwave uV that can penetrate many meters under the surface. There is going to be no radiative transfer, mediocre conduction, and fairly slow convection going on, at least relative to the atmosphere for that energy absorbed below the top. The top skin of the ocean is going to radiate its energy out and drop in T quite readily but that skin is either mm or sub mm in thickness and must wait for more energy to arrive from convection and conduction from below or yet more energy from above. Also, the ocean offers a tremendous heat sink ability.

    Hence, some incoming energy doesn’t result in immediate exhibition of noticeably higher T or have the ability to radiate immediately. While it is a tiny fraction of the overall, it is also the part which varies the most due to varying solar conditions. There are doubtlessly other possible examples of these ‘second tier’ scenarios. Just how much of a significant effect they can have remains to be seen. A mechanism that can effect daylight cloud cover by a % or so might be a ‘second tier’ level occurance but it would virtually have primary control over the system.

  508. Posted Mar 6, 2008 at 8:09 AM | Permalink

    507 (cba): all very good, but Erl is talking about 150nm radiation. First, I don’t think it even reaches the troposphere; second, I don’t think it would penetrate even sub mm into the oceans. But, again, all this has been hashed out in thousands of comments here and elsewhere, so we [I, at least] should not join that carousel. I thought the reason for the discussion was the alignment of peaks in various time series, and whether they go up or down [people conflating signs and all]. If there is no alignment then further discussion of the cause of the alignment seems rather pointless. So, on to alignment: I see none in the plots supplied. Do you? There are standard ways of resolving this issue. Let’s do it the way it should be done.

  509. Philip Mulholland
    Posted Mar 6, 2008 at 8:56 AM | Permalink

    Leif #472

    In the debate over solar influence, cosmic rays are sometimes ‘blamed’ for provided condensation nuclei. Here is a different twist:

    This New Scientist editorial continues the theme of bacteria and snowflakes:
    Editorial: Do bacteria control our weather?

  510. Erl Happ
    Posted Mar 6, 2008 at 9:06 AM | Permalink

    Leif
    Here is a diagram that clearly establishes the reduction in OLR across the Indian and Pacific oceans during the El Nino event of 1997-8. There is no need to concern ourselves with confusing indices with positive and negative signs. We know that the ocean warms during an El Nino event and here we see that OLR diminishes at the same time.

    The converse is the La Nina event when OLR increases and the ocean cools. Today’s map of sea surface temperature anomalies at the bottom shows the result in terms of ocean temperatures right across the tropics and also at high latitudes. That map is there so that we can easily judge the area affected by the changes in OLR.

    Now, the puzzle is: What causes the fluctuation in cloud that is plainly responsible for the change in the balance between reflectance of solar irradiation and ocean absorption of thermal energy?

    OK, 150nm radiation does not get to the surface. Lots of other wave lengths between 0 and 600 in the UV and Far UV spectrum do penetrate the troposphere. They all energise the atmosphere directly and some reach the surface. Wave lengths shorter than the visible have the capacity to energise the atmosphere directly. We can see from the chart of radiation at the 150-154 nm wave length that these wave lengths fluctuate very considerably over the short term, a time period much shorter than the solar cycle.

    I can see that peaks in UV irradiance correspond with the highest temperatures that are reached in the troposphere in the period of record. I don’t need a statistical test to see the point.

    If you need to employ that degree of sophistication to satisfy yourself that there is a relationship, and you have the skills (I don’t) then by all means do the responsible thing and compute it. I would love to get a copy of the short wave irradiance data in numerical form if it is available. I will send you a case of red wine in exchange.

    There are only two sources of warmth for the atmosphere, short wave radiation from the sun and the various processes by which the Earth imparts heat to the atmosphere. If the atmosphere varies in temperature to a much greater degree than the Earth itself over periods of time greater than a day or a week then it is UV radiation that is responsible. This follows from the thinness of the atmosphere, its incapacity to store heat and the rapidity with which its heat dissipation processes work.

  511. Posted Mar 6, 2008 at 9:26 AM | Permalink

    510 (Erl):

    I can see that peaks in UV irradiance correspond with the highest temperatures that are reached in the troposphere in the period of record. I don’t need a statistical test to see the point.

    But I do. I know there are many lurkers out there. Can we have a vote? Who else can see the correspondence? and, just as important, who cannot see it?. Now, why don’t the much larger solar cycle variation not cause a much larger temperature increase if every little blip do? So, small peaks have effect, large do not? Where is the dividing line? and why is there one?

  512. Posted Mar 6, 2008 at 9:32 AM | Permalink

    510 (Erl): does MgII index suffice? see: http://wwwsolar.nrl.navy.mil/susim_uars_mgii_index.html for explanation and http://ozone.sesda.com/solar/ for data. MgII actually have a better change of reaching deep into the atmosphere as its wavelength is 279.9 nm.

  513. Andrew
    Posted Mar 6, 2008 at 10:31 AM | Permalink

    Erl is trying too hard, I think. Erl, you need to remove the effects of El Nino and volcanoes (random, I think) first and other effects before the solar cycle emerges from the noise. Here’s an excellent example:

    http://www.spacecenter.dk/publications/scientific-report-series/Scient_No._3.pdf/view
    :)

  514. Posted Mar 6, 2008 at 12:22 PM | Permalink

    513 (Andrew): I think Erl is not trying to get a solar cycle signal [and certainly not by removing ENSO], but rather to have some [other] solar parameter [I don't quite know; it has been a moving target - FUV, cosmic rays, aa-index, solar wind, ...] be the CAUSE of the ENSO [el Nino, el Nina, etc] [and all consequences of that]. But it is not clear to me from [the otherwise extremely detailed and voluminous] postings what is to be explained and how.

  515. Raven
    Posted Mar 6, 2008 at 4:43 PM | Permalink

    Erl,

    I see a lot of overlap between the angle you are persuing and Linzden and Spenser’s work on the Infrared Iris.

    http://www.weatherquestions.com/Roy-Spencer-on-global-warming.htm

  516. Erl Happ
    Posted Mar 6, 2008 at 5:39 PM | Permalink

    Raven,
    Thanks for the link. You are dead right in your comment. ENSO is climate change and the period is not 11 years. Spencer and Lindzen have the common sense necessary to keep the various forces in perspective.

    Quote from Roy Spencer:

    Al Gore likes to say that mankind puts 70 million tons of carbon dioxide into the atmosphere every day. What he probably doesn’t know is that mother nature puts 24,000 times that amount of our main greenhouse gas — water vapor — into the atmosphere every day, and removes about the same amount every day. While this does not ‘prove’ that global warming is not manmade, it shows that weather systems have by far the greatest control over the Earth’s greenhouse effect, which is dominated by water vapor and clouds.

    And I would add this. Temperatures in the troposphere vary widely across time, much more so than at the surface. With temperature varies relative humidity and cloud cover. And the variation in cloud cover is the key to the warming and cooling of the oceans where the Earth stores warmth.

    What causes the atmosphere to warm and cool? It is short wave radiation from the sun that shows a very strong variation within the solar cycle. How? Because the wave length is short enough to directly energise the atmosphere. There is enough at the surface to cause sunburn, give you headaches and cause melanomas. Hold up you hand to the sun and look at the inside of your fingers. UV radiation changes the colour of paint, denatures plastics and at very short wave lengths, so I am told, penetrates concrete.

    Now, in terms of an index to measure short wave radiation it probably does not matter much which wave length is chosen because they all vary together, the shortest waves varying the most. But Leif could probably tell us about this.

  517. Posted Mar 6, 2008 at 5:51 PM | Permalink

    516 (Erl):

    Now, in terms of an index to measure short wave radiation it probably does not matter much which wave length is chosen because they all vary together, the shortest waves varying the most. But Leif could probably tell us about this.

    The shorter the wave length, the larger the variation, but also the less penetrates into the atmosphere. The MgII-index is probably a good compromise, so make a three-month running average and compare with your temperature graph [don't forget to put the 7% annual variation back in].

  518. Erl Happ
    Posted Mar 6, 2008 at 6:09 PM | Permalink

    511 (Leif)

    Now, why don’t the much larger solar cycle variation not cause a much larger temperature increase if every little blip do? So, small peaks have effect, large do not? Where is the dividing line? and why is there one?

    Not just a good question but a great one. The swift changes to create small peaks are more important because the effect on relative humidity and cloud cover requires it. Gradual change will not do the trick. The atmosphere is always playing catch up. Extra warmth at the surface increases sea surface temperatures, promotes the circulation of the air, causes evaporation, increases relative humidity and promotes cloud cover.

    Water vapour is the Earths refrigerant gas, it limits the temperature gain in the tropics, carries heat into the upper atmosphere where it can be readily radiated to space (CO2 is irrelevant where the density is low enough and the temperature cold enough) and in the form of clouds is our sun shield effectively diverting about 30% of solar radiation so that it never reaches the surface.

    When short wave radiation collapses so does temperature in the troposphere and up goes cloud cover. OLR increases in response. The oceans cool. From what I can see in terms of OLR effects most of the change in cloud cover that is clearly associated with ENSO effects is between Madagascar and 170° West.

  519. Posted Mar 6, 2008 at 7:08 PM | Permalink

    518 (Erl): OK, so here is the MgII-index [a proxy for the UV that penetrates into the troposphere and sunburns you]. The data is plotted as monthly means. The red curve is the MgII-index at the Sun. The blue curve is the MgII-index at the Earth [which is presumably what we should use]:

    Now, which temperature curve [with one month resolution] do you wish me to compare the blue curve with? Tell me, and I’ll add it to the plot.

  520. Posted Mar 6, 2008 at 7:16 PM | Permalink

    519 (me): beats me why image showed nicely in the preview, but not in the comment. Trying again:

  521. Posted Mar 6, 2008 at 7:22 PM | Permalink

    529 (me): small confession to make: the MgII-index as an index is actually independent of the distance to the Sun and as such does not vary over the year, but if we use MgII as a proxy for the intensity of UV, then that intensity must be modulated by the distance squared. This is for the nit-pickers only.

  522. Erl Happ
    Posted Mar 6, 2008 at 7:33 PM | Permalink

    512 Leif

    does MgII index suffice?

    The following two figures seem material to this question. My question to you is which wave lengths heat the atmosphere below the tropopause and of those, which ones vary sufficiently to promote significant change in temperature?

    Is that second figure, with all its symmetry, the result of measurement or calculation? Just how representative is it of the facts? Its static for a start, not dynamic, and it needs to be dynamic.

    I can see significant penetration of the troposphere in the tropics by very short wave lengths in the top diagram that does not seem to be acknowledged in the bottom diagram.

  523. Posted Mar 6, 2008 at 7:51 PM | Permalink

    522 (Erl): The top one shows that there is NO penetration or heating for shorter than 300 nm. The bottom one is calculation, but from well-understood physics, so not in doubt [although 30 years old]. Also note that 100 mb is at 16 km height, so above the troposphere [even in Tropics]. So, the high-energy short-wave length stuff don’t get into the troposphere.

  524. Erl Happ
    Posted Mar 6, 2008 at 7:57 PM | Permalink

    519. Leif

    Now, which temperature curve

    Roy Spencers UAH Alabama temperatures in the troposphere.

    I sense a straw man here. UV at the Earth is very different between the hemispheres and can not be ascertained from any site at the surface. Big seasonal variation. What you need is a measurement from the active region in the upper atmosphere, de-seasonalised. I presume that the index at the sun is the more appropriate of the two but for the effect that I am looking at you need a better resolution.

    May well be that the average is inappropriate and you should be looking at the average maxima on a 10 day basis. The beauty of the 150-154nm composite from spacecraft is that it gave you all the information, trends in both minima and maxima. It’s the big sudden peaks that produce the result that is significant.

    If you want to apply a statistical test you need to be looking at area above the curve with a defined slope of increase to take into account the time factor.

    Averages will suck the guts out of the thing.

    Do you know that the average of a daily maximum and minimum temperature can lie either side of an average of 24 hourly temperature measurements by as much as two degrees, and either way. And yet we base all our estimates of Earthly temperature on just two measurements, the minimum and the maximum.

  525. Erl Happ
    Posted Mar 6, 2008 at 8:12 PM | Permalink

    523 (Leif)
    Troposphere runs up to 20km in the tropics.

    Looks to me like significant penetration of the upper atmosphere at the 200 to 230nm level and no data below 200 nm. The 250nm should be a reasonable proxy for this wave length.

    Looks to me also that UVB has maximum penetration and 280-315 is where we should be looking for the effect of this much less variable wave length.

  526. Posted Mar 6, 2008 at 8:40 PM | Permalink

    524 (Erl): Let’s take these one at a time:

    Roy Spencers UAH Alabama temperatures in the troposphere.

    You could have saved me the trouble of digging up the URL, but, OK, Google can.

    I sense a straw man here.

    This is a serious accusation actually. Would you put up a straw man? so, how can you say that I did? I’ll assume you are serious and honest, so you may assume the same of me.

    UV at the Earth is very different between the hemispheres

    We are looking at the same Sun. Maybe you mean on the ground? But you use UV data measured by satellite, so clearly you didn’t mean on the ground.

    and can not be ascertained from any site at the surface.

    Here we agree, because the FUV does not penetrate to the surface or to within 16 km of it.

    Big seasonal variation. What you need is a measurement from the active region in the upper atmosphere, de-seasonalised.

    This does not make sense, as the seasons regulate the radiation budget at any given site.

    I presume that the index at the sun is the more appropriate of the two

    This is wrong, I think, because we need to take into account what is actually received at the Earth. Imagine, the Earth had a very eccentric orbit with aphelion being ten times as distant as perihelion, so that we receive 100 times less energy at aphelion. Should we still use the value at the Sun and ignore the 100 times change in actually received radiation?

    but for the effect that I am looking at you need a better resolution. May well be that the average is inappropriate and you should be looking at the average maxima on a 10 day basis. The beauty of the 150-154nm composite from spacecraft is that it gave you all the information, trends in both minima and maxima. It’s the big sudden peaks that produce the result that is significant.

    Apart from the fact that 150 nm doesn’t penetrate to the troposphere, we now have a specific research project: examine the high-resolution data and try to identify and align peaks. I’ll do that, but since the temperature data you showed are 3-month averages, it becomes a real problem to identify one-day or five-day or even 10-day variations with a three-month smoothed curve. How would one do that? There could be many short bursts within a three-month interval, so alignment might either be meaningless or trivially true [because you can always find a local maximum among many within a wide window]. We shall see how it comes out.

    If you want to apply a statistical test you need to be looking at area above the curve with a defined slope of increase to take into account the time factor.

    I don’t understand this. Maybe you mean that one sets a slope in advance as to what constitutes a rise. What slope did you use or would you suggest?

    Averages will suck the guts out of the thing.

    They will remove the subjective guesswork alright. If that also removes the effect, then what?

    Do you know that the average of a daily maximum and minimum temperature can lie either side of an average of 24 hourly temperature measurements by as much as two degrees, and either way. And yet we base all our estimates of Earthly temperature on just two measurements, the minimum and the maximum.

    Where I live, we can have a 50F temperature change in an hour, but in the long run this averages out and things are not so screwy as to keep me awake at night. Maybe you are preparing for an attack [an audit - to stay with the CA-spirit] on the temperature curve as not being good data, should the analysis turn out not to support the expectations…

  527. Posted Mar 6, 2008 at 8:49 PM | Permalink

    525 (Erl): From the always useful Wikipedia:

    The troposphere is the lowest of the Earth’s atmospheric layers and is the layer in which all “weather” occurs. It begins at ground level and ranges in height from an average of 8 km (5 miles/26,400 feet) at the poles to 17 km (11 miles/58,080 feet) at the equator[1]. At the equator, the stratosphere begins at roughly 17 km (11 miles) in altitude, and it may reach as high as 50 km (31 miles) from the earth’s surface. It is at its highest level over the equator and the lowest over the geographical north pole and south pole. On account of this, the coolest layer in the atmosphere lies at about 17 km over the equator. Due to the variation in starting height, the tropopause extremes are referred to as the equatorial tropopause and the polar tropopause.

  528. Posted Mar 6, 2008 at 10:14 PM | Permalink

    526 (me,Erl):

    Erl: but for the effect that I am looking at you need a better resolution. May well be that the average is inappropriate and you should be looking at the average maxima on a 10 day basis. The beauty of the 150-154nm composite from spacecraft is that it gave you all the information, trends in both minima and maxima. It’s the big sudden peaks that produce the result that is significant.

    Me: Apart from the fact that 150 nm doesn’t penetrate to the troposphere, we now have a specific research project: examine the high-resolution data and try to identify and align peaks. I’ll do that, but since the temperature data you showed are 3-month averages, it becomes a real problem to identify one-day or five-day or even 10-day variations with a three-month smoothed curve. How would one do that? There could be many short bursts within a three-month interval, so alignment might either be meaningless or trivially true [because you can always find a local maximum among many within a wide window]. We shall see how it comes out.

    Erl: Averages will suck the guts out of the thing.

    Now, we are proceeding with our research project: The issue is the time resolution. I suggested a month would be good, but Erl clings to a shorter time [presumably because the curve didn't look what he expected - this is a typical, normal, and reasonable reaction]. We have from the SORCE satellite an MgII-index [at the sun] with a time resolution of ~3 hours. Here is the year 2003 [starting March 1st, and not yet modulated by distance]:

    The vertical lines are 27 days apart. Hmmm, There is a strong solar rotational modulation [not unexpected]. This means that within any three-month average window of temperature as shown in comment #501, there will always be three strong peaks to align with whatever temperature peak we choose to consider. This makes the alignment procedure somewhat meaningless, so clearly we’ll have to average the solar data to remove the ever-present rotational signal. Since 27 days is close to a month, monthly averages [as we had in the Figure in comment #520] should do just fine. So, now we can progress to the next phase [tomorrow] comparing monthly averages. As for averaging sucking the guts out of the effect we are looking for [actually looking for something - anything - might be somewhat unscientific unless we have a definite mechanism and are looking for confirmation of that one] so be it. We cannot avoid the averaging if one of the series already is an average.

  529. Posted Mar 6, 2008 at 10:21 PM | Permalink

    Here is the whole SORCE series. Note, the solar cycle signal, the rotational signal, and the occasional activity burst signal:

  530. Erl Happ
    Posted Mar 7, 2008 at 7:12 AM | Permalink

    Leif,
    Sorry for my inattention. Vintage has started this week so there is a lot on. 6.30 AM to 9.30 PM today with a bit of time in the morning to jot a post.

    Yes, three month averages should be fine. Clouds can disappear for a month in the middle of a La Nina cooling event. I documented that in the paper at http://www.happs.com.au/downloaders/Cloud_temp_tropo.pdf

    However, practically speaking, La Nina and El Nino events can last for up to three years so if there is an effect, as I am sure there is, there should be a parallel increase in the short wave data.

