Svalgaard #4

Continued from here.

571 Comments

  1. Andrew
    Posted Mar 17, 2008 at 12:52 PM | Permalink

    Leif, I wouldn’t quite describe it as “smoothing away” the whole Dalton, that makes it sound bad. It was necessary to do the smooth, because excluding trees makes it difficult to say whether individual peaks and troughs really happened at the same time. Some detail is lost. But its better than dealing with precipitation issues etc. related to trees.

  2. Posted Mar 17, 2008 at 12:59 PM | Permalink

    1 (Andrew): It was not meant to be bad. But for people that seek support for the supposition that low solar activity as we had for 40 years at the Dalton minimum would automatically result in lower temperatures it might look bad that there is no hint of a dip, especially if they use Craig’s curve as support for their theory.

  3. cba
    Posted Mar 17, 2008 at 2:30 PM | Permalink

    Leif,
    That paper was a snoozer(Hines & Halevy). I could only skim it. Still wound up snoring for close to an hour after page 2. The appendix wasn’t much better.

    I had some questions from recent posts and readings.

    1. If Wolf used the needle to adjust his SSN values upward, the n why should there be an additional correction added?

    2. Why is the SORCE data so low if it is supposed to be so accurate? It’s somewhere around 4.9 W/m^2 below the ACRIM data which corresponds to a value that is effectively the BB emission of the photosphere at 1 AU?

    3. What is the method used to extract the orbital position modulation of the TOA from the data, used either for SORCE or for ACRIM (which is the one I need since SORCE already has it)? Is it just a daily orbital position correction from 1AU? That’s nice for analyzing the sun but it’s a royal pain trying to deal with that when concerning actual TOA incident radiation.

  4. cba
    Posted Mar 17, 2008 at 3:09 PM | Permalink

    BTW,

    I heard consensus was reached for a low group and high group, where SS 24 will be 90 low up to 139.9 high and both groups agree it started in Feb 2008.

  5. Posted Mar 17, 2008 at 5:33 PM | Permalink

    3,4 (cba):

    1: Wolf and SSN: Wolf used the amplitude of the Declination, dD, (which varies from place to place) instead of the East component, rY, which does not vary. The formula is (approx): rY = H dD / 3438 where H is the total horizontal component, dD is measured in minutes of arc, so if H = 20,000 nT (typical for Europe) and dD is 10 minutes of arc, rY = 20,000 * 10 / 3438 = 58 nT. H varies from 0 at the magnetic poles to 36,000 at the Equator. Lastly, Waldmeier screwed up the record in 1947 causing another ~20% adjustment.

    2: SORCE and the others are NOT accurate as far as absolute calibration (in Watts/sqm) is concerned. We cannot do that better than 5-10 Watts. Relative accuracy from days to day is very accurate. The stability over time for all the other instruments [except SORCE] is not so good. The detectors ‘age’ and the big problem is the quantify that ageing. For SORCE, they compare the Sun to a set of non-varying stars, so that even a 1000 years from now they can check against the same stars.

    3: Astronomers have known for 400 years how to calculate the distance to the Sun. I can give you a formula [long and complicated, depending on accuracy needed], but you can also find it the internet. Just the distance in AU [unless you worry excessively about accuracy - 1 AU is NOT the correct, real distance semi-axis, but close enough]

    4: I don’t know what you have heard. I participated in the last telecon about a week ago and nothing has changed: Minimum in March 2008 +/-6 months; maximum either 90 or 140 depending on which camp you are in. The 139.9 is excessively precise, but may be a typo on your part.

  6. Posted Mar 17, 2008 at 5:37 PM | Permalink

    3 (cba):

    That paper was a snoozer(Hines & Halevy). I could only skim it. Still wound up snoring for close to an hour after page 2. The appendix wasn’t much better.

    I warned you! but I think that any such paper will be boring in the extreme, even Steve McI’s. Just look at his Georgia Tech presentation. That is the nature of the beast and there is no ‘royal road’ to this.

  7. cba
    Posted Mar 17, 2008 at 5:50 PM | Permalink

    hi Leif,

    http://www.swpc.noaa.gov/ftpdir/weekly/Predict_high.txt

    http://www.swpc.noaa.gov/ftpdir/weekly/Predict_low.txt

    are what i’m referring to – for whatever it might be worth.

  8. Posted Mar 17, 2008 at 6:25 PM | Permalink

    7 (cba): At least they now have a low and a high file. Their first inclination was to have only one file with the average of low and high, on the argument that many users have automated systems that cannot handle two files, but was looking at …/weekly/Predict.txt

    But wait, they do have the …/weekly/Predict.txt file! It contains this hilarious disclaimer:

    # Prediction values are based on an average of the ISES panel cycle 24 forecasts
    # of 13-month running smoothed values. The panel does not consider this to be a
    # correct interpretation of their predictions.

    BTW, their two files are not consistent with the graphs at SWPC, especially in 2008. I have pointed this out to them, but, apparently, they cannot change the tables nor the graph, except once a year…

  9. cba
    Posted Mar 17, 2008 at 8:12 PM | Permalink

    I think it’s worse than that. There was something about issue 1697 with new numbers in a table that evidently no one noticed when it was released that the numbers were new.

    I eyeballed the SORCE data compared to the ACRIM on about the same scale. They match fairly close but it looks like SORCE doesn’t vary quite as much in the jiggles as we get to later data. I’m tempted to take a difference and just see what the result is (other than an average of 4.9 W/m^2 higher for the ACRIM. Somehow, absolute accuracy of 0.5% is not really impressing me much. Supposedly the ACRIM stuff (at least most of the instruments involved) had calibration capability that could compensate for instrument degradation.

  10. Dennis Wingo
    Posted Mar 17, 2008 at 8:31 PM | Permalink

    Leif (#5)

    4: I don’t know what you have heard. I participated in the last telecon about a week ago and nothing has changed: Minimum in March 2008 +/-6 months; maximum either 90 or 140 depending on which camp you are in. The 139.9 is excessively precise, but may be a typo on your part.

    I don’t understand this. If the telecon was last week and the minimum is +/- March 2008 six months, are we not already past the – part? In looking at the daily data there is no uptick of CY-24 spots yet and the number and magnitude of CY-23 spots continues to decline.

    What am I missing here?

    (though it is interesting that the 10.7 cm flux remains fairly high for the number of spots, supporting some of your papers of late regarding the minimum during quiet times).

  11. Posted Mar 17, 2008 at 9:31 PM | Permalink

    9 (cba): issue 1697 ??? what is that?
    About the TSI, I can only refer you #721 of Svalgaard #3. The TSI-groups still struggle mightily with all this, with no end in sight.

    10 (Dennis):

    If the telecon was last week and the minimum is +/- March 2008 six months, are we not already past the – part? In looking at the daily data there is no uptick of CY-24 spots yet and the number and magnitude of CY-23 spots continues to decline.

    If tomorrow and the next few months we see enormous solar activity then the smoothed minimum will be pushed back to, say, October of last year. That is the – part. Personally I think minimum will be closer to the extreme _ part, but I’m only one on the panel [and several people had earlier expected minimum on 2006 or earlier 2007, and there is this 'I stand by my data/forecast/voodoo/whatever' syndrome we all have to live with].

    F10.7 still has a way to go, and TSI [SORCE] still shows a downward trend, and the corona is still not ‘flat’, so we are not here yet. On the other hand, cosmic rays may have peaked, as the expected shape of the peak should be sharp [alternating every other cycle: sharp, flat, sharp, flat, sharp, ...]:

  12. Posted Mar 17, 2008 at 9:36 PM | Permalink

    11 (me): “extreme + part”, of course.

  13. Dennis Wingo
    Posted Mar 17, 2008 at 9:53 PM | Permalink

    (#11)

    Leif

    Ok, I get that. So in your scenario, the CY 24 sunspots are going to be fairly few and far in between for the rest of 2008 in order to keep the smoothed min centered around March 08.

    As far as TSI goes, that is a mystery to me why they cannot calibrate better but I guess that the problem is that the sensors degrade over time and it is really hard to get a good sensor calibration that holds up under a varying radiation environment.