    I am really gratified that you are going to the trouble of analysing the data.

    I remember Dennis Wingo saying that the US air force had done a lot of work measuring short wave length radiation at the top of the atmosphere. Ideally, the data should be coming from the place where the heating effect is postulated. Neither the surface or outer space is appropriate. That said, if peak radiation as measured by a satellite in low Earth orbit lines up with peak temperatures in the troposphere it would be reasonable to suggest that there is an effect, would you not agree?

  531. Posted Mar 7, 2008 at 1:57 PM | Permalink

    530 (Erl):

    if peak radiation as measured by a satellite in low Earth orbit lines up with peak temperatures in the troposphere it would be reasonable to suggest that there is an effect, would you not agree?

    Absolutely, yes. But, on the other hand, if they do not line up, then you would have to agree that there is no effect, right?

    So, let’s see how we fare.
    First, we plot MgII [at Sun, and at Earth] together with the temperature anomaly for the Tropics according to Roy Spencer:

    It is clear that there is no obvious solar cycle nor annual cycle in dT to correspond with the cycles in MgII, so we follow Erl’s theory that it is only the short-term spikes that have effect, and remove a 1-year running mean from both time series:

    The time scale is not suitable for comparing the curves, so we expand it a bit; here are the first ten years:

    There are some peaks that line up [green dots] and some that do not [red dots]. Here are the next ten years:

    and the next:

    As many peaks do align and equally many do not, how does one know if this is just random or if there is a pattern or system to it? The standard way is to correlate the two series:

    The uniform ‘cloud’ of points around zero shows that there is no correlation and that the alignments are just random noise. Now, perhaps the ‘effect’ is a bit delayed, so we can also correlate MgII with dT one months [and two, and three, etc] later. This is shown by the red dots [for 1 month], and, again, there is no correlation. At this point it is time to follow the data and say that they do not corroborate the claimed effect. Most proponents of a theory don’t give up yet, but suggest trying a different latitude band, a different solar index, different smoothing, different time interval, different this and different that, in the hope that something will show up. This is a normal [if somewhat deplorable] human trait, but the increasingly cramped antics will soon doom the theory to be ignored [for good reason] by mainstream science.

  532. Jim Arndt
    Posted Mar 7, 2008 at 3:01 PM | Permalink

    Hi,

    Leif, did you try and compare it the the geomagnetic index? Temperature that is. Also I think Joe D’Aleo did TSI to PDO and got some very interesting results using an 11 year smooth.

  533. Larkin Lowrey
    Posted Mar 7, 2008 at 4:22 PM | Permalink

    I’m not nearly qualified enough to participate in this thread… but I’m going to embarrass myself anyway.

    When I look at Leif’s top MgII chart in post 531, I see a “disturbance” in the dT immediately after the minimums (1988, 1998). Granted, there is also a dT “disturbance” in 1983, but that observation intrigues me. My first reaction, as an engineer, is that this might indicate resonance in a (loosely) coupled system. I am wondering how small periodic variations in solar output might resonate strongly wrt the ENSO, and other oceanic oscillations.

  534. Posted Mar 7, 2008 at 11:00 PM | Permalink

    532 (Jim A):

    Leif, did you try and compare it the the geomagnetic index? Temperature that is.

    Well, here we go with “try this, try that” :-)
    Keeping the same format as above we get for the aa-index and dT:

    I’ll come to the same conclusion: “no effect”.

    Also I think Joe D’Aleo did TSI to PDO and got some very interesting results using an 11 year smooth.

    Both PDO and TSI have decadal variations and so might just go along for a while without being related.

    533 (Larry):

    this might indicate resonance in a (loosely) coupled system

    Assuming that they are coupled at all, and THAT is the issue. We have no good evidence that they are [although lots of people claim so - please don't serve us with yet another list].

  535. Posted Mar 7, 2008 at 11:23 PM | Permalink

    534 (me): The ‘cloud’ on the correlation plot is a bit off-center because the aa-index has a long tail of very large values, pulling up the mean, but no tail at the low end [it can't go negative]. This does not alter anything. There are also standard techniques to transform this ‘log-normal’ distribution to a ‘normal’ distribution, but, again, they don’t make any difference.

  536. Posted Mar 7, 2008 at 11:42 PM | Permalink

    Just to show what it looks like when there is a correlation, we plot MgII against the sunspot number [SSN]. Using the same format as above we get:

    If we hadn’t removed the running 1-yr mean, but simply correlated the monthly values we would have found the very high ‘coefficient of determination’ R^2 = 0.92.

  537. maksimovich
    Posted Mar 8, 2008 at 1:31 AM | Permalink

    re 535 LS

    The ‘cloud’ on the correlation plot is a bit off-center because the aa-index has a long tail of very large values

    The AA index statistcs is biased by large events

  538. Erl Happ
    Posted Mar 8, 2008 at 4:57 AM | Permalink

    Leif
    Thanks for your effort in analysing the data. However, there are a few things that you need to consider:

    1. You must have an old copy of Wikipedia! What I read is: “The average depth of the troposphere is about 11 km (7 miles) in the middle latitudes. It is deeper in the tropical regions (up to 20 km (12 miles)) and shallower near the poles (about 7 km (4 miles) in summer”.
    2. From the diagram that I posted at 522 there is an effect on the upper atmosphere by radiation at 200 nm down to about 300 mb pressure or about 8.3 km. We have no data for shorter wave lengths in that diagram. Can wave lengths shorter than 200 nm be detected in the upper troposphere? If the figure is to be believed radiation at 200nm is detectable in the upper troposphere. Any short wave radiation will have a heating effect. I am not persuaded by your proposition that the calculations of 30 years ago are correct because you say they are ‘based on well established physical principles’. Many argue the greenhouse effect of CO2 on the same basis. Others argued that the Earth was flat and that the sun revolves around the Earth using the same appeal to authority.
    3. My proposition is conditional and I would not expect the sort of correlation that you could demonstrate with standard statistical method. I do not expect a correlation between every data point on both series and neither should you. There is a rate of increase in the energy supplied by short wave radiation to the upper atmosphere that will produce a fall in relative humidity and cloud cover as long as the increase in evaporation brought on by the resulting increase in sea surface temperature (and increasing wind speed) does not keep pace. If short wave radiation ascends in such a manner (over a short enough interval with has sufficient heating power) points after such a sufficiently rapid advance will correlate with temperatures in the troposphere, and at the surface and there will also be a relationship with outgoing long wave radiation. Other points will not. Those that do correlate are probably best discovered by examination of the result for different trend rates of increase. The type of presentation used for the 150-154 nm data is well suited to this. The validity of the effect is easily established by visual inspection. I know of no statistical test that will do the job. The phenomena are complex and multifactorial. You sell me badly short. The statistical test you use is fine for two variables unaffected by any third factor impinging on the dependent variable. However, it lacks the sophistication to test this proposition. You throw out the baby with the bath water.
    4. Complex phenomena demand deductive thinking. I hope that one thing that I have clearly established is that during El Nino events outgoing radiation falls as energy is swallowed up by the heat sink that is the ocean. If you don’t like the notion that the increase in sea surface temperature is due to falling cloud cover due to heating of the atmosphere by short wave radiation, go ahead and suggest a better one. Up-welling of warm waters is not on. If the point is contestable then go ahead and say so.
    5. What explanation do you have for the much greater rate of temperature fluctuation in the troposphere than at the surface? How can this happen. If the surface is the sole supply of warmth for the troposphere this defies the laws of physics.
    6. What explanation do you have for the fall in temperatures in the tropics at sunspot minimum, a fall that is plainly unrelated to the gradual fall in TSI? What explanation do you have for the big El Nino that occurs reliably at the commencement of each solar cycle when short wave radiation and geomagnetic activity are rapidly increasing? What other aspect of solar activity could be responsible?

    But there is an even more fundamental question to consider. Can ENSO phenomena that includes a periods of warming and cooling across the entire tropics be properly considered as due to an ‘internal oscillation of the climate system when variations in OLR are plainly involved. Obviously, such events affect the Earths heat budget. In trying to discover why the Earth has warmed can we afford to ignore these short term changes in the heat budget? The long term is made up of lots of short terms. I have demonstrated that we have swung between La Nina dominance and El Nino dominance several times in the last 150 years. The last three solar cycles were heavily El Nino dominant and the three previous, that coincided with cooling in the northern hemisphere were La Nina dominant.

    So, I remain unconvinced by your statistical test. Nor am I intimidated by bullying that concluded your last post. ‘Most proponents….etc’.

    You are not really entering into the argument at all.

    Far easier to criticise than to chance one’s hand.

  539. Sean Houlihane
    Posted Mar 8, 2008 at 5:38 AM | Permalink

    Erl,

    It seems to me that Leif is being extremely tolerant in his efforts to help you discover that your theory is simply not supported by the evidence. What you are doing here seems similar to the attempts to correlate tree ring widths to temperature. You feel there must be a connection, which you have identified. If you want to progress your theory, I suggest it is up to you to locate data which you expect to show statistical support for your theory before the analysis is done, and accept that the absence of any correlation is equivalent to proof that your theory is flawed. In that case, move on and look for a different mechanism. I believe there is something, but you have not found it yet.

  540. Posted Mar 8, 2008 at 6:31 AM | Permalink

    537 (maks…): The ‘knee’ in your curve is an artifact of the definition of aa. [logarithmic up to aa = ~70, then increasing more slowly]. I’m not sure why you made your comment. To refute something or to support something…

    538 (Erl): I did not expect you to change your mind [although I did have, at least, the hope]. I have put in a large effort in trying to see what you do. Up to this point I have clearly failed to see the light by applying the tools that I’m familiar with, but I have labored to show you why I, at least, am not convinced. Maybe your theory cannot be approached by standard scientific methods.

  541. Erl Happ
    Posted Mar 8, 2008 at 7:23 AM | Permalink

    540 (Leif) and 530 (Sean)
    Correlation is never proof of causation and the opposite should also be acknowledged. Where several interacting factors are involved the lack of correlation is explicable.

    Here we have a trend in temperature in the troposphere that is unrelated to surface phenomena. Demonstrably, the troposphere can warm while the surface cools and there is a simultaneous drop in measured cloud cover and a fall in outgoing long wave radiation. There is no other source of heat than the sun. It follows that some emanation from the sun is responsible. The only thing that seems possible is short wave radiation that has the capacity to heat the atmosphere directly is responsible for this temperature increase.

    I have a lot more faith in deduction than statistical analysis. Had there been a high correlation we would have been none the wiser.

    What solar studies lack most is a plausible description of modes of causation for the variations in temperature that are plainly evident. I go where my observations lead me.

  542. Erl Happ
    Posted Mar 8, 2008 at 7:46 AM | Permalink

    Perhaps we can run a correlation analysis on the UAH data for the troposphere (tropics) and the Hadley surface temperature data. When we discover it is less than perfect we can ask ourselves why? How much discrepancy is allowable before we question whether there is a source of heat for the troposphere other than the Earth itself.

    Correlation analysis is an aid but no substitute for analysis of cause and effect.

    While we are at it, can we solve the question of why the stratosphere cools as the surface warms via statistical analysis.

    End of the day you need a good reason why two things are related. Sometimes we even discover that apparently unrelated phenomena are closely dependent. Cooling of the stratosphere related to increasing moisture, less ozone, less aerosol content, less interception of short wave radiation.

  543. Posted Mar 8, 2008 at 7:55 AM | Permalink

    540, 538 (Erl): A more constructive approach might be that “some of the peaks” line up and only those that do are of importance. Then one could study if there are other factors that only happen during the peaks that line up [let's called then the 'golden' peaks]. This would be a perfectly valid approach, but then the golden peaks will have to be accepted a priori, not by cherry picking from a graph without justifying each pick. There is a similar situation with geomagnetic storms. After a century long struggle, people finally accepted that geomagnetic storms were caused by explosions on the Sun hurtling material our way. But lining up the explosions with the storms was not so simple, because about half of the explosions did not cause the expected storm: some did, and some did not. We now know that only if the matter ejected from the Sun has a southward pointing magnetic field embedded in it does a large storm ensue. That is the missing ingredient. This can be elucidated by studying the individual events, not by saying that only visual inspection and not statistical methods work. So, to make progress you should make a list of golden events [and another of the duds], then see what is so special about each list. The onus is completely on you for this, but it is a way to go forward. Ad hom-arguments [bullying, babies, etc] are bad style. Frustration should be channeled into constructive work. The fun is in the journey, after all, not in the arrival.

  544. Posted Mar 8, 2008 at 8:39 AM | Permalink

    541,542 (Erl): about ‘correlation': almost everything is correlation in the end. That is the experimental aspect of science. You have a theory, it makes predictions, you compare the prediction with observations. You do that by correlating the predicted values with the observed values. Correlations can also guide you to the theory. Your own comment #501 is a prime example of a correlation [lining up peaks] that was supposed to support a theory. All I did was to quantify that correlation so that it was not just eyeballing.
    “I have a lot more faith in deduction” and “I go where my observations lead me” seem to be slightly contradictory. “There is no other source of heat than the sun. It follows that some emanation from the sun is responsible” ascribes all climate and weather variation to the Sun which is way too extreme and factually wrong.
    Erl, as someone said, you are trying too hard.

  545. Erl Happ
    Posted Mar 8, 2008 at 9:32 AM | Permalink

    Leif

    543: I take your point. And thank you for the illustration. It is very apt. And I agree The fun is in the journey.

    544 You say I am factually wrong. My comment on heat sources related to the troposphere alone. It should not be taken as a statement applying to climate in general or climate change in general. Climate is a site specific thing. What other sources of heat are there for the troposphere than the Earth and the sun? Have I missed something?

    I don’t mind being called a trier but ‘trying too hard’ is derogatory and best left unsaid.

  546. Posted Mar 8, 2008 at 9:45 AM | Permalink

    541 (Erl):

    The only thing that seems possible is that short wave radiation that has the capacity to heat the atmosphere directly is responsible for this temperature increase.

    The short wave radiation is less than 1% of the total and the VAST majority of the short wave radiation is absorbed way up in the atmosphere. The part penetrating to the troposphere to “heat the atmosphere directly” is truly minuscule and on energy grounds alone cannot be responsible for the temperature increase. Apart from the lack of observational evidence [correlation] I don’t see how your ‘direct heating’ by UV could work. So, neither observation nor [simple] theory seem to work for me here.

  547. Posted Mar 8, 2008 at 9:52 AM | Permalink

    545 (Erl): you are factually wrong is assuming that only emanation from the Sun can be responsible for the temperature variations on the time scale we were discussing. Do I really have to belabor this?

    “Trying too hard” to explain everything in the whole chain. I believe in lowering the aim a bit when the going gets rough… And I am NEVER derogatory in intent.

  548. Erl Happ
    Posted Mar 8, 2008 at 9:59 AM | Permalink

    Occurs to me that the simplest way to ascertain whether short wave radiation penetrates the troposphere or not is to simply get up there and look for it, now, and during the coming El Nino when SC 24 cranks up. What is the capital expense on a short wave radiation detector?

  549. Erl Happ
    Posted Mar 8, 2008 at 10:06 AM | Permalink

    Leif

    Do I really have to belabor this?

    Yes you do. What time scale do you have in mind?

    “Explaining the whole chain”. Every link is an important part. The mechanism that explains the little shifts may well be the same mechanism involved in the big ones, on the centenneial scale at least.

  550. Posted Mar 8, 2008 at 10:15 AM | Permalink

    549 (Erl): time scale of from 1 month to 1 year. You yourself used 3-month averages thus defining the time scale. Since you made the claim, you set the time scale. Except that you may be a moving target mentioning centennial time scales in #549, after having told us that on a time scale of a solar cycle there is no effect, only “the sharp increases count”. So, you would like me to belabor why on a time scale of 3 months ONLY the sun has any effect at all. Let me mention Pinatubo to begin with.

  551. Posted Mar 8, 2008 at 10:27 AM | Permalink

    548 (Erl):

    Occurs to me that the simplest way to ascertain whether short wave radiation penetrates the troposphere or not is to simply get up there and look for it

    People have been doing that already, and the observations match the calculated values closely:

    JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 110, D09202, doi:10.1029/2004JD005397, 2005

    Airborne measurements of aerosol, ozone, and solar ultraviolet irradiance in the troposphere

    Costas Varotsos

    Abstract
    [1] Measurements of the distributions of the aerosol characteristics and solar ultraviolet irradiance were conducted by using instrumentation flown on a Falcon aircraft over the entire Greek area from the sea up to the middle troposphere. To study the impact of the aerosol abundance to the solar ultraviolet irradiance at various altitudes, complementary observations of ozone concentration, relative humidity, and temperature have been also performed using instrumentation flown on free balloons. In addition to a detailed description of the observed aerosol characteristics we attempt an improvement of the earlier proposed theoretical algorithms regarding the absolute values of the solar ultraviolet irradiance at various tropospheric altitudes. [...] Moreover, upon using the obtained aerosol distribution data in the radiative transfer calculations, we find that the calculated values of the solar ultraviolet irradiance correlate well with the observed ones, showing an 4.3 ± 0.1% km−1 increase for altitudes ranging from the ground to 6.2 km.

  552. kim
    Posted Mar 8, 2008 at 10:50 AM | Permalink

    Uh, oh, hippogryphs ahoy. He connects them to blooms of fish populations. And climate.
    ===========================

  553. Posted Mar 8, 2008 at 10:57 AM | Permalink

    554 (kim): “and climate”… but I thought Erl had just taught us that only the Sun can do that. So, I guess poor Gerhard loses…

  554. Posted Mar 8, 2008 at 11:06 AM | Permalink

    551,553-558: Folks, we are drifting OT. Enough.

  555. cba
    Posted Mar 8, 2008 at 4:40 PM | Permalink

    Erl, %541

    Your comment on shortwave uv being the only thing that can heat the atmosphere is in error. There is longwave solar incoming that has some absorption inbound and there is longwave outbound. While the inbound is unlikely to vary much compared to uV incoming, it’s absorption is likely to depend on water vapor content (as well as clouds) so may offer some serious variability in this realm and may well have some association with the ENSO activity. Also, the outbound IR does have some absorption activity all the way out, even though the vast majority is near the surface. It too is affected most seriously by h2o vapor and by clouds which also tend to affect just how the surface T varies.

    While I haven’t yet found some really good incoming measurements before SORCE and I haven’t found good albedo measurements prior to that either, I have just done some simple balance calculations based upon the SORCE and NOAA albedo averages of the last few years and made a 33 yr energy balance for aphelion and perihelion months. It’d be nice to find some.

    It is a bit interesting so far. If I recall it correctly, we have a difference in albedo from jan to jul of 0.29x to 0.31x. Incoming energy is about 90W/m^2 greater in jan due to perihelion. Outgoing energy is quite a few W/m^2 lower in Jan than in Jul (I think). Also, I believe the result is that the net budget is about 10-15W/m^2 absorption in Jan and an average of around 3 W/m^2 excess emission in Jul – at least up to around 2004 when excess absorption in Jan dropped to under 5W/m^2 and excess emissions in Jul dropped to almost 0.

    Note, I do need to double check the data to make sure there’s not some shift due to offsets to a wrong month but if not, evidently something changed in 2004, either in the data collection or in the data. If this is real, it could turn out to be rather interesting. It’s also rather apparent that I’m going to have to keep downloading additional months to determine all the monthly averages because there is quite a variety in the budget during the year.

  556. Jim Arndt
    Posted Mar 8, 2008 at 5:24 PM | Permalink

    Hi,

    Here is graph for solar cycle length and temperature

  557. Pat Keating
    Posted Mar 8, 2008 at 5:27 PM | Permalink

    557 cba
    Interesting work again, cba.
    Is the lower flux since 2004 due to albedo?

  558. Jim Arndt
    Posted Mar 8, 2008 at 5:29 PM | Permalink

    Hi,

    Here is a paper for CRF and Cloud cover.

    http://star.arm.ac.uk/~ambn/345epb.pdf

  559. Posted Mar 8, 2008 at 8:44 PM | Permalink

    I have been working with Roy Spencer’s temperature data as per Erl’s suggestion. Now, looking at the data more closely, I notice that there does not seem to be any annual variation in the NH and SH data [tropics, I can understand]. Why is that? What are these data anyway? I assumed they were deviations from some reference level, but does that include a ‘reference’ annual variation that is subtracted out?
    Anybody knows?

  560. Erl Happ
    Posted Mar 8, 2008 at 9:21 PM | Permalink

    (cba)

    Your work on OLR is fundamental. Can you treat the tropics as a unit?

    You say:

    Your comment on shortwave radiation being the only thing that can heat the atmosphere is in error.

    You are correct and my statement needs copious qualification. It should have been followed with an “all other factors being equal” clause. Outgoing long wave is attenuated as part of the atmospheric heating process. The diagram at 522 shows lots of incoming wave lengths attenuated in the atmosphere and as you point out the degree of attenuation will depend upon moisture vapour, cloud, ozone and aerosols, the latter two especially in the stratosphere where moisture levels are low enough to allow them to persist for a long time. So, the degree of atmospheric heating also depends upon water vapour content, clouds, volcanoes and other forms of pollution.

    I am looking for sources of atmospheric heating that are not dependent upon such factors, important as they are. So, I suggest that the same wave lengths that split the oxygen molecule in the stratosphere and create the ionosphere, (some ions existing in a free state down into the troposphere I have read somewhere) can be responsible for the same effects in the troposphere. These wave lengths are short enough to energise atmospheric gas, nitrogen and oxygen, directly, and can also disassociate the water molecule I believe. Recombination is a function of atmospheric density so the life of an ozone molecule or a free ion in the troposphere is short. But the impact of the process is energy gain.

    Cooling of the stratosphere is prima facie evidence of a reduction in the interception of solar radiation in the 20- 50 km zone above the troposphere. Cooling has been continuous over the period of the satellite record (except for short term blips readily linked to volcanoes) and is much greater in amplitude than the degree of warming at the surface. What happens to that portion of radiation that is not absorbed in the stratosphere? It doesn’t turn around and go back again.

    So, my hypothesis, based on the observed heating of the troposphere and disappearance of clouds in the tropics, (while there is simultaneous cooling at the surface) is that those wave lengths that are short enough to energise atmospheric gas are responsible. It is atmospheric gas that is the base constituent and only the energising of this stable base constituent can generate the sort of fluctuations one sees in temperatures across the tropics. The other constituents of the atmosphere vary independently. No one suggests that they are tied to ENSO.

    This mechanism then involves the sun in a temperature fluctuation that is tied to variations in short wave energy emission. Those variations are very much in evidence in my post at 501. and at: http://i249.photobucket.com/albums/gg220/erlandlong/Radiationandcloud.jpg

    The mechanism has a short term component that is related to ENSO and a long term component related to changes in the stratosphere brought about by changes in the troposphere due in turn to the change in the balance between heating and cooling over time (ENSO).

    Is it really a co-incidence that those three big El Ninos mark the start of the new solar cycle? Now, there is a testable proposition! Nothing like the start of a new solar cycle for a big upswing in short wave radiation.

  561. Posted Mar 8, 2008 at 10:43 PM | Permalink

    561 (Erl):

    So, I suggest that the same wave lengths that split the oxygen molecule in the stratosphere can be responsible for the same effects in the troposphere.

    Erl, those wave lengths are effectively removed [down by many orders of magnitude] from the incoming beam by Oxygen (below 242 nm) or by Ozone (above 242 nm) before they reach the troposphere. This process is not governed by cooling or such, but only depends on how many molecules the incoming photons encounter, which is well known.

  562. Erl Happ
    Posted Mar 8, 2008 at 11:18 PM | Permalink

    Leif,
    For a discussion of indices and their peculiarities see here: http://www.junkscience.com/MSU_Temps/Warming_Look.html

    And here

    http://www.junkscience.com/MSU_Temps/Sources.html

    Can not raise the direct link for the UAH explanation of their data.

  563. Posted Mar 8, 2008 at 11:23 PM | Permalink

    562 (me,Erl): Maybe I can show that best with this Figure:

    Take, for instance, a best case of a wave length of 205 nm. At a height of 20 km, the surviving flux is only 1/100 of what it was at 50 km height. So the energy left is only 1/100 of what it was at 50 km, which is already only a tiny, tiny fraction of the solar irradiance. There are simply not enough surviving photons to do any heating below 20 km.

  564. Erl Happ
    Posted Mar 8, 2008 at 11:31 PM | Permalink

    562 (Leif)

    My point is that if the molecules of ozone and the particulate matter were there to intercept the radiation the temperature would rise, not fall.

    This talk from the recent SORCE conference is pertinent to the question of stratospheric warming and cooling. In either event there are implications for the temperature for the troposphere and the height of the tropopause and the temperature there

    :

    Climate Forcing Since 1960: What Does the Moon Have to Say?
    Richard A. Keen [richard.keen@colorado.edu], University of Colorado, Dept. of
    Atmospheric and Oceanic Sciences, Boulder.
    About once per year, on average, the moon is eclipsed as it passes into the Earth’s
    shadow; at these times the moon can be used as a remote sensor of the global average optical
    depth of stratospheric aerosols of volcanic origin. The moon is visible during a total lunar
    eclipse due to sunlight refracted into the shadow (umbra) by the Earth’s atmosphere
    (primarily by the stratosphere). Stratospheric aerosols can affect the observed brightness of
    the eclipsed moon, and the optical thickness of the aerosol layer can be determined from the
    difference between observed and predicted brightness of a lunar eclipse.
    Eclipse data from 1960 to 2007 and from 1880 to 1888 suggest that the Pinatubo
    eruption in 1991 perturbed the stratospheric aerosol layer at least as much as that of Krakatau
    in 1883. The aerosol veil from Pinatubo disappeared between 1993 and 1996. Since 1997,
    stratospheric aerosols have been at or near background levels; this is the longest period with
    a clear stratosphere since before 1960.

    The converse is stratospheric warming, due to greater short wave flux or the presence of more molecules to absorb the flux that is coming through.

  565. Posted Mar 8, 2008 at 11:31 PM | Permalink

    563 (Erl): Thanks for the links. So, the numbers are deviations from MONTHLY averages. This explains why there is no annual variation. But, of course, raises other questions [which perhaps can be answered by study of the stuff] such as the fact that the monthly averages themselves change with time, so maybe there is a reference period or something. I’ll assume that the folks at UAH have done a reasonable job, so I don’t have to worry about such details.

  566. Erl Happ
    Posted Mar 8, 2008 at 11:33 PM | Permalink

    Sorry, last sentence should not be in quote.

  567. Posted Mar 8, 2008 at 11:36 PM | Permalink

    565 (Erl): at 205 nm [and all wavelengths below 242 nm] the absorption is done by Oxygen of which there is plenty at all times [about 21%]. No cooling or warming, etc, is relevant [as I already pointed out]. Who cares about aerosols etc when good ole O2 does the job perfectly well.

  568. DeWitt Payne
    Posted Mar 8, 2008 at 11:37 PM | Permalink

    Leif #560,

    …but does that include a ‘reference’ annual variation that is subtracted out?

    Yes.

  569. Posted Mar 8, 2008 at 11:38 PM | Permalink

    567 (Erl): it doesn’t matter because the O2 concentration is constant [21%] anyway.

  570. Posted Mar 8, 2008 at 11:41 PM | Permalink

    569 (DeWitt): so this is not controversial, so I don’t need to worry about it, but can assume that EVERY climate researcher in the whole world would not have a problem with the Spencer series? Or is this the usual can of worms?

  571. DeWitt Payne
    Posted Mar 8, 2008 at 11:46 PM | Permalink

    Leif (571),

    It’s SOP for all the temperature anomaly records, GISS, Hadley, RSS. It allows a comparison between geographic locations (SH vs NH say) at any time of the year.

  572. Posted Mar 8, 2008 at 11:48 PM | Permalink

    572: explain the jargon, SOP?

  573. Erl Happ
    Posted Mar 8, 2008 at 11:59 PM | Permalink

    564 (Leif)

    Is that a peak I see between 190 and 215nm? Looks like plenty of photons above 300 nm too. Can we not expect energy delivery to the atmosphere between 300 and 600 nm?

    Can we consider that less photons may still have an effect when there are but one tenth to one sixteenth of the atmospheric molecules present to be worked on?

    Can we also consider that if this diagram relates to average conditions there are large variations about the average with big reductions in atmospheric density in the middle of the day?

    Can we expect any acceleration effects from the reduction in cloud cover?

  574. John Baltutis
    Posted Mar 9, 2008 at 12:08 AM | Permalink

    Re: #573

    572: explain the jargon, SOP?

    Standard Operating Procedure.

  575. maksimovich
    Posted Mar 9, 2008 at 12:11 AM | Permalink

    Can we expect any acceleration effects from the reduction in cloud cover?

  576. Erl Happ
    Posted Mar 9, 2008 at 12:48 AM | Permalink

    567 (Maksimovich)

    Where does this data come from? And since you are such a handy fellow can you also obtain anomalies for outgoing long wave radiation at the top of the troposphere for the same latitudes?

  577. Posted Mar 9, 2008 at 5:21 AM | Permalink

    575 (John,DeWitt): SOP; does everyone adhere to the same standard?

  578. Erl Happ
    Posted Mar 9, 2008 at 6:20 AM | Permalink

    576 (Maksimovich)
    Seems to be a direct relation between your cloud cover index and the SOI. Increasing cloud occurs as the index rises towards La Nina cooling events. Fall in cloud cover is associated with a fall in the index towards El Nino events. Within, there is a large seasonal variation associated with falling cloud cover in Northern Hemisphere summer. This perturbs any possible connection with the UV index over short time scales. However, from the figure, there appears to be an inverse relationship with magnetic flux when seen over the period of an entire solar cycle.

    A relationship also appears in the data for outgoing long wave radiation with radiation increasing with cloud cover in the move towards La Nina cooling. For the relationship see #510.

  579. Posted Mar 9, 2008 at 7:05 AM | Permalink

    574 (Erl):

    Is that a peak I see between 190 and 215nm?

    yes, that is the best we can do. Reduction only by a factor of 100. Outside that peak, nothing gets through below 300 nm.

    Looks like plenty of photons above 300 nm too. Can we not expect energy delivery to the atmosphere between 300 and 600 nm?

    yes, that is called sunlight and sure does heat the surface with little heating of the air [which is transparent to light]

    Can we consider that less photons may still have an effect when there are but one tenth to one sixteenth of the atmospheric molecules present to be worked on?

    worked on by 1/100th, and the total energy absorbed is thus puny.

    Can we also consider that if this diagram relates to average conditions there are large variations about the average with big reductions in atmospheric density in the middle of the day?

    The density doesn’t matter [and there is no big reduction]. What matters is the number of molecules and that stays the same.

    It would seem to me that your mechanism have no way to work.

  580. Posted Mar 9, 2008 at 7:16 AM | Permalink

    Erl, this OLR data might be some use:

    http://www.cdc.noaa.gov/cdc/data.interp_OLR.html

    The data is a global grid at 2.5 x 2.5 deg resolution starting Jun 1974 through to Dec 2007 in either daily or monthly chunks, so you can extract data for whatever areas you like from it.

    You will need some software that can read NetCDF format files – for example, NASA GISS has free downloads of PANOPLY that makes neat maps from the data:

    http://www.giss.nasa.gov/tools/panoply/

    – such as this OLR map for Dec 2007:

    – which shows how effective La Nina is at dumping IR from the NINO regions of the Pacific into space.

    A program such as “R” may have a NetCDF reader package that could make extracting exactly what you want in the form you want it easier.

  581. cba
    Posted Mar 9, 2008 at 9:05 AM | Permalink

    Pat,

    I have to say simply that I don’t know at present. My guessis that it probably is as it’s not due to TSI, but…

    Also, there’s a lot more effort I need to do as this is just looking at a single month (aphelion) and a single month perihelion and there may be other curiosities with time lags.
    Spring break is right around the corner and I may get a bit more time to play but I’ve still got to keep ahead of the game as I am extremely rusty on stuff since I spent a whole career away from this, unlike Leif with his dual one. Consequently, my efficiency (lack there of) is overcome only with significant extra effort.

  582. Erl Happ
    Posted Mar 9, 2008 at 9:56 AM | Permalink

    580 (Leif)
    How then do you interpret the diagram below from:

    Cloud Occurrence, Cloud Overlap, and Cloud Microphysics from the First Year of CloudSat and CALIPSO Data
    Jay Mace With contributions from: Roger Marchand, Mark Vaughn, Qiuqing Zhang

    http://lasp.colorado.edu/sorce/news/2008ScienceMeeting/

    Looks to me like there is a fairly regular flux of heat to the tropical troposphere that reduces cloud presence above low latitudes. This is the place where evaporation is greatest where one might expect lots of cloud up to the highest altitudes. Apparently, there are some real gaps there down to quite low levels.

    How much evidence do you need to shake your faith in the insularity of the troposphere. There is just 20% (?) of the mass of the atmosphere above it and that at a very tenuous density as you have been at pains to point out to me on more than one occasion. There is also the possibility that upper atmosphere neutrals are subject to some displacement by electromagnetic influences and a definite influence on density in the stratosphere due to changing flux of short wave radiation.

  583. Posted Mar 9, 2008 at 11:38 AM | Permalink

    583 (Erl): The reason there are no clouds in the upper tropical troposphere is not that the clouds have evaporated away due to heat, but that it is too cold for clouds to form [they freeze out and fall down]. The tropical upper troposphere is the coldest and driest place on Earth. This is good, because if it were not so, water vapor would rise into the stratosphere, get disassociated into Oxygen and Hydrogen; the latter then escaping into space, leaving the Earth as dry as Mars in the end.

    The stratosphere is indeed only 24% of the atmosphere, but that is plenty enough for the Oxygen [and Ozone] to remove all the UV below 300 nm, preventing it from getting further down.

    As to how the neutrals are influenced by ‘electromagnetic’ forces, you tell me.

    I don’t have any ‘faith’ to be shaken in any of this. This is all cold, hard calculation and observation. I think I have shown here that the correlations you were advocating [although you say you don't believe in correlations] don’t support your ideas and that the physics you invoke don’t support your ideas either. But none of this detract from the value your ideas may have to you. Now, all this is complicated enough [and not my field] that I may be wrong here and there. Let more knowledgeable people correct me as they come across my errors. I’m always willing to listen and to learn [but it has to make sense and your stuff just doesn't].

  584. Posted Mar 9, 2008 at 1:40 PM | Permalink

    583 (Erl): If you want to learn more about how Solar UV could influence the Troposphere, you might study: Terry Nathan,On the Connection Between Solar Spectral Irradiance, Planetary Wave Drag and the Zonal-Mean Circulation also from the SORCE 2008 meeting. Here, at least, the physics looks better.

  585. pochas
    Posted Mar 9, 2008 at 1:53 PM | Permalink

    This may be of interest to those searching for a solar variability/climate link.

    http://www.awi.de/en/news/press_releases/detail/item/stratosphaerische_ozonchemie_als_wichtiger_faktor_fuer_atmosphaerische_stroemungsmuster_identifizier/

  586. DeWitt Payne
    Posted Mar 9, 2008 at 2:23 PM | Permalink

    Leif #578,575

    (John,DeWitt): SOP; does everyone adhere to the same standard?

    AFAIK (As Far As I Know). Removing seasonal variation from anomaly plots seems to be very common. Cryosphere Today, for example, does it for their sea ice anomaly plots. A signicant difference between anomaly plots is the baseline period. However that is usually just a constant offset.

  587. Posted Mar 9, 2008 at 5:04 PM | Permalink

    588 (DeWitt): Removing a ‘standard’ [unvarying] reference seasonal variation seems a dangerous thing to do. In my own field, there are quantities that vary with the seasons, and where the seasonal variation itself vary from year to year. Subtracting an unvarying reference variation would seem to introduce a signal where there is none. A good example is the ‘quiet day Sq variation’. This is a change through each local day of a geomagnetic element, e.g. the declination. This diurnal change is brought about by a electric current vortex in the E-layer at 110 km height. The vortex stays fixed in relation to the Sun, while the Earth rotates underneath it, thus creating an apparent change through the of the magnetic effect of the current at an observer rotating with the Earth [e.g. an observatory]. The strength of the current depends on the ionization of the ionosphere which in turn depends on the solar zenith angle [varying through the year] and the solar FUV flux [varying over the solar cycle] and on some other [not understood] effects than can vary the current strength by up to a factor of two from day to day. Subtracting a reference seasonal variation will introduce a seasonal variation in years with low solar activity and not remove enough in years with high solar activity. But, if this is not controversial in climate research, I’ll not try to make it so, but shall adopt the SOP.