    It is funny that we use 1358 as our median number in the space solar power design world.

  14. Posted Mar 17, 2008 at 10:03 PM | Permalink

    13 (Dennis): The SORCE calibration of TSI is really neat: observing the same set of non-varying stars from time to time. As I said earlier, a 1000 years from now we can check the calibration.

    And, yep, the SC 24 spots will have to be feeble for some time.

  15. maksimovich
    Posted Mar 17, 2008 at 11:37 PM | Permalink

    The forecast of the CR behavior based on the predicted values of the global characteristics of the solar magnetic field, thick line – data of CR intensity observations (Moscow NM), thin line – the predicted CR variation up to 2013 based on data of Kitt Peak Observatory (upper panel) and based on data of Stanford Observatory (bottom panel).

    Belov et al 2004

  16. maksimovich
    Posted Mar 17, 2008 at 11:59 PM | Permalink

    Regarding the absence of a climatic anomaly with regard to Krakatoa in the temperature series used by Archibald previous thread,Stenchikov et al in thier review of Volcanics for the AR4 GCM had this intersting comment on the “ensembles”.

    The observed value no doubt has substantial temporal sampling errors, and the observations for the earliest eruptions considered may suffer from inadequate geographical sampling or other data quality issues. It is noteworthy in this respect that the global-mean surface temperature record used in this study surprisingly shows no global cooling following the very large 1883 Krakatau eruption [Jones et al., 2003]. However, the surface air temperature reconstructed by Hansen and Lebedeff [1988] shows a more sizable cooling effect of Krakatau, which indicates the level of uncertainty in the observations themselves, especially for earlier volcanic events.

  17. cce
    Posted Mar 18, 2008 at 12:40 AM | Permalink

    Leif,

    You’ve probably been answered this, and if so just link to your past comments, but what is your take on the cosmic ray theory of climate change?

    I linked these two papers on RealClimate but there wasn’t much discussion. There’s been a lot of empahsis on the cosmic rays themselves, but people seem to have forgotten about the clouds.

    Arguments against a physical long-term trend in global ISCCP cloud
    amounts

    http://cimss.ssec.wisc.edu/clavr/amato/2006GL028083.pdf

    Observed interdecadal changes in cloudiness: real or spurious?

    http://meteora.ucsd.edu/~jnorris/reprints/NorrisGwattRevised.pdf

    Somewhat related, I also see Lockwood has two new papers (money-walled)

    http://journals.royalsociety.org/content/w8221618164412nt/

    http://journals.royalsociety.org/content/e810603423v14t37/

  18. Pete
    Posted Mar 18, 2008 at 4:50 AM | Permalink

    #17 cce RE: The new Lockwood papers. (one is joint with Frohlich)

    Thanks for pointing this out. These are obviously a follow-up to their Paper from last year, which as you know was published FREE and amongst much publicity! I can’t read these as I’m not paying £27 ($49) for each of them!

    I’m sure Mike Lockwood is a very respected solar scientist. From what I have read of the abstracts he certainly now fully states his opinion that PMOD is more reliable than ACRIM. I’m sure Prof. Willson will have something to say on this! This is a continuing debate however. He says they explore why this is in the paper we can’t read.

    Now onto the temperature comparisons. Firstly, it disturbs me from the abstract that he states there is declining Solar activity from 1987 for 20 years, and over the same period surface temperatures have been streadily rising. This is just wrong. There has certainly been a rise in temps up to 1998. But in reality we have 1 Decade of increase and then 1 Decade of flat temperatures. To imply that we have a continuing declining Solar Activity and at the same time a rising temperature, and they are in linear opposition continually across the whole period opens up some disbelief why he would say this.

    Now, over recent weeks I’ve been studying the 1991 Pinatuba eruption, well within the timeframe of this Study (1987 to 2007), or 20 Years. (Too short for a Solar study anyway surely….!) It is the overwhelming opinion of Science that this well studied eruption reduced Global Temperature by 0.4C to 0.5C for a 2-3 years period. You’d be surprised the effect this has on the trend from 1987 to 2007 shown in this study.

    Below I present 4 graphs. The first one is the pure Hadley temperature 1987 to 2008. This has an overall 21 year increase of +0.42C, or the well known and published +0.2C / decade.

    Now, in my 2nd graph I have removed Pinatubo effects by replacing a progressive adjustment upwards from 1992 to 1994 starting at +0.3 and reducing to 0 by 1995. Actually I was being very safe by not using the higher 0.4-0.5 published values. As shown in the 2nd graph this actually reduced the upward trend 1987 to 2008 to +0.215C, or +0.1C / Decade. A halfing of the Warming actually produced by non-volacanic forcings.

    So, the Volcano actually forced down, in an artificial manner, the trend line at the start of the 20 year period in question here. I’m happy this is accurate as I restored the ‘lost’ El Nino Peak of 1992/1993 which is now clearly seen.

    Now to the bottom two graphs which cover Solar Cycles 22 and 23. The pure unadjusted Hadley trend is -0.0003C for Cycle 22, so just noise really, no change. Cycle 23 is +0.009C, so again what can we say? Minimum to Minimum more or less no change, whether you use ACRIM or PMOD it lies in the middle of the two.

    May be this is the truth.

    All in all, it seems to me that over the past 20 Years there was a Step change in temperature in the lead-in to the 1998 El Nino, and any other forcing was too small to significantly effect the global temperature.

  19. Andrew
    Posted Mar 18, 2008 at 5:12 AM | Permalink

    Doug Hoyt stated just the opposite in regards to which is better, PMOD or ACRIM:

    Douglas Hoyt Says:
    August 12th, 2007 at 1:18 pm

    The PMOD solar irradiance reconstruction is definitely incorrect. It is based upon an assumption that the Nimbus-7 ERB radiometer suddenly got more sensitive towards the end of its life and based upon that assumption they adjust the solar irradiance values down. As someone who spent more than 20 yeats designing, building, and analyzing radiometers and more than 10 years on the N-7 radiometer, I say unequivocally that the PMOD adjustment is physically impossible and erroneous. The Willson irradiance is correct. The Lockwood paper is nonsense.

    http://www.warwickhughes.com/blog/?p=131#comments

    I remember thinking at the time that if they ignored that data, they were being quite sloppy. But then, I doubt advancing our understanding of the sun/climate link was the point. Did they give a reason for doing so?

  20. Posted Mar 18, 2008 at 5:27 AM | Permalink

    17 (cce): The GCRs and clouds: the observational evidence is weak, mainly because the data is of dubious quality, c.f. the papers you cited.

  21. Posted Mar 18, 2008 at 5:56 AM | Permalink

    18,19 (pete,Andrew): In my book climate is a 30-year thing. So making statements as many people have based on a few months worth of data (cold 2008), or a decade’s worth of data does not make much sense. My comment #721 still summarizes my opinion about TSI: no trend has been demonstrated. Lockwood: he has been wrong before ['doubling of Sun;s magnetic field'] and may be again when it comes to detailed explanations based on preferred data. Time will tell.

  22. TrueSceptic
    Posted Mar 18, 2008 at 6:37 AM | Permalink

    Dr Svalgaard,

    As you can imagine, your threads here have generated much interest wherever people discuss climate change.

    Can I ask you whether you consider the following to summarise correctly your views?

    But what he is going to say about the sun (he repeats this over and over) is that either it isn’t the big driver, or else climate senstivity to solar is very high. Which to me seems really obvious, for example, at the minimum solar has got to be 2x any greenhouse effect (including water vapor).

  23. David Archibald
    Posted Mar 18, 2008 at 6:41 AM | Permalink

    Re 17, Lockwood is out and proud as a messianic warmer, and these days is programmed to write papers disproving the solar – climate. He outed himself in a letter to the Telegraph newspaper: “Climate change is by far the greatest threat to everyone’s standard of living. Unlike political parties, companies, media stars, works of art, consumer products and even social trends and national economies, a scientific reality is immune to spin.
    (Prof) Mike Lockwood, Southampton University ” As such, anything he publishes these days is highly suspect.