  588. Posted Mar 9, 2008 at 5:49 PM | Permalink

    From the Grapevine:

    Geophysical Research Abstracts,
    Vol. 10, EGU2008-A-06330, 2008
    SRef-ID: 1607-7962/gra/EGU2008-A-06330
    EGU General Assembly 2008

    Is the correlation between solar proxies and clouds affected by ENSO and volcanic events?
    M. Voiculescu, I. Usoskin, K. Mursula
    Studies of correlation between the cloud cover and solar proxies suggest that variations of solar activity can affect the cloud cover at Earth. Atmospheric processes are important when studying the extent of the link between the solar activity and global climate change and it was suggested that climatic or terrestrial quasi-periodic and sporadic phenomena, such as ENSO and/or major volcanic eruptions, which do affect the cloud formation, may influence the results of statistical studies of the Sun-cloud relation. Using partial and total correlation analysis, we show that removing ENSO and volcanic years from the full-set analysis does not alter the results. Moreover, some relations, as for instance the anticorrelation between low clouds and UV irradiance, are improved. This supports the idea that the solar signal affects clouds sirectly. An interesting result relates to an area in the eastern Pacific, where the full-set analysis showed that the relationship between cosmic ray induced ionization and clouds is opposite to the global one. This odd correlation is no longer observed when the “problematic” years are removed. We conclude that, although removing years of strong ENSO and volcanic eruptions has no important effect on global correlation patterns, caution must be paid when interpreting the results of correlation studies in some areas, prone to be affected by extreme internal climate processes.

  589. Posted Mar 9, 2008 at 8:52 PM | Permalink

    In the discussion of the heating by short wave solar radiation it is good to know what the integrated solar irradiance from 0 to 200 nm is: 0.10 W/m^2. It is instructive to compare that with the heat flow out of the interior of the Earth which is 0.09 W/m^2…

  590. EW
    Posted Mar 10, 2008 at 4:20 AM | Permalink

    If someone is interested in reading a related article to that mentioned by Leif, I can send a .pdf (ewcz@seznam.cz):

    Mirela Voiculescu,Ilya Usoskin, and Kalevi Mursula
    Effect of ENSO and volcanic events on the Sun–cloud link
    Advances in Space Research
    Volume 40, Issue 7, 2007, Pages 1140-1145

    Results of correlation studies between solar proxies and clouds suggest that there is a solar effect on the occurrence of clouds. However, there is a possibility that terrestrial quasi-periodic and sporadic phenomena, such as ENSO and/or major volcanic eruptions, which have an effect on the cloud formation, may influence the results of statistical studies of the Sun–cloud relation. We show that removing ENSO and volcanic years from the full-set analysis does not alter the results. Moreover, the correlation between clouds of different type and two solar proxies, UV irradiance and cosmic ray induced ionisation, is partly improved. This supports the idea that the solar signal affects clouds directly. An interesting result relates to an area in the eastern Pacific where the full-set analysis showed that the relationship between clouds and cosmic ray induced ionization is opposite to the global one. When ENSO and volcanic years are removed this odd correlation disappears, suggesting that in this particular area, the ENSO effect prevails over solar effects.

  591. Erl Happ
    Posted Mar 10, 2008 at 7:39 AM | Permalink

    Leif
    Re wave lengths greater than 300nm and your statement that:

    yes, that is called sunlight and sure does heat the surface with little heating of the air [which is transparent to light]

    and also your statement in #591

    In the discussion of the heating by short wave solar radiation it is good to know what the integrated solar irradiance from 0 to 200 nm is: 0.10 W/m^2. It is instructive to compare that with the heat flow out of the interior of the Earth which is 0.09 W/m^2…

    From Wikipedia:

    Non-ionizing radiation (or, esp. in British English, non-ionising radiation) refers to any type of electromagnetic radiation that does not carry enough energy per quantum to ionize atoms or molecules — that is, to completely remove an electron from an atom or molecule.[1] Instead of producing charged ions when passing through matter, the electromagnetic radiation has sufficient energy only for excitation, the movement of an electron to a higher energy state.

    The non visible (to humans) spectrum from 300nm to 400nm encompassing Middle ultraviolet 200-300nm, UVB 280-315nm, NUV 300-400nm, UVA 314-400 nm are therefore active in influencing atmospheric temperature. It is considered important that sunscreen block both UVA and UVB indicating the prevalence of these two at the surface. Their concentration at the top of the troposphere is considered a considerable hazard to flight crew in commercial airliners.

    Diagram at #565 shows all above 280nm fully active down to the surface.

    Flux in these wave lengths is accompanied by strong flux in all other wave lengths, many of which are capable of exciting (heating) atmospheric molecules as conveyed by the diagram in #522. It is therefore incorrect to assume that the heating of the atmosphere during an episode of increased UV flux will amount to that provided by the energy of wave lengths below 200nm alone (your post #591). Why 200 nm?. Is this because of the strong activity at the ionising wave lengths between 195 and 215 nm (see diagram in #565)

    The UV effect on cloud cover via atmospheric heating is alive and well. And furthermore it is likely to amount to as much as the variation in TSI over the entire cycle. Given that water vapour content of the air increases exponentially with temperature there are major knock on effects on irradiance at the surface. This is all the amplification that is needed to justify a very strong solar effect on climate. What’s more there seems to be a strong relationship between cloud cover and ENSO phenomena. Arguably, the predominance of La Nina events over the last three solar cycles is responsible for a good deal of the warming since 1976.

    What sort of correlation would we get in Maksimovich’s diagram at #577 on less than a solar cycle time interval. Would a low correlation invalidate the effect? What is this effect due to? Is this the effect referred to in the abstract at #590? Is it not the same effect that I have been talking about for a couple of weeks only to be told that the mechanism does not exist?

  592. kim
    Posted Mar 10, 2008 at 7:47 AM | Permalink

    You mean El Nino events in your next to last paragraph.
    ======================

  593. Posted Mar 10, 2008 at 8:23 AM | Permalink

    593 (Erl): the reason I have focused on the short wave lengths and not the near UV to which the atmosphere is transparent [therefore no heating] is simply that you repeatedly invoke them as the causative agent:

    522 (Erl):

    I can see significant penetration of the troposphere in the tropics by very short wave lengths in the top diagram that does not seem to be acknowledged in the bottom diagram.

    562 (Erl):

    So, I suggest that the same wave lengths that split the oxygen molecule in the stratosphere and create the ionosphere, (some ions existing in a free state down into the troposphere I have read somewhere) can be responsible for the same effects in the troposphere. These wave lengths are short enough to energise atmospheric gas, nitrogen and oxygen, directly, and can also disassociate the water molecule I believe. Recombination is a function of atmospheric density so the life of an ozone molecule or a free ion in the troposphere is short. But the impact of the process is energy gain.

    Also, you bring up the variability of the 150-154 nm [one of your diagrams], you see a peak at 195-215 nm, etc. My discussion has had the goal of pruning all that irrelevant very short wave length stuff away. It seems that you are now saying that the near UV (300-400 nm) to which the atmosphere is transparent [to wit: it sunburns you] is the main agent in heating the troposphere. And that the influence of solar activity in this band is important. If so, we have made progress and can continue the pruning.

  594. Posted Mar 10, 2008 at 8:50 AM | Permalink

    595 (me): to reduce nit-picking, I note that when I said “is the main agent in heating the troposphere”, I, obviously, was implying “causing El Nino events”.

  595. Erl Happ
    Posted Mar 10, 2008 at 9:03 AM | Permalink

    594 (Kim)
    Dead right. Always have trouble with that.

    596 (Leif)
    Right too. Plead guilty to sloppy definition. Must be more careful.

  596. Posted Mar 10, 2008 at 9:34 AM | Permalink

    597 (Erl): it is more than just the definition; the whole fundamental assumption ["suggestion"] is off. This is important, because no progress can be based on a wrong basis. But maybe it doesn’t matter, because it didn’t make sense before, so the sooner it is abandoned the better is the chance that you can get on a better track. And, I don’t think that that particular assumption was central to your work, anyway. I think that your key point is that ENSO is externally imposed rather than internally. This may be so [I have no idea and no opinion], but it devalues your key point to be hostage to something clearly wrong [FUV heating the troposphere, or Geomagnetic Activity somehow doing it].

  597. Jim Arndt
    Posted Mar 10, 2008 at 10:46 AM | Permalink

    Hi,

    Leif, Hi please check comment #558 (didn’t post very timely some sort of issue). It shows solar cycle length and temperature. Also Comment #560 CRF and clouds. Still trying to see if fits it all together.

  598. Posted Mar 10, 2008 at 11:04 AM | Permalink

    599 (Jim A): the #558 correlation has been discredited [rightly so in my opinion, see here]. The CRF and Clouds is still something many believe [not me, though, but I'm open to be convinced].

  599. Jim Arndt
    Posted Mar 10, 2008 at 11:42 AM | Permalink

    Hi,

    Leif #600 The problem I have with that is in Fig #2 they, Mann and Jones, reconstructs of Temperature. The CRF is interesting in that it maybe effected by the Earth’s magnetic field. So you may not have a homogeneous change in CRF and maybe concentrated by field strength. I have been looking at some magnetic field maps and temperature maps and will update you that, need time to do all of this.

  600. Posted Mar 10, 2008 at 12:06 PM | Permalink

    601 (Jim A): the problem with the CRF is that it depends on BOTH the Earth’s magnetic field and AND the Sun’s magnetic field. The Earth’s field reconstruction is in pretty good hands, but the Sun’s is not. Then, the CRF is gotten from 14C or 10Be data, and those in turn may be contaminated by climate or volcanoes, plus that the temperature reconstruction is also doubtful [to wit: this whole blog], so lots of unknowns.

  601. Jim Arndt
    Posted Mar 10, 2008 at 1:45 PM | Permalink

    Hi,

    Leif, Thank you. Reconstructions are doubtful at times when using isotopes for proxies. A lot of proxies and don’t want a proxy a proxy.

  602. See - owe to Rich
    Posted Mar 10, 2008 at 2:01 PM | Permalink

    Leif,

    Re #558 and its “discrediting”, the discrediting analysis does not apply to my own work
    here, since that uses a 1+1+1 filter consistently, and is based on the HadCRUT3 series. My work does, admittedly have a problem with Cycle 24, but the way Cycle 24 temperatures are going my model may end up on track again!

    I shall be posting a graph or two there quite soon. And I haven’t forgotten my promise to you to see if I can do anything with sunspot numbers rather than cycle lengths.

    Regards,
    Rich.

  603. See - owe to Rich
    Posted Mar 10, 2008 at 3:30 PM | Permalink

    Sorry, “problem with Cycle 24″ should have read Cycle 23.

  604. Posted Mar 10, 2008 at 4:03 PM | Permalink

    605 (Rich): but you can still do some ‘what-if’ research. What if cycle 24 were to be long as well [as I think] or short [as Dikpati thinks]?

  605. cba
    Posted Mar 10, 2008 at 4:37 PM | Permalink

    Leif,

    http://www.climateaudit.org/index.php?p=166 takes one to the thread analyzing the demon and laut 2004 paper and there are some references to other papers as well from there. It would seem there is a lot of questions concerning just how adaquately demon and laut debunked the the cloud / cr / magnetic connection.

  606. Posted Mar 10, 2008 at 6:25 PM | Permalink

    607 (cba): Is there not some confusion as to what is being debunked here? I thought it was solar cycle length => temperature and not crf => clouds?
    My own opinion is that the whole premise is silly. F-CH&L justify using the length because it is a proxy for the size of the cycle, but why use a proxy when the real thing is available? And the smoothing over 30+ years reduces the number of degrees of freedom so much that almost all significance is lost. Consider two 100-year time series with 100 yearly values each. Then smooth them so that only two values remain in each, and presto, perfect correlation. This is, of course, extreme, but, I hope, illustrates my point. But, hey, if you think the procedure has merit, I’m not going to lose sleep over that.

  607. Andrew
    Posted Mar 10, 2008 at 7:29 PM | Permalink

    My understanding was that Gleissberg found short cycles had high sunspot numbers and strong eruptive activity while long cycles were characterized by low maxima and fewer solar eruptions. But maybe his findings have really “held”? Laut also purported to show that clouds diverged from cosmic rays, but in fact this was the result of a calibration problem that Svenmark corrected for. Laut blindly declared jihad on all alleged correlations of Solar with climate. I’m not sure if I trust him, but I don’t regard the solar cycle length thing highly anyway, given how long this cycle has been.

  608. Posted Mar 10, 2008 at 8:35 PM | Permalink

    609 (Andrew): Many people have remarked that strong cycles are short and that weak cycles are long, on average. This is not a strong rule, but only a tendency.

    ——–

    Scafetta & West have an ‘opinion’ piece in the March 2008 issue of ‘Physics Today’, and end with this: “If climate is as sensitive to solar changes as we suggest [their 2007 JGR, 112 D24S03 paper], the current anthropogenic contribution to global warming is significantly overestimated. We estimate that the Sun could account for as much as 69% of the increase in Earth’s average temperature, depending on the TSI reconstructions used [and hence indirectly on the sunspot number time series]. Furthermore, if the Sun does cool off, as some solar forecasts predict will happen over the next few decades, that cooling could stabilize Earth’s climate and avoid the catastrophic consequences predicted in the IPCC report.”

    This is well trodden ground, but interesting that it should appear in Physics Today.

  609. DeWitt Payne
    Posted Mar 10, 2008 at 10:26 PM | Permalink

    Leif #589,

    Subtracting an unvarying reference variation would seem to introduce a signal where there is none.

    Interesting. That may well explain something on another thread about correlations between temperature anomalies and CO2 anomalies.

  610. Raven
    Posted Mar 10, 2008 at 10:45 PM | Permalink

    I have seen previous discussions of 33 month vs. 66 month intervals between cycles. I never realized that the former group covers all of the moderen cycles (see http://users.telenet.be/j.janssens/Spotless/Spotless.html#Main ).

    Dr. Svalgaard noted before that pre-modern observers probably missed small sunspots which would introduce a bias into the sunspot cycle max and the number of spotless days. I was wondering if this same observation error could explain the 66 month intervals (i.e. the first spotless day for the purposes of this calculation actually had spots). This would mean the current 50 month delay is taking us into uncharted territory.

  611. maksimovich
    Posted Mar 10, 2008 at 11:31 PM | Permalink

    Re Erl(many posts).

    An interesting conundrum.If in El Nino years the zones of deficit radiative flux were perturbed in some way,would not the surplus zone also be effected and by an effect and not a causal mechanism.

  612. Pete
    Posted Mar 11, 2008 at 3:05 AM | Permalink

    #610 – The paper Leif refers to is here:

    S&W

  613. Pete
    Posted Mar 11, 2008 at 3:09 AM | Permalink

    Also, for those who don’t have a sub to PhysicsToday.org, Scafetta has it on his web site:

    March Opinion

  614. cba
    Posted Mar 11, 2008 at 6:59 AM | Permalink

    Leif,

    I thought it curious that length vs count would be used. What’s more, SS length could potentially be an additional factor.

  615. Posted Mar 11, 2008 at 7:14 AM | Permalink

    612 (Raven): Although I have shown it before, here is an updated version [actually, the previous one is updated too - as it is pulled from the updated version on my website] of the accumulated days since the 10th and the 1st spotless day:

    Since the first spotless day can occur almost at random after solar max I prefer the 10th spotless day as my marker. You will notice the very large spread using the 1st and the much tighter curves used the 10th. Anyway, the point is that the curves do not fall into two distinct groups, 33 vs. 66 months. They wander all over the place. The end of cycle 23 [green curve] leading to cycle 24 falls right within all the other ones, so we are not in uncharted territory. The red curves are the cycles with high Rmax, the blue are for low Rmax.

  616. Andrew
    Posted Mar 11, 2008 at 7:19 AM | Permalink

    Any particular reason why you emphasized cycles 14 and 19?

  617. Erl Happ
    Posted Mar 11, 2008 at 7:20 AM | Permalink

    613 (Maksimovich) Afraid I am all at sea with this graph.

    My opinion: The zone of intense radiation in the tropics warms the ocean variably depending upon cloud cover. There is a lag in the distribution of warmth to higher latitudes depending on ocean transport and over shorter time scales due to air movement, this depending upon the relative strength of the ITCZ in relation to other convective forces at higher latitudes. We have a lot of warm moist air streaming away from the tropics at the surface currently. This affects precipitation patterns. Tropical cyclones are few but the movement of masses of warm, moist tropical air, particularly down the East coast of Australia,(Africa, Argentina) is very noticeable.

    The cool waters currently present in the north Pacific probably had their genesis in the 1996 La Nina illustrating the long lag in ocean transport. It’s faster in the Atlantic.

    It seems that the continental areas do not change much in their OLR behaviour whereas the oceans, particularly the central Pacific does. It is here that the largest change in OLR occurs with the strong flux in reflective cloud. Less cloud allows the ocean to warm and while this is happening OLR is relatively low in the cloud free zone. With the current La Nina there is an expansion of the area of heaviest OLR into the western Pacific, normally an area where OLR is lighter. With all the cloud the Eastern coast of Australia has experienced a very cool summer.

    If there are hemispherical effects it is due to the relative proportion of land and sea.

    My best resource seems to be http://www.cdc.noaa.gov/map/time_plot/

  618. Posted Mar 11, 2008 at 7:38 AM | Permalink

    618 (Andrew): You must be looking at the old version. Sitting in your cache somewhere. Flush the cache by shutting down the browser. Anyway, 14 was just the smallest and 19 the largest cycle of the bunch. Don’t mean much, hence not emphasized on the new version.

  619. See - owe to Rich
    Posted Mar 11, 2008 at 11:54 AM | Permalink

    Andrew #609: “I don’t regard the solar cycle length thing highly anyway, given how long this cycle has been.”

    I don’t fully understand this remark, unless you are meaning that a long Cycle 23 should have led to lower Cycle 23 temperatures. But, according to my model, a long Cycle 23 will have its biggest effect (0.07 deg/year) on Cycle 24 temperatures, and some further effect (0.03 deg/year) on Cycle 25 temperatures. The fact that temperatures are currently plummeting, in fact faster than my model would suggest (modulo my sorting out within cycle effects), is highly interesting to me, and I believe supports the hypothesis. Though, as Leif says, relating it to low sunspot numbers for a long interval might be a better way to analyze the data.

    Regarding Leif’s “what if”, if Cycle 24 is 13 years long then I predict Cycle 25 temperatures to be 0.07*3 = 0.21C cooler than if Cycle 24 is 10 years long (apart from the fact that there will be 3 more years of global warming from CO2 by the time we get to the end of Cycle 25 :-)).

    Rich.

  620. Andrew
    Posted Mar 11, 2008 at 1:00 PM | Permalink

    My issue is that the original correlation was “too” good, to clarify. Your methodology is interesting to me, however.

  621. Raven
    Posted Mar 11, 2008 at 5:20 PM | Permalink

    Here is another strange correlation to consider: http://www.fao.org/DOCREP/005/Y2787E/y2787e03b.htm#FiguraB

    shorter days == higher temps?

    less time for OLR to escape at night?

  622. Andrew
    Posted Mar 11, 2008 at 6:08 PM | Permalink

    Raven, isn’t that from the paper that “rejects” General Relativity?

  623. Posted Mar 11, 2008 at 6:16 PM | Permalink

    623 (Raven): The text says: “Shifting the ACI curve by 4 years to the right (Fig 2.3B) results in almost complete coincidence of the curve maximums of the early 1870s, late 1930s, and middle 1990s.”
    Yet, Figure 2.3B says “LOD shifted 6 years to the right”. Clearly some confusion at a fundamental level. Now, a warmer [and higher] atmosphere should result in a longer day, so some correlation is to be expected…

  624. Andrew
    Posted Mar 11, 2008 at 6:45 PM | Permalink

    Leif, and here I was thinking it must be faked! An actual mechanism connecting the two. To bad it goes the wrong way!

  625. Andrew
    Posted Mar 11, 2008 at 7:29 PM | Permalink

    Actually, I had another problem with the Solar Cycle length, thing, to. That is, the extremely long and short cycles when you go futher back.

  626. Erl Happ
    Posted Mar 12, 2008 at 3:35 AM | Permalink

    Leif
    How do you see this? http://scienceandpublicpolicy.org/images/stories/papers/monckton/whatgreenhouse/moncktongreenhousewarming.pdf

    Strong statement linking radiation from the tropics and temperature change.

  627. Posted Mar 12, 2008 at 4:25 AM | Permalink

    628 (Erl): From the paper you cited:

    Does observation demonstrate that increases in outgoing long wave radiation in the tropics lead to increases in observed global temperature? The answer to that question is Yes

    Somehow, this does not strike me as so strange. See #498.

  628. Posted Mar 12, 2008 at 4:45 AM | Permalink

    629 (me): looking at the dT of #531 and at the figure on page 13 of the paper cited, one is hard pressed to see any warming at all over the period 1978-2008 [this was brought to my attention by an innocent bystander], so why are people debating AGW? :-)

  629. Erl Happ
    Posted Mar 12, 2008 at 8:22 AM | Permalink

    Leif,
    My somewhat theoretical interpretation for what it is worth:

    Let’s assume that the energy transmitted to the top of the troposphere is constant. The Earth is variably shielded by cloud. Cloud is a strong reflector of solar energy. For some reason the cloud disappears in a part where cloud cover is usually strong. The immediate result is a fall in OLR in that part.

    A warmer body must emit more long wave radiation. However, to warm, the body must store energy and it can not both store and emit it. Contradiction in terms. The reduction in radiation during an El Nino warming event is most obvious in the central and western Pacific. As it gains temperature the water increases long wave emission. It’s also moving westwards at the surface. For the temperature to rise the rate of absorption of energy must exceed the rate of emission. A generalised rise in emission is not inconsistent with a fall in emission in the area where the clouds are no longer present to reflect the usual large proportion of solar energy. In these newly cloud free locations emission falls. This is documented in figure 9 on page 9 of my paper at http://www.happs.com.au/downloaders/Cloud_temp_tropo.pdf which I am sure you have diligently perused.

    The interesting thing is that during an El Nino event the entire area of the tropics warms to a variable extent.

    Take away the cloud cover over the entire tropics and there would be an immediate fall in OLR as the oceans absorbed energy. As temperatures increased OLR would naturally rise with increasing temperature. However, in the tropics the response is not so much in temperature but in evaporation. How can the atmosphere sustain a reduction in cloud cover for more than a year under these circumstances? How is it that the ocean can warm, absorbing energy and yet the energy emitted can actually increase? Good question for a solar physicist.

    Is there a problem with the assumption that energy coming through the troposphere is a constant?

  630. Erl Happ
    Posted Mar 12, 2008 at 8:38 AM | Permalink

    630 (Leif)
    One thing I do notice on returning to your post 531 is that you could drop the temperature curves for SC 23 onto that for SC 22 and they would look pretty similar. Notice too, how the amplitudes of the fluctuation in both variables tend to vary together? It’s very hard to give up this notion that the energy incident at the top of the troposphere is not a constant.

  631. Posted Mar 12, 2008 at 9:23 AM | Permalink

    631-632 (Erl):

    the body must store energy and it can not both store and emit it. Contradiction in terms.

    If I stick an iron poker into a fire, I observe it get hotter, it begins to glow red hot. When I remove the red-hot glowing poker and grab it with my hand, it reminds me forcefully that the poker can at the same time both store and emit energy. I don’t see the contradiction.
    And the energy input to the upper troposphere is not constant, it varies 0.1% over the solar cycle, and several times as much on a time scale of days. The difficulty in understanding if the solar radiative forcing is strong enough to change climate hinges on how sensitive the climate system is to such small changes. You might remember that that was my initial question and very reason to be here. Sadly, I have not seen any serious and physically viable discussion of that fundamental issue. Instead, you have seen the stuff that I get bombarded with. I see it as a public duty to answer to these things the best I can, across very diverse levels of scientific training.

  632. Andrew
    Posted Mar 12, 2008 at 9:42 AM | Permalink

    Leif, Erl is suffering from interpreting “store” as “trap”.

  633. EW
    Posted Mar 12, 2008 at 10:30 AM | Permalink

    #625

    Based on this multidecadal periodicity of LOD, and the fact that LOD runs ahead of dT by 6 years, a gradually descending dT may be expected around 2005.

    And doesn’t it descend or stagnate these days?

  634. Posted Mar 12, 2008 at 11:50 AM | Permalink

    635 (EW): all you wanted to know about Earth Rotation (LOD) and Atmosphere

  635. jae
    Posted Mar 12, 2008 at 12:26 PM | Permalink

    Leif: MarkT linked this paper over on the BB. No name on it, but it’s an interesting summary. Any comments? (apologies if it’s been linked here previously)

  636. Posted Mar 12, 2008 at 1:15 PM | Permalink

    637 (jae): The paper has an optimistic tone that should give pause, phrases like “perfect vindication” and the like. The TSI correlations are based on Hoyt and Schatten’s old (1992) reconstructions which everybody in the field now recognize as having a much too large variation [the 'update' in 2005 is just tagging on the last couple of cycles without changing the early data when the increase happened]. In addition, there is the usual suspects [Friis-Christensen etc]. All in all, a one-sided rehash of old stuff. I, personally, would not attach much weight to it.

  637. jae
    Posted Mar 12, 2008 at 2:02 PM | Permalink

    Thanks, Leif. Kinda what I thought.

  638. Andrew
    Posted Mar 12, 2008 at 3:01 PM | Permalink

    Leif, I wonder, what would Hoyt himself say about that? Its not that I would give his opinion of his data more weight than yours, just that I’m curious.

  639. cba
    Posted Mar 12, 2008 at 3:02 PM | Permalink

    Paper refers to a perfect verification specific to Shindell and Labitzke and Van Loon’s work on a specific topic. It didn’t look like there was a tremendous dependence totally on Hoytt etc. Also, it looked like the TSI from them was overdoing it a bit.

  640. Andrew
    Posted Mar 12, 2008 at 3:11 PM | Permalink

    cba, I’m bothered by the fact that the Labitzke and Van Loon verifivation graphics are labeled as Strataspheric heights and then tropospheric ones. I don’t get that. Also, I don’t get the point of associating high Geopotential heights in the Arctic with magnetic storms and the expansion of the Auroral ring…In short, its interesting, and has some nice graphs, but some of it makes very little sense to me.

  641. Posted Mar 12, 2008 at 3:44 PM | Permalink

    640 (Andrew): Last I communicated with Hoyt he did not express any particular view on the old paper. You can compare the H&S TSI with other, more recent, ones on my website. Click on “Seminar-LMSAL.pdf (Seminar at LMSAL, 2008)” and scroll down to page 20 [the last, if you don't want to look at the rest]. Note how there is a progressive trend towards smaller and smaller TSI-variation as you go to more recent reconstructions.

  642. Andrew
    Posted Mar 12, 2008 at 4:12 PM | Permalink

    Thanks for the info, Leif. Doug is a conributor to that site, as I understand it, but I wouldn’t know if he endorsed its use or not.

    List of Icecap people:

    http://icecap.us/index.php/go/experts

    Douglas V. Hoyt, Solar Physicist and Climatologist

    Douglas V. Hoyt is a solar physicist and climatologist who worked for more than thirty years as a research scientist in the field. He has worked at NOAA, NCAR, Sacramento Peak Observatory, the World Radiation Center, Research and Data Systems, and Raytheon where he was a Senior Scientist. He has conducted research on issues related to climate change, changes in solar irradiance on all time scales, and the sun-climate connection. His most recent publication is the book “The Role of the Sun in Climate Change”. He has published nearly 100 scientific papers on solar irradiance variations, the greenhouse effect, atmospheric transmission, aerosols, cloud cover, sunshine, radiative transfer, radiometers, solar activity, sunspot structure, sunspot decay rates, and the history of solar observations.

  643. Andrew
    Posted Mar 12, 2008 at 4:15 PM | Permalink

    Also, his page on Warwick Hughes’s site:

    http://www.warwickhughes.com/hoyt/climate-change.htm

  644. Andrew
    Posted Mar 12, 2008 at 4:31 PM | Permalink

    Looking at the last slide, I’m curious as to what you think of the apparently high variation in reconstructions relying on isotopes? I apologize if this has come up before. As I understand it, you suggested that there may be some other source of Cosmic Ray variation, correct?

  645. Posted Mar 12, 2008 at 4:44 PM | Permalink

    646 (Andrew): Most reconstructions tie TSI to the Sun’s magnetic field. There has been the notion [starting all the way back with me in 1977] that the Sun’s magnetic field has more than doubled since ~1900. Recent work by several people [incl. myself] indicate that this doubling very likely did not happen, hence all reconstructions and calibrations [of response to cosmic rays] that rely on such doubling are suspect. I don’t have the space to go into details here, but several [many, actually] of the papers on my website address various aspects of this research. In the interest of full disclosure, I must add that not all researchers accept my findings, especially if I disagree with theirs. This is a normal human reaction, sometimes dressed up as a virtue ["I stand by my earlier results"].

  646. Andrew
    Posted Mar 12, 2008 at 5:25 PM | Permalink

    Hi Leif, I guess what I’m wondering is, has all previous work been using the magnetic field doubling thing? This is what I mean when I refer to “isotopes”:

  647. Posted Mar 12, 2008 at 7:00 PM | Permalink

    648 (Andrew): The short answer is “yes”. If you look carefully you will notice that they are share a distinct increase from 1900 to about 1950. That is the “doubling” they all key on to. The real answer is much longer, involves sunspot number counts in the past, cosmic ray calibration non-linearities, etc, etc. The ‘take-away’ message is that these reconstructions are not ‘settled’. Even the simple fact that they differ so much should tell you that…

  648. Posted Mar 12, 2008 at 7:29 PM | Permalink

    649 (me): I don’t remember about Bard and Cowley [but AFAIK, they also do it], the current wisdom is that the strength of the Heliospheric magnetic field [HMF]determines the modulation of cosmic rays [CRs] [I don't think so, but what does that matter when we discuss work by others that do?]. We only have good direct measurements of cosmic rays back to ~1952, and some back to ~1935 [but with uncertain - or at least, disputed - calibration]. We have spacecraft measurements of HMF back to ~1960. Correlation of the CR modulation with the measured HMF, allows us to calculate HMF if we know CR. The problem with this is that since 1960, the HMF has not varied enough to get a good correlation. Using geomagnetic activity, people try to estimate HMF back to ~1850, and we get a longer time base for the correlation HMFCR. Now, we have to correlate 10Be or 14C to CR during modern times. Atom bomb explosions kind off screwed up the 14C data, and the upper few decades of 10Be from ice cores are not reliable [so I'm told], so the HMF calculated from geomagnetic activity back to ~1850 [or ~1900] becomes important [the 'doubling']. Now, the HMF depends somewhat on the number of sunspots, so we can also try to calculate relations SSN HMF CR 10Be (say). TSI also depends on the sunspot numbers, so maybe we can say: TSI SSN HMF CR 10Be climate; as the deposition of 10Be depends on snowfall. For some of these we have other ways of cross-checking, but not very precisely, so we have a long chain of uncertain [and disputed] connections. The bottom line is that this whole thing is wide open and not ‘settled’ at all, although you will find people that say that it is settled and that everything is understood [at least by them]. In this respect the solar field is no different from the climate people [although somewhat more civil].

  649. Posted Mar 12, 2008 at 7:35 PM | Permalink

    650 (me): when I string things together like: TSI SSN, I mean to say TSI <-> SSN. I forgot to ‘escape’ the less-than-sign, etc.

  650. Andrew
    Posted Mar 12, 2008 at 8:07 PM | Permalink

    Yikes, I had no idea how difficult it was to reconstruct solar activity/CRF variability!

  651. cba
    Posted Mar 12, 2008 at 8:13 PM | Permalink

    #642, Andrew,
    I’m not taking much away from that paper in that area. Supposedly it’s showing a relationship or effect from stratosphere to troposphere. Pretty much the rest is lost on me at present.

    Scafetta & West’s paper tend to favor Wang 2005 for providing the TSI reconstruction. Evidently it provides a time relaxation which is more in the ball park for a real phenomenon. It provides somewhat high sensitivity. Leif’s is close to that but evidently without a rising trend. But that brings up the question of what happened with the Maunder minimum etc. if there was almost no variation or how did the earth survive with a super sensitivity to TSI. Note that I can believe that GHGs are far less in effect than solar variation due to absorption/emissivity relationship.

  652. Andrew
    Posted Mar 12, 2008 at 8:21 PM | Permalink

    cba, I’m beginning to suspect that Cosmic Rays vary considerably independently of solar variability. I don’t have definitive proof (note all the issues with recons/measurements by Leif above) and I can’t explain why this would be the case, but if so, CRF could explain some of apparent correlations of TSI with climate, without the hypersentivity. This is the idea that this discussion has me toying with, but I may abandon it if it gets to be to akward (so far its only mildly akward).

  653. Phil.
    Posted Mar 12, 2008 at 9:50 PM | Permalink

    Leif what was the status of the recent sunspots, which cycles were they attributed to? Thanks.

  654. Posted Mar 12, 2008 at 9:59 PM | Permalink

    655 (Phil): all spots the last several years except the tiny one in January 2008 were old cycle 23 spots.

  655. Posted Mar 12, 2008 at 10:11 PM | Permalink

    someone [maybe in a different thread] referred to this Figure

    which has the caption: “The annual number of magnetic storms is represented by each bar of the histogram. Superimposed is the smoothed sunspot number. The dashed lines indicate solar minima and the dotted lines indicate solar maxima. Note the correlation of magnetic activity with solar activity and the apparent increase in magnetic activity with time”.

    The number of storms is determined as the number of days when the average geomagnetic aa-index exceeded a certain value. Since aa is incorrectly calibrated [too small before 1957 and other problems] the Figure gives an incorrect impression of geomagnetic activity with time. The Dst-index is specifically designed to measure storms. Dst is negative, and the more negative, the greater the storm. Jeffrey Love of the USGS has compiled Dst back to 1905 and we are in the process of taking it further back. A great storm has Dst < -150. Here is the number of such storms per year since 1905 [the red curve is the sunspot number, SSN]:

    it should be clear that there is no increase of storminess with time. The number of storms simply followw the sunspot number.

  656. Pete
    Posted Mar 13, 2008 at 4:09 AM | Permalink

    The point Leif brings up about a doubling of the magnetic field over the 20th Century now being challenged is interesting, and also from the mainstream point of view, as it changes what we understand to be the TSI also.

    It is well recognised by all mainstream scientists – one of the few areas there is no diagreement – that the temperature rise from around 1900 to 1940 (+0.4C) is firmly in the ball park of Solar changes. The GCM’s have even produced this. From the two scenerios run by the GCM’s of ‘natural’ only and ‘natural + Greenhouse Gases’ – it is not possible for them to reproduce the bulk of the 1900-1940 Warming without a substantial increase in Solar Forcing over this period.

    So, Leif is actually, but does he realising it(!?), is also opening another very big can of Worms here – which is: without the increasing Solar trend thought to have occured 1900-1950, how do we exlain this Warming now!? A puzzle indeed for all Climate Scientists.

  657. cba
    Posted Mar 13, 2008 at 7:26 AM | Permalink

    Andrew, 654

    Cosmic rays vary due to the fact they originate out in the cosmos. Currently, we’re in an arm of the galaxy (milky way). Eventually, we’ll orbit around to the other side of the galaxy and be closer to whatever is out that way. Our warm secure little niche here in the arm is not always the case, we can bobble up and down and back and forth a bit as we orbit around. There’s also supernovae, star formation and all sorts of things going on in a random fashion. Hence there seems to be long term variations due to our orbit around the milky way and just plain random events going on. Shaviv and / or Svensmark have discussed such things in their papers. Name has been dubbed cosmoclimatology.

    The sun also emits lots of lower energy cosmic rays (mostly protons despite the misnomer of ray). It looks like there are some moderately high powered ones from there too – according to one person I know. The really high powered ones don’t matter as they’re rare.

  658. Posted Mar 13, 2008 at 7:33 AM | Permalink

    658 (Pete): of course I realize the implications for climate science. That’s why I’m here. The problem is even bigger than the 1900-1940 period, there are the LIA and the MWP too, which may be without solar forcing as well. If the TSI has less than a 0.1 % to play with, if the HMF did not fall to [near] zero during the MM, if therefore the CRs were not out of the ordinary, etc, then we either have to accept hypersensitivity or no solar effect, and our understanding [and the GCMs] is in trouble. That is why this all is so very exciting, much more so than some bristle cone pines. Since this is so big, there is and will be a lot of resistance against the whole notion. Nobody wants the boat rocked too much.

  659. Posted Mar 13, 2008 at 7:45 AM | Permalink

    659 (cba): The variations of CR due to events [except a supernova next door] in the Galaxy and the Sun’s movements within the Galaxy are so slow that we can discount them as climate effectors on the time scale of centuries [although lots of fun speculation can be made about what happens over millions of years]. The modulation of CRs by solar activity is what might be of interest.

  660. Erl Happ
    Posted Mar 13, 2008 at 8:07 AM | Permalink

    633 (Leif) 634 (Andrew)

    The oceans have low albedo and a great capacity to absorb energy. Clouds have relatively high albedo and virtually no capacity at all to absorb energy. Remove the cloud and albedo diminishes. The oceans then absorb energy. Outgoing Long wave Radiation must contemporaneously fall to a degree equivalent to the energy absorbed and stored by the ocean. OLR will return to normal once the ocean has increased in temperature (or provides sufficient evaporation, or warms the atmosphere by conduction) sufficient to return OLR to its pre-existing level.