  24. Posted Mar 18, 2008 at 7:07 AM | Permalink

    23 (David A): as always, watch your language ["messianic"].

    As such, anything he publishes these days is highly suspect.

    A person’s private views have no bearing on the science he produces. It’s like: “he’s a Christian, so all his works on cosmology are suspect”. You write a paper, it goes to peer review, if accepted, it is [perhaps] read by the scientific [sub]community, if sound, it is built upon and remembered, otherwise it is just ignored. This whole process basically works, although not always well. But in the end, the good stuff endures.

    22 (TrueSceptic):

    Can I ask you whether you consider the following to summarise correctly your views?

    But what he is going to say about the sun (he repeats this over and over) is that either it isn’t the big driver, or else climate sensitivity to solar is very high.

    so far, so good, the following:

    Which to me seems really obvious, for example, at the minimum solar has got to be 2x any greenhouse effect (including water vapor).

    is your own opinion.

    On the whole, I’m not trying to spread a message or to be a protagonist for any particular ‘view’. I have studied the long-time behavior of Sun for a long time, and believe that there is interest in my findings, which I’m happy to share. I’m not always correct [as opposed to some people], e.g. when I thought [in 1977] that the Sun’s magnetic field had doubled, but, at least, I admit my mistakes [or rather, mainstream scientists just ignore them]. I find joy in the journey, not in being right. If some time-traveler from the future would give me all the answers, I would not want to know. If, however, some time-traveler from the past would tell me something…

  25. Andrew
    Posted Mar 18, 2008 at 7:18 AM | Permalink

    My opinion on Lockwood is that, given how many people arguing for the sun’s role reference his work on the “doubling” of the open solar flux, he had a crisis of conscience, and rather speedly put together a paper trying to establish that the sun couldn’t explain recent warming. I think it was rather sloppy, if he has any understanding of climate/weather, becuase everyone knows that it isn’t hotest when the sun is highest in the sky, its hotest two hours later, or even if he doesn’t, apparently ignoring some relevent data, or at least downplaying it. If (I stress if) ACRIM is more accurate, that renders the entire paper meaningless.

    I’m not saying he did anything wrong intentionally, of course, but I think he allowed his emotions to rule him a bit.

  26. Posted Mar 18, 2008 at 7:22 AM | Permalink

    25 (Andrew):

    I think he allowed his [Lockwood's] emotions to rule him a bit.

    Some people are like that; not much to do about it…

  27. Posted Mar 18, 2008 at 8:00 AM | Permalink

    22 (TrueSceptic): The problem I’m having [over and over :-) ] is that some people have things backwards, their argument goes something like this: “Since the LIA and the MWP are obviously caused by the Sun [what else?], then obviously, the Sun must vary accordingly”. This is putting the cart before the horse, IMHO.
    That the Sun and Weather/Climate must be related is so ingrained that it is hard to unshackle those chains. I’m holding in my hand volume XVI of Encycl. Brittanica 9th edition [~1895]. Geomagnetic activity and its relationship to solar activity is discussed in the entry “Meteorology”. On page 181, one reads that “an increase of diurnal temperature range most probably denotes an increase of solar energy, and we are thus led to associate an increase of solar heat with a large development of spots. This, however, is a point which requires further investigation”.

    Now 110 years later, we don’t seem to be much closer…

  28. TrueSceptic
    Posted Mar 18, 2008 at 8:03 AM | Permalink

    Re 24:

    is your own opinion.

    No, not mine. I was quoting someone else who thought he knew your mind better than you did!

    I am pleased to see this

    On the whole, I’m not trying to spread a message or to be a protagonist for any particular ‘view’. I have studied the long-time behavior of Sun for a long time, and believe that there is interest in my findings, which I’m happy to share. I’m not always correct [as opposed to some people], e.g. when I thought [in 1977] that the Sun’s magnetic field had doubled, but, at least, I admit my mistakes [or rather, mainstream scientists just ignore them]. I find joy in the journey, not in being right. If some time-traveler from the future would give me all the answers, I would not want to know. If, however, some time-traveler from the past would tell me something…

    This is how it should be.