    An increase in temperature in the tropics over a period of one or two years (a process of energy absorption that will of itself reduce OLR) that is accompanied by an increase in OLR above and beyond the pre-existing state, requires an incremental increase in the energy supplied to the system equivalent to the energy needed to increase the water temperature (stored energy) plus the energy required to sustain the increase in OLR (emitted energy). The interesting question is what disturbs the cloud albedo in the first instance and maintains that state of reduced albedo against the countervailing tendency for increased evaporation to restore the cloud cover.

    Yes, the hot poker will emit more radiation once heated. Equally, a cold poker placed in a furnace will reduce the heat emission of the furnace as a whole. Now, consider how far this analogy differs from the one we are considering, the Earth in relation to solar energy. The heat absorption capacity of the ocean is relatively much greater than the poker. The incident energy from the sun can not be compared to the heat of a furnace. Consider the disparity in surface temperatures.

    I am not denying that OLR for the tropics or the Earth as a whole will increase during an El Nino event. I am asking you to consider where the extra energy is coming from that will maintain a higher ocean temperature and increased OLR.

    For the relationship between OLR and the SOI (both narrowly focused upon the area of the Pacific where OLR diminishes during an El Nino event (massive increase in stored energy) check the data for yourself.

    Here are the sources:
    OLR: http://www.cpc.ncep.noaa.gov/data/indices/olr I used the anomalies which graphs very well against the Southern Oscillation Index
    SOI: http://www.bom.gov.au/climate/current/soihtm1.shtml

    I graphed the period 1996-2000 covering the big El Nino event of 1998 and the following La Nina.

  661. Posted Mar 13, 2008 at 8:19 AM | Permalink

    662 (Erl):

    I am asking you to consider where the extra energy is coming from that will maintain a higher ocean temperature and increased OLR.

    Well, it is not coming from the Sun.

  662. Phil.
    Posted Mar 13, 2008 at 11:01 AM | Permalink

    Re#662

    The oceans have low albedo and a great capacity to absorb energy. Clouds have relatively high albedo and virtually no capacity at all to absorb energy. Remove the cloud and albedo diminishes. The oceans then absorb energy. Outgoing Long wave Radiation must contemporaneously fall to a degree equivalent to the energy absorbed and stored by the ocean.

    This is not true, the ocean will continue to emit OLR just as it did before unless the temperature rises in which case it will increase. In the case where the ocean is covered by cloud the OLR won’t make it past the cloud but will be replaced by OLR from the cloud tops which will be colder than the ocean and therefore emit less OLR.

  663. jae
    Posted Mar 13, 2008 at 11:26 AM | Permalink

    About half of the solar energy is in the IR part of the spectrum. Where does all this energy go when it’s cloudy?

  664. cba
    Posted Mar 13, 2008 at 11:49 AM | Permalink

    665 Jae
    Much of the incoming IR is close to visible and subject to similar reflectivity – at least some of it. Other incoming will be absorbed by the cloud.

    Leif,
    I thought there was some reasonable evidence indicating CR modulation at the LIA as well as evidence that the LIA was a global event. Unless you want to start looking at the crab nebula’s supernova in 1054 as prime suspect, aren’t you going to have to start looking at these factors as affecting the limits to just how small the solar variability can be?

  665. Posted Mar 13, 2008 at 11:56 AM | Permalink

    666 (cba): The CR modulation continued straight through the LIA and the MM, showing that the solar cycle was operating. What we don’t know is how the sun managed to hide the spots. Maybe they were there, but so small [but numerous] as to escape detection. Or maybe the spots are not the main players in the solar cycle after all, but only incidental effects. We don’t know. This is one of the biggest mysteries in solar physics. THE biggest one is simply: what is a sunspot? What keeps it together?

  666. Andrew
    Posted Mar 13, 2008 at 12:49 PM | Permalink

    Yes, Leif is correct, Be shows 11 year cycles even during the Maunder Minimum (though it does appear to show low CRF):

    Be and SSN don’t perfectly line up.

  667. Andrew
    Posted Mar 13, 2008 at 12:53 PM | Permalink

    BTW, Leif, what is the difference between Group and Wolf SSN? Wolf doesn’t go back far enough, but it appears to show small cycles during the late Maunder Minimum, when Group is totally flat.

  668. cba
    Posted Mar 13, 2008 at 12:53 PM | Permalink

    I think someone has the theory that we could have passed through a dense region of interstellar space (something quite a bit more than what you see this time of year that has the name zodiacal lights). Whether such is possible to reduce the TOA in the incoming solar to provide needed power reduction and whether there is any evidence for such an event are questionable.

    As I understand, solar wind does much or most of the modulating and coronal holes are where one sees variations concerning the solar wind. Is this tied to SS only ? Or, does one have a variable sun manifesting activities and intensity including magnetic?

  669. Posted Mar 13, 2008 at 1:15 PM | Permalink

    669 (Andrew): Rudolf Wolf defined the Wolf number as W = k*(10*g + s) where g is the number of groups [of spots] and s is the number of spots. Hoyt and Schatten defined the Group number as G = q*(13*g) where k and q are calibration constants [depending on who does the counting - this is not QUITE correct, but close enough for here]. W and G disagree because of disagreements about k and q. It is like the temperature anomalies: most of the differences comes from ‘adjustments’ rather than observations. IMHO, both W and G are wrong [in different ways]. Of course, people that feel that W fits better with their pet ideas think that W is correct and G is wrong, and vice versa. None of them entertain the possibility that BOTH are wrong.

    670 (cba): CR modulation is a BIG topic. In my opinion [not shared by most CR physicists], the important factor is the ‘geometry’ of the heliospheric current sheet, but the details are not important for the discussion about effects of the Galaxy.

  670. cba
    Posted Mar 13, 2008 at 2:17 PM | Permalink

    well, long term galactic influences are not unimportant any more than prospect of snowball earth. They just aren’t relevant for the present which brings back the question of alternatives to TSI variations or spectral variatio n or hypersensitivity.

  671. Posted Mar 13, 2008 at 2:46 PM | Permalink

    672 (cba): would be true if the galactic effects have been demonstrated, but at this point they are just fun speculation.

  672. cba
    Posted Mar 13, 2008 at 3:36 PM | Permalink

    demonstrated? other than waiting around for a few hundred million years to collect several cycles worth of data what would you consider as beyond just speculation? Both Shaviv and Svensmark have studied this and analyzed and presented evidence that supports the idea. My recollections are that this is a viable theory on par with the alternatives – what few there are. It doesn’t solve the short term – at least without sort of variability, magnetic, spectral, power or whatever – and most likely one would think – from the sun.

  673. Posted Mar 13, 2008 at 4:58 PM | Permalink

    674 (cba): There are problems with those long cycles. Maybe the most severe is the assumption that the spiral arms are static. They are not. They are ‘dynamic’ waves that travel through the [not symmetric, but barred] Galaxy. But, hey, it is good clean fun and I would not exclude the modulation of CRs due to that mechanism. It is the next step, that I have more problems with: that CR controls the climate. But since we don’t really know what does, one should not be dogmatic about it [so I ain't].

  674. Posted Mar 13, 2008 at 5:00 PM | Permalink

    675 (me): and I think that S and S are not really interested in the first link of the chain, but that their interest is in the second [weaker] link.

  675. cba
    Posted Mar 13, 2008 at 7:02 PM | Permalink

    well, differential rotation would appear to be a problem if the arms aren’t a pressure wave. Shaviv is an astronomer with research interests in Eddington radiated limited stellar contraction if I recall. I don’t think he would make a seriously poor mistake concerning galactic structure. Of course there too, pressure waves are a theory, albeit rather popular and I don’t think the realizations that rotation of even the galactic structure depends on the size of the mass in the middle (super massive super small object – perhaps a black hole) is going to change that. As for barred spiral, that too is in heavy debate as we may be a spiral, an SBb or SBc barred spiral or yet some category in between.

    I’m not quite sure what you’re categorizing for link one and two – but then I just came off of two 13+ hr days that make up the middle of my week and this time around monday morning felt like a full bore friday afternoon.

    It’s almost a safe bet that clouds control the climate. (probably something like lindzen’s iris). It’s got to be a massive self adjusting feedback mechanism – that can become fouled up by lots of snow & ice on occasions. Regardless of GHGs energy is transported by the water cycle in addition to radiative mechanism.

    CR is one of a number of possible contributors to this cloud thing, making it nice and messy. There is also always the prospect that one of our ‘constants’ is in fact a variable and no one is looking at it yet.

  676. Posted Mar 13, 2008 at 7:11 PM | Permalink

    677 (cba): the links in the chain: A=(galactic position,cosmic rays) -> B=(cosmic rays,clouds) -> C=(clouds,climate). While A and C are not much in doubt, the weak [but interesting] link in the chain is B. And that is the one I have a problem with. Discussing A and C are discussing straw men.

  677. Chris Knight
    Posted Mar 13, 2008 at 7:44 PM | Permalink

    #667 (Leif): We do know how some of the spots were hidden, the clue is cold summer weather, mild winter weather which indicates that the sun was covered by clouds. Extreme floods, frosts and storms reported by multiple 17th C. diarists. We also know of widespread duststorms due to droughts, famines, plagues, political, religious and international wars in part due to these causes.

    There were periods of elevated Volcanicity in the 1600s: Fogo 1638, Llaima, Komaga 1640, Kelut, Awu 1641, Long Island 1662, Gamkonora 1673, Tongkoko 1680, and several others, at least 18 major events over the 50 year period, accompanied by frequent El Nino events.

    We also know that early Astronomers like Flamsteed and Halley considered Sunspots as mere curiosities, fit only to show that the sun rotated, and that the spots were on the sun’s surface; accurate mapping of the night sky for navigation was their main function. I guess they had to sleep sometimes. We know they worked in rather polluted city atmospheres with unwieldly equipment that often needed more than one assistant in order to make and record an observation. Flamsteed, in particular was very poorly paid, and equipped the observatory himself, and disagreed vehemently with some of his colleagues. He was known to destroy stuff he disagreed with, because he was professionally jealous, and his wife kept the contents of his observatory after his death.

    We now know that there were some extreme volcanic events around the world – especially during the Dalton period – Laki/Grimsvotn in 1783-85 which has reported smogs of hydrogen fluoride gas ( I wonder how that affected the optics of the time?) across Europe, followed by perhaps 10 years of cool UK summers despite several El Nino events, which led to warm wet winters at the same time. The solar minimum was observed in 1784, I believe, and the next, reported about 14 years later. Perhaps Laki’s dust veil brought forward the observed minimum in 1784 by a year or so?
    The 1810/11 minimum also followed by about a year, another powerful volcanic event, which was recorded in ice cores from both poles, from one or possibly 2 near equatorial volcanoes, which have yet to be definitely associated with the evidence. Summer temperatures (UK HadCET)were again depressed by 2-3 deg C for about 9 years, again with several El Ninos at the same time, at which time Tambora erupted in the year without a summer 1815. The next solar minimum was in 1823, just a year after Galunggung erupted. and then ended the Dalton minimum. There should have been 5 minima during this period – only 4 were counted, and three of them recorded during times of global atmospheric dimming.

    What about the Cold winters though? Well, the 2007/8 Arctic ice recovery and the NH winter may be a clue. Additional winter precipitation over continental landmasses leads to snow and an extension of winter temperatures, both seasonally and latitudinally. Maybe not every year.

  678. cba
    Posted Mar 13, 2008 at 8:34 PM | Permalink

    Leif,

    I’m not sure one could really call A as being a done deal. It too is an assumption that won’t be verified any time soon – that CR varies as we approach the edge or go outside the edge of the arm. However, any consequences of such are so far in the distant future as to be not of concern. As I recall Svensmark is definitely into the Cloud experiment at Cern but I don’t recall reading anything of Shaviv on that.

    Last thing I recall reading was correlations between CR and cloud cover were shown on a cross hatched map suggesting most was more around the UK than the S. Pacific and that results were either null or inconclusive for most of the rest of the globe.

    My understanding of navigation was that while lots of work was being put in to stellar positions, the real problem was longitude which was timing. Newton was somewhat involved in that but with jupiter’s moons. Such serious astronomy for navigation was really totally immpractical and the solution became the chronometer – william harrison. Precision stellar positions were a popular thing back then – so was comet hunting but neither were that crucial for navigation. After all, the sun was a bit more practical a target for lattitude measurement and no astronomical target was suitable for longitude.

  679. Posted Mar 13, 2008 at 8:43 PM | Permalink

    679 (Chris): Yeah, that was also the argument most solar astronomers came up with when Maunder [and Spoerer, before him] first claimed that there were no spots during what we now call the Maunder Minimum. Jack Eddy looked at the problem again in the 1970s and concluded that the dearth of spots was real. It is a curious fact that his main argument that convinced most people was the LIA and the 14C record. Hoyt and Schatten looked again in the 1990s and concluded that the observers did not miss any spots. In my opinion they overstate their case. They give for each observer for each year a table showing for each day how many spots that observer saw. For, say, the 19th century there would be a reasonable number of ‘missing days’ where no observation was made [mostly because of weather]. But for the 17th century observers there were no missing days! recorded in Hoyt’s and Schatten’s tables. This is not credible as the observers were in Europe [e.g. London] and certainly the weather [and maybe the pollution] would at times [often? as the weather was miserable during the LIA] hinder observations. Possibly H&S misinterpreted a statement like ‘during all of 1666 I did not see a single spot’.
    Most astronomers today accept the lack of spots during those ‘Grand Minima’ because of the LIA and 14C [and also 10Be] arguments, but there is problem with the solar cycle modulation of CRs without sunspots, so maybe it is time to revisit the Grand Minima. If we were to enter one in the next cycle or two, we might simply observe what goes on.

  680. Posted Mar 13, 2008 at 8:53 PM | Permalink

    680 (cba): the difference between ‘not in doubt’ and ‘a done deal':
    ‘not in doubt’ == we don’t know what else
    ‘a done deal’ == we ‘know’ it’s not something else :-)

  681. Posted Mar 13, 2008 at 9:18 PM | Permalink

    679 (chris): the volcanic connection extends probably also to the cosmic ray record. See my paper at SORCE 2008 on my website. Click on “TSI From McCracken HMF.pdf (TSI Reconstruction 1428-2005, Santa Fe, SORCE 2008)” and go to page 2.

  682. Posted Mar 13, 2008 at 9:24 PM | Permalink

    683 (me): grrr. website

  683. David Archibald
    Posted Mar 14, 2008 at 2:29 AM | Permalink

    Given the failure of NASA’s prediction panel to pick the 23/24 minimum, I feel compelled to step in and provide some guidance with my recent paper presented at the New York Climate Conference earlier this month, Solar Cycle 24: Implications for the United States.

    It is available at: http://www.lavoisier.com.au/papers/articles/ArchibaldMarch2008.pdf

    Apart from my prediction of a Solar Cycle 24 amplitude of 45 and minimum in July 2009, my contributions to science include calling Dr Hansen’s 350 ppm danger level for CO2 “spectacularly idiotic”.

    http://www.lavoisier.com.au/papers/articles/ArchibaldMarch2008.pdf

  684. Posted Mar 14, 2008 at 6:20 AM | Permalink

    685 (David A): Using the obsolete Hoyt&Schatten TSI diminishes the value of what you are trying to say. And I’ll admonish you to avoid characterizations like “s… i…”, true or not.

  685. Posted Mar 14, 2008 at 7:58 AM | Permalink

    Here is David Hathaway’s latest ‘butterfly diagram’ for cycle 23 and 24 [the little blip at 30N near the right-hand edge]:

    One interpretation of this is that solar minimum is at least half a year away [certainly not right now as predicted by the Panel, although we did claim a +/- 6 month uncertainty].

  686. cba
    Posted Mar 14, 2008 at 12:11 PM | Permalink

    David,
    a most interesting paper.

    how volcanism fits in isn’t really covered though.

    I concur with Leif that hansen may not be S & I but rather might be sinister, cunning, greedy, and/or political. It’s at present not a fact in evidence and as just stated, is something with numerous possible alternatives. Alternatives should be left to the mind of the listener after being presented with the facts that lead to such conclusions.

  687. kim
    Posted Mar 14, 2008 at 3:40 PM | Permalink

    The road to Hell is paved with good intentions, but people forget that the road to Paradise is similarly paved. I believe his intention was to save the Earth, and he simply lost his way.
    ================================

  688. Posted Mar 14, 2008 at 3:45 PM | Permalink

    685,688,689 (all): It is OT to comment on somebody’s intentions, motivation, character, or personality.

  689. kim
    Posted Mar 14, 2008 at 4:12 PM | Permalink

    Sure, and sorry about that. Is there any meaning in the fact that most of the last spots from Solar Cycle #22 are Northern Hemisphere, and most of the last from #23 are in the Southern?
    ==========================

  690. Posted Mar 14, 2008 at 4:40 PM | Permalink

    691 (kim):

    Is there any meaning in the fact that most of the last spots from Solar Cycle #22 are Northern Hemisphere, and most of the last from #23 are in the Southern

    Probably not. It is quite normal that you have such asymmetries and they change from cycle to cycle. During the MM, almost all the spots were in the South. Here is a plot of number of spots in each hemisphere [blue=N; red=S] for the last five cycles:

  691. cba
    Posted Mar 14, 2008 at 5:10 PM | Permalink

    Leif,

    How much info is there there on the SS during the maunder minimum? There is enough to say they were most all in the south?

  692. Posted Mar 14, 2008 at 5:28 PM | Permalink

    693 (cba): Yes, whenever observers reported a spot (and some 50 or so were observed, IIRC, versus thousands in a normal cycle) it was almost always in the South.

  693. kim
    Posted Mar 14, 2008 at 5:30 PM | Permalink

    How about during the Dalton?
    ==================

  694. Posted Mar 14, 2008 at 5:40 PM | Permalink

    694 (me): and most of the spots were very large and long-lived, while today most spots are very small and short-lived.

  695. Posted Mar 14, 2008 at 5:45 PM | Permalink

    695 (kim): Dalton? This is very hard to answer, because [and this is a little known secret] there were very few observations. In fact, Rudolf Wolf couldn’t find any for several years just after 1800 and had to ‘make up’ sunspot counts from a relationship with auroral sightings. Hoyt and Schatten did find some few and scattered observations, but it is fair to say that we don’t know with any precision how many spots [and where they were] there were during the Dalton minimum.

  696. kim
    Posted Mar 14, 2008 at 5:49 PM | Permalink

    Very interesting, and thank you. It seems the sun was dreaming, as may I have to over this.
    ========================================================

  697. kim
    Posted Mar 14, 2008 at 5:55 PM | Permalink

    Was it failing to flip magnetic poles, then?
    ==========================

  698. cba
    Posted Mar 14, 2008 at 5:57 PM | Permalink

    large and long lived versus small and short lived sure suggests something different between then and now.

    The simple fact that a spot is large (and cooler than the rest of the photosphere) is going to cause a reduction in TSI while the spot is near the central area of the disk (rather than on the limb). Seems like a exoplanet crossing the face of a star can reduce the apparent luminosity we see by a good 0.1%, perhaps more, making it plausible to find them with small telescope systems. I wonder just how much reduction in TSI is accomplished by your big spot or group? Offhand it would seem plausible that it could be somewhat more than 0.1%

  699. Posted Mar 14, 2008 at 6:06 PM | Permalink

    699 (kim): “flip poles”? we don’t know. Most researchers would say that the solar magnetic cycle was still operating. The reason being that if the poles stopped flipping it would be hard to start them again.

    700 (cba): Large spots are surrounded by large facular areas that on the whole compensate for the darkness of the spots [before and after the spot], but it it not known if that was also the case [likely, though] during the MM and to what degree.

  700. kim
    Posted Mar 14, 2008 at 6:23 PM | Permalink

    It’s hard to start a stopped train, but you do it by working on one pole.
    ==========================================

  701. Posted Mar 14, 2008 at 6:26 PM | Permalink

    700 (cba): small telescopes? Here is one:

    the real problem was poor seeing and poor optics.

  702. Posted Mar 14, 2008 at 6:36 PM | Permalink

    702 (kim): this is something that is being worked on: reversal of solar poles

  703. Chris Knight
    Posted Mar 14, 2008 at 6:38 PM | Permalink

    Arlt R. Solar Physics. February 2008
    DIGITIZATION OF SUNSPOT DRAWINGS BY
    STAUDACHER IN 1749–1796

    Abstract. Original drawings by J.C. Staudacher made in the period of 1749–1796 were
    digitized. The drawings provide information about the size of the sunspots and are therefore
    useful for analyses sensitive to the sunspot area rather than Wolf numbers. The total
    sunspot area as a function of time is shown for the observing period. The sunspot areas
    measured do not support the proposition of a weak, “lost” cycle between cycles 4 and 5.
    We also evaluate the usefulness of the drawings for the determination of the positions of
    the sunspots for future studies.

    refutes Usoskin et al.,2001 in that it indicates 1795 was a solar mimimum, rather than a maximum.
    It puts minima at 1758,1766.5,1776,1785,1795.

  704. Posted Mar 14, 2008 at 6:45 PM | Permalink

    705 (Chris): Nice! The “lost cycle” was not accepted by most solar researchers anyway. But it is necessary that ‘crazy’ ideas be given a chance now and then. [They might turn out to be right, e.g. Maunder's minimum, missing solar neutrinos, etc]

  705. David Archibald
    Posted Mar 14, 2008 at 8:51 PM | Permalink

    Re 688, as a geologist who has kicked a lot of rocks, volcanism is over-rated. We haven’t had a big eruption for a while and apparently the atmosphere is continuing to clear. According to warmer theory, that should make the planet hotter, but we have had ten years of cooling now, and a lot of cooling in the last year. If fact, half the children alive today have only ever experienced global cooling.

  706. jae
    Posted Mar 14, 2008 at 9:02 PM | Permalink

    686, Lief:

    685 (David A): Using the obsolete Hoyt&Schatten TSI diminishes the value of what you are trying to say. And I’ll admonish you to avoid characterizations like “s… i…”, true or not.

    Well, I don’t understand “diminishes” here. Are David’s general conclusions correct or not? If not, why not? You are sounding much like an AGW apologist here ?

  707. Posted Mar 15, 2008 at 6:07 AM | Permalink

    708: JAE

    I understood Leif’s comment to mean that being possibly right for the wrong reason(s) is not good science, and harms the valid portions of the paper.

  708. Posted Mar 15, 2008 at 7:00 AM | Permalink

    708 (jae):

    are general conclusions correct or not?

    Is not the correct question. The conclusions can be correct or not quite independent of the reasoning that led to said conclusions. The is a tribe in darkest Africa that believes that the beating of tam-tam drums will restore the sun after a solar eclipse. So far they have not been wrong.

    That solar activity is declining is something that I won’t disagree with. That CO2 generally is good for us [and the plants] is something I won’t disagree with. That we are not staring disaster from global warming in the face is something I won’t disagree with. Where I have a problem is the poor way the ‘evidence’ is brought to bear.

    I could go through the paper point for point, but I first thought that the poor quality of the paper speaks for itself. But apparently not.

  709. Francois Ouellette
    Posted Mar 15, 2008 at 9:48 AM | Permalink

    #710 Leif,

    One can only agree with you. I’ve looked at this whole “climate conference”, and, quite frankly, scientifically it’s a joke. I think that if AGW is to be disproved, it has to be with scientific rigor, and inventiveness. It’s not enough to always talk about things we don’t know, you’ve got to find those things too! Skepticism is OK only if you make something of it. If there are better explanations for the warming, what are they? Not just handwaving, but solid theories, models, proofs.

  710. Chris Knight
    Posted Mar 15, 2008 at 12:17 PM | Permalink

    #705 (me), #706 Leif

    Back to Arlt for a moment if not too OT, please. If the cycles 1-4 were 8.5 y, 9.5 y, 9 y, 10 y for minima at around 1758, 1766.5, 1776, 1785, and 1795, does this say anything about the relative size of the peaks at the time, i.e. short cycle, longer cycle, repeated? Could anything be inferred from this predicting cycles 5-6, since we really have very sparse observations at that time?

  711. Posted Mar 15, 2008 at 12:29 PM | Permalink

    712 (Chris): First, the minimum is 1795 is not well determined as he is running out of data. It could easily be that 1796 and 1797 were low too, which would put the minimum a year or two later. Second, the [short cycle high activity, long cycle low activity] rule is not absolute. Whether a long cycle predicts a low next cycle is refuted by e.g. cycle 20 -> cycle 21. The sun doesn’t quite go by our numerology, unfortunately :-)

  712. David Archibald
    Posted Mar 15, 2008 at 8:07 PM | Permalink

    I know I am doing good when warmers deride my work. The best example of that is this post on Realclimate after a paper of mine from June 2007: http://www.realclimate.org/index.php/archives/2007/10/my-model-used-for-deception/

    When the warmers complain, you know you are pushing the right buttons. We are in a strange world in which NASA is too afraid to make a solar cycle prediction. After calling the beginning of Solar Cycle 24 for a couple years now, they have gone silent. My guess is that their political instinct tells them that another wrong prediction will show them to be floundering about and clueless.

    Dr Svalgaard, has Dr Hoyt abandoned his TSI reconstruction with Dr Schatten and recognised the superiority of your more recent one? The funny thing is that the Hoyt and Schatten reconstruction has a better fit with other data series than Dr Svalgaard’s invariate Sun.

  713. Posted Mar 15, 2008 at 8:48 PM | Permalink

    714 (David A): The problem with your paper [and now you] is that you make unsubstantiated statements “with absolute certainty”. E.g.:

    When the warmers complain, you know you are pushing the right buttons

    How do you know that I and Francois and tucker are ‘warmers’. Didn’t I explicitly (#710) state: “That we are not staring disaster from global warming in the face is something I won’t disagree with”?

    NASA is too afraid to make a solar cycle prediction

    How do you know that NASA is afraid? I’m one the NASA Panel, and can tell you that the reason the Panel did not make a ‘consensus’ prediction [in spite of NASA's wish for one] is that the science does not allow us to make a choice at this point between a high and a low prediction, so we acknowledged that and gave two predictions, reflecting the diverse views within the Panel.

    they have gone silent

    If you at all follow this, you might look at Bieseckers presentation at SORCE 2008: Predictions of the Solar Cycle, Past and Present

    has Dr Hoyt abandoned his TSI reconstruction

    That is not the way science works. Old and obsolete papers are not publicly ‘abandoned’, but quietly forgotten. The paper was correct in the context of the data and the knowledge at the time the paper was written.

    The funny thing is that the Hoyt and Schatten reconstruction has a better fit with other data series than Dr Svalgaard’s invariate Sun

    First of all, there is nothing ‘funny’ about anything. Science is serious business [although 'fun' to do]. Maybe the best way to illustrate that you are ill-informed is this:

    The most modern reconstructions are the ones marked “Wang”, “Kriv[ova]“, Preminger, and my own. Note how closely they agree (within half a Watt/sqm) [Wang and Krivova are still hampered a bit by the non-existent 'doubling' of the HMF in the 20th century], compared with the [light grey] Hoyt and Schatten curve that wanders all over the place and agrees with none of the other curves.

    As a further example I can just take one from your paper [there are dozens]. On page 10 you show a graph of the temperature at DeBilt as a function of the solar cycle length. The most critical data point [which mostly determines the correlation] is at a solar cycle length of 16 years, except that there is no cycle with this length [you even show a table on page 25]. And on and on. You do the fight against the AGW people a disservice with such sloppy work [I ordinarily wouldn't use such a word, but it is, sadly, apt].

  714. David Archibald
    Posted Mar 15, 2008 at 11:17 PM | Permalink

    Re 715, Dr Svalgaard, [snip]

    It is now March, 2008 when NASA says it would be solar minimum, and +/- six months doesn’t cut it. That is a margin you could drive a truck through, and is meaningless. Biesecker had nothing to say at SORCE 2008 so he dredged through history to put together a talk. There are some useful things in that talk though that seem to have been forgotten in the last few years:

    Slide 4: Evidence is accumulating that long-term solar variability is an important factor affecting climate.
    Slide 5: Influence of planetary tidal action – solar cycle period related to planetary alignments
    Slide 6: All about rate of change of Sun’s angular momentum about COM of the solar system – the third squiggle down looks pretty predictive.

    Now let’s look at that compilation of TSIs that you provided, specifically at the last five years. All of the others are now well underneath your floor, a floor that you have said means that the Sun doesn’t vary enough to affect climate. But the data that all the TSIs for this period is fresh and not subject to subjective interpretation. How do you explain away this discrepancy?

    I will re-examine that De Bilt graph, but thankfully Butler and Johnson’s graph published in 1996 shows an even better correlation. And we have moved on, to Hanover, New Hampshire. Friis-Christensen and Lassen-type analysis has done the Atlantic crossing to a temperature series near you. A New England beachhead has been established and won’t be dislodged.
    [snip]

    Steve: David archibald, please cut down the histrionics.

  715. David Archibald
    Posted Mar 15, 2008 at 11:24 PM | Permalink

    Galactic Cosmic Ray – Low Level Cloud Correlation Anyone?

    Months ago I promised to do battle against Dr Svalgaard’s invariate Sun using Dye-3 Be10 data. Well it has been worth waiting for. The upper graph is the Dye-3 data with the solar minima plotted on it. There is a good correlation with the Little Ice Age/Modern Warm Period.

    The next graph inverts the Be10 data and plots it against Craig Loehle’s temperature reconstruction. What comes from that is that there are long periods of very good correlation and one long period in which the correlation is less immediately obvious. The correlation from 1424 to 1600 is spectacular. The correlation from 1830 to 1935 is also very good. It is evident that the Be10 explains 110% of the warming of the 20th century. GCRs control the Earth’s climate for hundreds of years at a time. We are currently in one of those periods.

    Be10 does not explain the Maunder Minimum, which means that the Sun has more than one way to cool the Earth. My recommendation to Dr Svalgaard is, instead of hammering the TSI flat, why don’t you tweak yours so that it agrees with other data sets?

  716. David Archibald
    Posted Mar 15, 2008 at 11:36 PM | Permalink

    Why did the graphs turn out black? If you stare hard enough you can make something out of the darkness. If anyone else wants to have a go at posting them I will email them to you. Mine is david.archibald @ westnet.com.au

  717. Jim Arndt
    Posted Mar 16, 2008 at 12:45 AM | Permalink

    Hi,

    Leif, I have been doing a lot of research lately on the effects of magnetism on climate. Many here try to make only one part to fit all. But it is much more than CRF, UV and TSI. You have CME and solar wind. The earths magnetic field effects. I have looked at the USGS Total Magnetic Intensity map and compared it to the RSS MSU (RSS does not measure the poles and GISS only extrapolates them)map with some very interesting results. It seems the there are more anomalies along the more intense fields in the magnetic intensity. For example there is less intensity at the tropics than in North America an it shows up in the RSS MSU. This is also true for Antarctica. More intensity more anomalies. CRF in effected by the earth magnetic field and is further effected by the angle in which the CRF hits the magnetic field. Then you take into account the CME and solar wind, then mix in a little volcanism and you have many factors to change climate that are completely natural. Who knows maybe too much looking at the computer screen. Still work in progress.

  718. cce
    Posted Mar 16, 2008 at 12:52 AM | Permalink

    I have a few questions for Mr. Archibald.

    The RSS satellite temperature analysis in your presentation has a 0.18 degree per decade warming trend for the globe. Do you consider this to be insignificant?

    What criteria did you use to select the 5 rural stations to represent the temperature change for the US?

    Why do you believe that swings in the temperature of Central England are evidence that the world undergoes similar swings?

    What is the source for the chart labeled “Medieval Warm Period – Little Ice Age”. Further, is it consistent with the chart that immediately follows, “The Holocene Optimum.”

    Thanks.

  719. Posted Mar 16, 2008 at 12:58 AM | Permalink

    716 (David A): Ok, I have said it before and will again: “global warming is good for us”. And the temperature now is where was 900 years ago. Does this make me a ‘warmer’?

    Bieseckers slides 4-6 were showing some of the [now] discredited misconceptions from the 1979 workshop. Remember he said to effect “we thought we knew it all back then, but alas…”

    About when minimum is, I agree that it is not now and that we might have to go to the end of the 6 months uncertainty quoted. Although there are right now two cycle 24 spot groups at high latitudes in the Southern Hemisphere. Let’s see if NOAA classifies them as ‘legit’ [they have to last long enough or be seen by at least two observers]

    Now let’s look at that compilation of TSIs that you provided, specifically at the last five years. All of the others are now well underneath your floor, a floor that you have said means that the Sun doesn’t vary enough to affect climate. But the data that all the TSIs for this period is fresh and not subject to subjective interpretation. How do you explain away this discrepancy?

    The people measuring TSI disagree wildly about their results. Maybe best shown by DeWitte’s presentation at SORCE 2008:

    Measured Total Solar Irradiance Cycle Variability: Status at the End of Cycle 23>

    His conclusions were:
    . 4 independent TSI time series are available during cycle 23.
    . All time series agree if standard ageing corrections are used.
    . Best composite = average of 4 series.
    . Good agreement with models assuming no solar minimum TSI variation.

    Maybe his last point is where I would hang my hat.

    why don’t you tweak yours so that it agrees with other data sets?

    Maybe you would do something like this, not me.

    Your comment #717 should be an embarrassment to you.

  720. Raven
    Posted Mar 16, 2008 at 1:00 AM | Permalink

    David Archibald says:

    Re 715, Dr Svalgaard, your throwaway lines are too numerous for you not to be a warmer. And you give the game away when you use a double negative – your English is too good for that, so you should have said,”I agree that global warming is not a disaster.” You used the double negative so that you couldn’t be counted in one camp or the other.

    I don’t see why this is a ‘if your not for me you are against me’ proposition.

    In any case, I also don’t see why you are so anxious to show that the invariant sun hypothesis is wrong. I see the hypothesis as a good thing from the perspective of people looking to demonstrate that the sun has a significant effect on climate since it means TSI cannot be used to explain solar related changes in the past. This makes indirect mechanisms a more plausible explaination for the recent temperature changes. You must remember that warmers can’t explain the warming until 1940 without a variable sun (unless they start fiddling with their aerosol fudge factors).

    Obviously, this means it is more difficult to show that the LIA was primarily due to sun, however, on balance the hypothesis is not bad. Personally, I think Dr Svalgaard’s hypothesis makes sense intuitively because I thought it was strange that TSI would jump around so much over 500 years yet decide to remain constant for the last 50.

    If I was in your position I would look at the temperature record – particularily the recent 8 years since the last max and try to determine if there is some combination of volcanic forcings, solar cycle linked non-TSI forcing and lower CO2 forcing that explains the current record. I am pretty sure that it could be done.

    Here is a link to lucia’s blog where she does some fairly robust statistical analysis that demonstrates that the trend since 2002 is either an 2-sigma statistical fluke or evidence that IPCC missed something. The fact that 2002-2008 coincides with the trailing edge of the solar cycle suggests that the latter is more likely.

    http://rankexploits.com/musings/2008/ipcc-projections-overpredict-recent-warming

    It is also worth noting that almost all skeptics agree that we have/will get a least a 1 degC of warming from CO2 no matter how much we might wish otherwise. If there is a truth to be found I suspect it will come from people who find a way to show that 40-60% of recent warming can be explained by the sun and that CO2 sensitivity has been overestimated.

  721. Posted Mar 16, 2008 at 1:11 AM | Permalink

    719 (Jim A):

    But it is much more than CRF, UV and TSI. You have CME and solar wind.

    back in #657 I showed that the number of magnetic storms [caused by CMEs] has not changed over the last 100 years. In numerous other places I have shown that the solar wind magnetic field has not changed either, so it is hard to attribute climate change to these factors. People that study CRs take the weakening of the Earth’s magnetic field into account when calculating the socalled ‘modulation parameter’.

  722. Posted Mar 16, 2008 at 5:02 AM | Permalink

    722 (Raven):

    Agreed. I am not a “warmer”, but a seeker of plausible theories to explain the warming that I see in the graphs and feel in my everyday life. Science allows us to make a list of these possible causes and one by one, cross them off the list if they do not produce the warming/cooling effect hypothesized. The list associated with solar science is long, as both you and Jim A mention in your recent posts. If Dr Svalgaard is correct about gross TSI being essentially invariate, then that is one possible cause of warming to cross off the list. Now we move onto the componentry of the TSI, and what, if any, link there is between UV, magnetic flux, solar wind, GCR’s etc, etc, etc to the climate change we SEE.

    I fear some (David A) are being too strident by putting a time constraint on good science in hopes of beating back the AGW crowd now. Let’s agree to let science take its customary pace. It may be after cycle 25 before our understanding of the sun is better than it is today. We need to accept that.

  723. David Archibald
    Posted Mar 16, 2008 at 6:11 PM | Permalink

    Re 722, Raven, the warming effect of CO2 is minute. I have done the sums, and it is 0.1 degree C over the last 100 years. The rest is the Sun. Forget aerosols, they don’t last long enough in the atmosphere.