  29. TrueSceptic
    Posted Mar 18, 2008 at 8:06 AM | Permalink

    Re 26: It doesn’t necessarily make someone wrong, though, does it?

  30. Posted Mar 18, 2008 at 8:12 AM | Permalink

    29 (TrueSceptic): no it doesn’t, and I broke my own rule of not commenting on people’s character or motivation. Maybe let my emotions get the better of me :-)

  31. TrueSceptic
    Posted Mar 18, 2008 at 8:13 AM | Permalink

    Re 27: It is so refreshing to read posts such as this. You have no agenda to prove it “must” be the Sun, or indeed to prove that it “can’t be”: you just tell it how it is.

    Ever since I’ve been reading up on climate change I’ve wondered why the Sun is such a favourite with those in the GW “sceptic” camp. The variations are so small and the recent trend is in contradiction to their claims (I’m talking 30 years, not 9, of course!).

  32. Andrew
    Posted Mar 18, 2008 at 8:23 AM | Permalink

    I think that internal variability, that is, trends which arise by chance can’t be ruled out as a possible alternative explanation. There’s also volcanoes:

    http://icecap.us/images/uploads/VOLCANOESANDCLIM.doc

    Which has always been the main competing explanation for the LIA, I think.

    To me, it is not so much the TSI variability as it is things closely associated with it. I think (but of course I can’t prove it) that strong secular trends in Cosmic Rays can emerge in times of general activity/in activity of the sun independently of any trends in TSI. It might seem ad hoc, but then again, if you look at Cosmic Rays and Sunspots, the only real similarity is the periodicity. The amplitudes vary independently of one another.

  33. Posted Mar 18, 2008 at 8:35 AM | Permalink

    32 (TrueSceptic):

    if you look at Cosmic Rays and Sunspots, the only real similarity is the periodicity. The amplitudes vary independently of one another.

    If you look at the plot in #15, you’ll see that the CGRs follow closely [albeit - and correctly - through a model] the ‘general magnetic field’ of the Sun. This field does follow the sunspot number. As always, the connections are subtle, but reasonably well understood. The CR modulation is also influenced by the geometry of the Heliomagnetic field. This is something I’ll comment on when I find the time…

  34. Posted Mar 18, 2008 at 8:37 AM | Permalink

    33 (me): See also the passable [inverse] correlation between SSN and CRs in #11.

  35. TrueSceptic
    Posted Mar 18, 2008 at 8:39 AM | Permalink

    Re 33: That was Andrew, not me. :)

  36. jack mosevich
    Posted Mar 18, 2008 at 8:39 AM | Permalink

    SO: what DID cause LIA and MWP? Maybe 2 different phenomena?
    Or maybe they didn’t really happen globally?

    Lief: the reason so many people look to the Sun to explain these phenomena
    is probably that it is the sole known source of energy.

  37. TrueSceptic
    Posted Mar 18, 2008 at 8:46 AM | Permalink

    Re 34: Is this why the Svensmark CR hypothesis is so popular with some: there has been no upward trend in recent TSI or sunspot frequency so something else solar-related is wheeled out to explain recent warming?

  38. TrueSceptic
    Posted Mar 18, 2008 at 8:56 AM | Permalink

    Re 36: AFAIK the climate science mainstream view is that they were local, or N hemisphere at most, and of course the GWS camp has been trying to show that they were indeed global. A transfer of heat can make as much difference as an overall increase, at least on a temporary basis, as we know from ENSO events. Were the WMP and LIA Dansgaard–Oeschger events?

    (I hope this is not OT.)

  39. cce
    Posted Mar 18, 2008 at 9:01 AM | Permalink

    Leif,

    It is often said that L&F rely entirely on PMOD, and if ACRIM were used, then the CR theory would be consistent. Do you agree with this?

  40. Raven
    Posted Mar 18, 2008 at 9:18 AM | Permalink

    TrueSceptic says:

    Re 36: AFAIK the climate science mainstream view is that they were local, or N hemisphere at most, and of course the GWS camp has been trying to show that they were indeed global. A transfer of heat can make as much difference as an overall increase, at least on a temporary basis, as we know from ENSO events. Were the WMP and LIA Dansgaard–Oeschger events?