    1.0 degree C is what the Stephan-Boltzman equation says for a doubling. The eminent climatalogists, Lindzen, Kininmonth and others, say that this will convert to 0.6 degrees in real world effect. Roy Spencer has done a recent paper coming to the same conclusion from atmospheric observations.

    The solar cycle length- temperature relationship pioneered by Friis-Christensen and Lassen means that we don’t have to know anything about TSI, Be10 or anything else to predict climate. On one data series, Armagh, Butler and Johnson got a beautiful correlation of 0.6 degrees per year of solar cycle length. I repeated on Hanover, New Hampshire and got 0.7 degrees per year of solar cycle length. That relationship means that the Sun controls climate very tightly. I’ll repeat it again – that relationship means that you don’t have to know or believe anything else about the Sun or climate. It’s game over.

    Re 720, I don’t see the warming trend that you see and measure. What I see is flat over 30 years. The criterion for the selection of rural station was that they had to be 100% non-contaminated by UHI. Thus they turned out to be mostly forestry research places. As it turns out some recent reconstructions of the US here on CA look very similar.

    Re your third question, I will put it to you the other way: Why would swings in world temperature not be reflected locally?

    The source of the graphic is based on an IPCC chart dating from 1990. Yes it is consistent.

  724. Alan Woods
    Posted Mar 16, 2008 at 7:46 PM | Permalink

    David, are you telling us that the science is settled?

  725. Andrew
    Posted Mar 16, 2008 at 8:06 PM | Permalink

    David, Spencer et al found a negative feedback. I dont remember them getting a specific value for the sensitivity. As for global versus regional, its quite well known that different records from different places show different things (ie State of Fear)

    Incidentally, I’m not sure why it matters if CS is 1 or .6, both are less than 2, which the alarmists tend to advocate for (or more!)

  726. Andrew
    Posted Mar 16, 2008 at 8:19 PM | Permalink

    Incidentally, there is nothing malicious about our critiquing your paper. That’s peer review, and trust me, its better to have your mistakes pointed out by your friends than enemies…

  727. David Archibald
    Posted Mar 16, 2008 at 10:41 PM | Permalink

    Re 726-727, climate science is a hobby of mine, my day jobs are in oil exploration and cancer research. It is a hobby that took me to New York though for the Heartland conference. On the way I gave a similar paper at an aviation conference in Hong Kong. It was a pleasure to see the stony faces of the carbon credit people (trying to sell indulgences to the airline industry) during my presentation.

    Yes, the science is settled. On the solar front, the solar cycle length – temperature relationship is seen in half a dozen cities so far. When I get a burst of energy, I will work through the rest of Europe. There are lot of other long term temperature series around the world. Someone with better stats than me will go on to determine error bars and similar. One beautiful thing about the solar cycle length – temperature relationship is that you don’t have to know the mode of action (in cancer science terms) in detail. It is a straight linear relationship that is repeatable over and over again. It is reproducible, therefore it is valid, and incontrovertible. Anybody who wants to get some papers to their name with minimal effort will start copying what I have done and start cranking them out.

    Thus the length of Solar Cycle 23 is very important to human affairs. One thing strange about the Biesecker presentation at SORCE 2008 was the worshipping of Dikpati theory and the ignoring of solar dynamo theory. Why was there felt to be a need to replace something that wasn’t broken?

    Roy Spencer presented at the Tuesday lunch in New York. His figure for doubling (based on his observations of the atmosphere) is 0.6 degrees for doubling. Yes that is a negative feedback from the 1.0 degree from Stefan – Boltzman. What I presented as a bar graph, he presented as a x-y graph with more and better data.

    I got through the temperature – CO2 effect in a couple of graphs. Others have written reams about the overlaps with water vapour, wings etc.

    The Idso work summarises the beneficial effect of increased CO2 on plant growth. So there you have it, all three aspects of the science are settled. Roy Spencer’s work provided the last brick to complete the structure. The strange thing is that the AGW people say they are doing it for the planet, while the science says that the Earth is better off with more CO2 in the atmosphere. It is not a case of precautionary principle and they could be a little bit right, they are completely wrong.

  728. cce
    Posted Mar 16, 2008 at 11:08 PM | Permalink

    Re: 725

    0.18 degrees per decade is the rate of warming of the RSS analysis from 1979 to the present. That is independent of what any of us “see”.

    You have used 5 temperature stations from the Southeast United States to infer the temperature change of the entire United States.

    Local temperatures are more variable than global temperatures.

    The graph from the First Assessment Report is Hubert Lamb’s temperature reconstruction of Central England. It was mislabled “global temperature” in the report. It uses growing seasons to infer temperature before thermometers.

    You have mislabled the scale. According to Lamb’s graph, the difference between the LIA and MWP in Central England was about 1.3 degrees (based on 50 year averages), not 4 degrees. A 4 degree temperature change is clearly incompatible with the “Holocene Optimum” graph, even given the coarse resolution.

  729. David Archibald
    Posted Mar 17, 2008 at 1:03 AM | Permalink

    snip

  730. cce
    Posted Mar 17, 2008 at 2:26 AM | Permalink

    re 731:

    Since you like drawing straight lines between endpoints, 1979 had an anomaly of -0.14 degrees. 2007 had an anomaly of 0.30 degrees. That is an increase of .44 degrees.

    But February’s anomaly was -0.007 degrees? What will we do? Waitaminute. January had an anomaly of -0.080 degrees. That’s an astonishing 0.073 degree increase for the month! That’s 0.88 degrees per year or an amazing 8.8 degrees per decade!

    The US hasn’t warmed in 72 years? Well, maybe all of our warmth went to Central England. That seems to be where all the action is.

    Your whole “paper” is one big “gotcha” for anyone who takes it seriously.

  731. David Archibald
    Posted Mar 17, 2008 at 3:54 AM | Permalink

    Re 732, sorry CCE, I have bigger fish to fry. I have been alerted to this paper in Geophysical Research Abstracts: http://www.cosis.net/abstracts/EGU2008/11765/EGU2008-A-11765.pdf?PHPSESSID=

    It is entitled “Does sunspot number calibration by the “magnetic
    needle” make sense?”
    K. Mursula (1), I. Usoskin (2), and O. Yakovchouk (3)

    I am told that it shows that Dr Svalgaard’s work is less than completely wonderful. A correspondent points out further: “Svalgaard claims the sunspot numbers for about 1880-1890 are underestimated by 58%. Spoerer, one of the most careful observers ever, got the same number of groups, within 10%, as the Royal Greenwich Observatory, and both of these results agree with Wolf’s reconstruction for this cycle. It is impossible for these observers to failed to have seen nearly 40% of the sunspot groups.”

    The solar physics community is reacting to Dr Svalgaard’s ripple of peculiar science in the literature. They have displayed more gumption than the climate science community in the face of Dr Hansen’s fiddling with historic temperature data.

  732. Posted Mar 17, 2008 at 6:30 AM | Permalink

    733 (David A):
    Here is my ‘rebuttal’ abstract for the upcoming AGU [American Geophysical Union] meeting in Florida in May:
    SP01: Sunspot Number Calibration by the “Magnetic Needle” Makes Sense: We show that the amplitude (dD) of the diurnal variation of the magnetic Declination is a reliable indicator of solar far ultraviolet radiation (FUV) and its proxy, the sunspot number, R [as was known already to Rudolf Wolf ~160 years ago]. FUV creates and maintains the E-layer of the ionosphere, determining the strength of the diurnal variation of the Declination. We show how the changes of sunspot number observers are faithfully reflected in discontinuities in the relationship between dD and R. Comparisons with other sunspot indices bear this out in a clear manner. On the whole, sunspot numbers before 1947 should be adjusted upwards by 20%, and before Wolf’s death by another 30%, with the net result that 20th century solar activity does not seem significantly larger than that of the 19th.

    Saying that the sunspot is wrong is serious business and will attract protesters. [By the way, dD and rY are equivalent for this discussion as one can be calculated from the other].
    As I showed in my talk at AGU Spring 2007 Wolf himself used the ‘magnetic needle’ to adjust the earlier [before 1849] sunspot count to match his own by 24%. More details can be found here. One of Wolf’s successors at Zurich, Waldmeier, around 1947 introduced a further ~20% jump on top of that as explained here.

    This is ongoing research and it will take some time before the dust settles. The situation is, in many ways, similar to what happened to my result that the geomagnetic aa-index was to small before 1957. First a vitriolic denial, then eventually followed by acceptance to the point that my detractors claim they themselves discovered the discrepancy. Same thing happened with my result that the Sun’s magnetic field did not more than double in the last 100 years as was previously claimed.

    The particular objection in the Mursula et al. abstract that “rY values do not actually imply that the observed Rz values in the 19th century are systematically underestimated” is disingenuous because Rz is the Wolf number already adjusted using the ‘magnetic needle’. My real issue is not the Wolf number, but the [Hoyt & Schatten] Group sunspot number, Rg, which is much lower than Rz before the mid 1800s. It is the Rg that is commonly used in TSI reconstructions.

    Without overburdening the discussion [although I would welcome a thorough analysis of the whole subject with y'all as referees] with detail, I’ll just consider the following [easy] case.

    I work with 11 stations before 1849 [the early group] and 8 after [the late group], but all in the 19th century. For all these we have good and reliable [this is another discussion point] geomagnetic data for several years. For each group of years [to increase the robustness] we form the average sunspot number <R> and the average range of the diurnal variation <rY>, then plot these averages for the different stations and years:

    The things to note are:
    1: The Wolf numbers (Rz) plotted as the blue diamonds have a good relationship with rY. This is not surprising since Wolf used rY [or rather dD] to secure this calibration of Rz.
    2: The Group numbers (Rg) plotted as pink dots after 1880 show the same good relation as Rz. This is not surprising either since Hoyt and Schatten constructed Rg to have the same mean as Rz after 1874.
    3: The Group numbers (Rg) for the early stations before 1849, plotted as filled red diamonds fall consistently below the pink dots. This is the important point. To have the early group numbers match the relationship of the late group of stations [sorry for the multiple use of 'group' with two meanings] we need to multiply them by 1.4. If we do that and plot the result [red open diamonds] we get a good match. So, Rg is too low by a factor of 1.4 before 1849 compared to Rz.

    An important point that is often missed is that both Rz and Rg involve ‘calibration’ factors that change from observer to observer. These factors are hard to come by for old data. Just as for the temperature record, the ‘adjustments’ are often larger than the real changes.

  733. Posted Mar 17, 2008 at 7:16 AM | Permalink

    734 (me): “As I showed in my talk at AGU Spring 2007 Wolf himself used the ‘magnetic needle’ to adjust the earlier [before 1849] sunspot count to match his own by 24%. More details can be found here.” one is good.

    And, David, you need to watch your language ["idio***", sarcasm: "less than wonderful", "worshipping", etc]. And, since people did direct several specific questions to you, you don’t duck by ‘going after bigger fish’. Their opinions and issues are important too.

  734. Posted Mar 17, 2008 at 7:20 AM | Permalink

    735 (me): grr, this one: (Calibrating Sunspot Numbers Using the Magnetic Needle). Sorry folks, I’m a bit fat-fingered this morning.

  735. Steve McIntyre
    Posted Mar 17, 2008 at 7:43 AM | Permalink

    #735. David Archibald, I’ve already asked you previously to cut out the histrionics. This is very tiresome that you should have to be asked again. If you want to post here, please stop the immaturity.

  736. David Archibald
    Posted Mar 17, 2008 at 7:47 AM | Permalink

    The clouds obscuring the graphs in 717 have cleared just in time. Experts have examined the Be10 – temperature relationship therein and pronounced as follows: “The Be10 data confirms that solar activity was lower in the 1800s than now, confirming the Hoyt and Schatten reconstruction and inconsistent with Svalgaard’s claims.”

    The Be10 data is hard data that can’t be tweaked and stroked to bend it to your will. It is hard data best consumed raw. No amount of volcanic wishing and hoping and dreaming will change the shape of that curve. It is the hard rock on which inaccurate TSI reconstructions will founder. The best temperature reconstruction going, Dr Loehle’s, agrees with the Be10 data for great chunks of time, especially in the critical Little Ice Age to Modern Warm Period transition. Two high quality data series agree with each other. What need is there for a convoluted story that tells us that we see with our own eyes is wrong?

  737. David Archibald
    Posted Mar 17, 2008 at 8:02 AM | Permalink

    Before the witchdoctors started running various NASA departments, NASA did some good science. We know it is good because they speak of a very high solar-temperature correlation. I refer to “Newly Found Evidence of Sun-Climate Relationships” by Kim and Huang which can be found deep in the NASA archives at: http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930017826_1993017826.pdf

    The abstract says in part:

    “Surprisingly the annual SAT responds to the solar total irradiance variations with a
    correlation coefficient as high as O.78. As yearly solar irradiance variations from 1978 to
    1990 were overlain on the “SAT box grid” ( Hansen, et aL, 1987), geographic patterns of
    Sun-climate correlation emerge which display a meridional component apparently driven
    by “the conditions of cloud coverage” and “the effectiveness of heat transport processes of
    the oceans”.
    The last one hundred years history of global SAT also match a solar irradiance
    model based on solar proxy data. A correlation coefficient of 0.82 was derived with
    appropriate parameterization of temperature response in long-term trends. Also, it has
    been observed that the derived correlation coefficient can typically peak when a built-in
    phase lag of 32-40 months is instituted in temperature response.”

    The paper speaks of a three year lag in response to solar activity, which means temperatures will keep going down to 2012 from my projected July 2009 solar minimum. That is plenty of time to sort the AGW believers out. And the beautiful thing is TSI is already under the lows of previous minima and we are likely to be still a year of minimum.


    Steve:
    David, please stop using terms like “believers” at this blog. I’m getting tired of your language. I haven’t personally studied the solar-climate correlations although it’s something that I’d like to do. Some scientists believe that the statistical work in such studies is inadequate and criticize me for not applying equal to the deconstruction of such data. All this data needs to be scrutinized just as carefully as bristlecones before making any triumphal announcements.

  738. Posted Mar 17, 2008 at 8:46 AM | Permalink

    738 (DavidA):

    Experts have examined the Be10

    Un-named experts don’t do much for me.

    Dr Loehle’s, agrees with the Be10 data for great chunks of time, especially in the critical Little Ice Age to Modern Warm Period transition

    Here is your unclouded graph:

    According to your own [I presume] annotation at the bottom, the correlation is poor half of the time. During the ‘critical’ transition where correlation was supposed to be good, the rounded rectangles highlight some clear mismatches. Looks poor to me. Following your own advice: “What need is there for a convoluted story that tells us that we see with our own eyes is wrong?” I refrain from a statistical analysis of just how poor it is.

    Concerning the long-term trend of the 10Be, consider:

    The 10Be curve within the red rectangle shows that there is not much systematic trend. Most of the curve falls within the rectangle. A few spikes above show higher activity [lower count] and a few spikes below show lower activity. Most of them [as you point out: "poor correlation"] unrelated to temperature. And some of them occurring at the same time as major volcanic eruptions, e.g. Hekla near 1700, Tambora and Mayon near 1815 [when, BTW, the temperature doesn't seem to have a dent - in spite of the year without a summer], and Krakatoa in 1883.

    What have volcanoes to do with 10Be? They spew out sulfuric aerosols that helps wash 10Be out of the atmosphere. There is some debate about the time scale [a few years vs. a decade], but that is normal in science.

  739. Posted Mar 17, 2008 at 8:51 AM | Permalink

    739 (David A):

    And the beautiful thing is TSI is already under the lows of previous minima

    You seem to ignore comment #721. Let me refresh your memory:

    His conclusions were:
    . 4 independent TSI time series are available during cycle 23.
    . All time series agree if standard ageing corrections are used.
    . Best composite = average of 4 series.
    . Good agreement with models assuming no solar minimum TSI variation.

  740. Andrew
    Posted Mar 17, 2008 at 9:19 AM | Permalink

    Leif, the year without a summer is smoothed out, becuase Craig used a 30 year smooth to allow for dating error. The dating isn’t perfect, so searching for such events may be a folly.

  741. Posted Mar 17, 2008 at 9:22 AM | Permalink

    739 (David A):

    Before the witchdoctors started running various NASA departments, NASA did some good science. We know it is good because they speak of a very high solar-temperature correlation. I refer to “Newly Found Evidence of Sun-Climate Relationships” by Kim and Huang

    The tone is not good here, neither is the science. Here is Figure 3 from said paper:

    The TSI-reconstruction is totally at variance with what we know today. The first many cycles do not follow the sunspot number [1957 was the biggest cycle max ever, and 1906 one of the smallest]. The last two cycles are simply calibrated wrongly. The RGO data stopped in 1975, so no wonder that the cycles after that look different. Compare this TSI reconstruction with almost any out there, even the Hoyt & Schatten one that you like and you’ll see how way off the mark this paper is.

  742. Posted Mar 17, 2008 at 9:29 AM | Permalink

    742 (Andrew): Thanks for that info. So Craig smoothed away the whole Dalton minimum, and got instead a clear local maximum from 1800 to 1824, at least according to David’s graph. So much for the theory that low solar activity produces cooling. Or am I reading this wrong? To each his own smoothing and cherry picking, it seems :-)

  743. Posted Mar 17, 2008 at 10:50 AM | Permalink

    739 (Steve):

    Steve said: I haven’t personally studied the solar-climate correlations although it’s something that I’d like to do. Some scientists believe that the statistical work in such studies is inadequate and criticize me for not applying equal to the deconstruction of such data. All this data needs to be scrutinized just as carefully as bristlecones before making any triumphal announcements.

    The reason for this inadequacy is, I believe, that most of such work is on the fringes of both solar and atmospheric science. The intersection of two fringes holds little significance and interest from either community, so nobody cares to really examine the claims. There are a few shining counterexamples. The best one I know of is by another Canadian, Colin Hines, analyzing the significance of a paper of which I was a coauthor. His analysis is here

    It could have been written by you! One problem with this kind of work is that most people’s eyes glaze over on about the second page.

    The fate of that original sun-weather effect discussed by Hines can be seen here. Still considered fringe-science by most people.

    P.S. Another coauthor was Walter Orr Roberts (1915-1990) who was the founding president of the University Corporation for Atmospheric Research (UCAR) and the first director of the National Center for Atmospheric Research (NCAR). I had the pleasure of sharing office with Walter at NCAR [in the little 'annex' off the parking lot to the left] in about 1974 for several months while grinding my way through the hundreds of magnetic tapes with atmospheric data using NCAR’s newest supercomputer [a CDC 7600; about a tenth as fast as my PC...].

Follow

Get every new post delivered to your Inbox.

Join 3,295 other followers