    The idea that the LIA and MWP were local is a myth created by the deeply flawed MBH98 reconstruction. This reconstruction has been deconstructed in detail on this site – the links on the left provide more information.

    A recent paper from Craig Loehl provides a robust analysis using non-tree ring proxies that demonstrates that the MWP and LIA were global, however, the error bars make it impossible to say whether the MWP was warmer than today but it is not possible to say that it was cooler either.

  41. Andrew
    Posted Mar 18, 2008 at 9:24 AM | Permalink

    TrueSceptic 37, the trend in TSI depends who you ask:

    http://acrim.com/ACRIM%20Composite%20Graphics.htm

    But we (skeptics) aren’t all “married” to the hypothesis (I find it the most compeling theory personally) and of course intrenal variability, independent of external forcing, as I said above, can’t be ruled out. After all, the A in AGW started right when the PDO “happened” to shift to the warm phase (more El Ninos).

    Leif, can that model be extended to infer CRF variability in the past, before measurements? There are of course some data which show trends and some not…

  42. TrueSceptic
    Posted Mar 18, 2008 at 9:49 AM | Permalink

    Re 41: Well, it so happens that we have a real, genuine expert on TSI here…

  43. TrueSceptic
    Posted Mar 18, 2008 at 10:04 AM | Permalink

    Re 40: I won’t reply to this because it would be a derail.

  44. jae
    Posted Mar 18, 2008 at 10:16 AM | Permalink

    32, Andrew:

    To me, it is not so much the TSI variability as it is things closely associated with it. I think (but of course I can’t prove it) that strong secular trends in Cosmic Rays can emerge in times of general activity/in activity of the sun independently of any trends in TSI. It might seem ad hoc, but then again, if you look at Cosmic Rays and Sunspots, the only real similarity is the periodicity. The amplitudes vary independently of one another.

    It is my understanding that most of the cosmic rays of interest are not produced by the Sun, but by other phenomena in the Universe. Thus, they can vary greatly (chaotically?). Although there is some shielding by the electromagnetic fields, causing a periodicity, the amplitudes could vary enormously. And at times there could be sufficient cosmic rays to “overwhelm” the fields. Is this not correct?

  45. Andrew
    Posted Mar 18, 2008 at 10:29 AM | Permalink

    TrueSceptic, I’m no “expert”, but Leif is, so there is an expert here. If you have TSI questions (as opposed to climate questions) you should probably ask him…

  46. Andrew
    Posted Mar 18, 2008 at 10:31 AM | Permalink

    jae, I wouldn’t know, ask Leif. My intuition (biased) says “yes” however.

  47. TrueSceptic
    Posted Mar 18, 2008 at 10:33 AM | Permalink

    Re 41, 45: Strange, I was responding to your “depends who you ask”.

  48. Andrew
    Posted Mar 18, 2008 at 10:39 AM | Permalink

    Okay, gotcha. Please be more clear about your meaning in the future. It is sometimes difficult to “get” the subtleties on internet discussions, so I misunderstood you.

  49. TrueSceptic
    Posted Mar 18, 2008 at 10:51 AM | Permalink

    Re 48: Sorry, you are right. What seemed obvious to me at the time is less so now. :)

  50. Posted Mar 18, 2008 at 11:41 AM | Permalink

    36 (jack): I didn’t say that the Sun didn’t cause LIA, I said that you cannot DEDUCE solar variability from the existence of LIA.

    39 (cce): The resolution of the TSI-question does not say anything about the CR-hypothesis. Me thinks that is a Darapti Syllogism.

    41 (Andrew): the 10Be and 14C measurements allow us to infer CRF in the past, provided we model the dependencies well enough, as we will eventually. We are not there yet [some people who know everything might disagree].

    44 (jae): that is correct; we distinguish between Solar Cosmic Rays [not called that anymore - now it's SEP, solar Energetic Particles] and Galactic Cosmic Rays, that come from a myriad of sources and therefore do not vary much from year to year.

    49 et al.: Very nice that you all avoid the bickering that is so common…

  51. Posted Mar 18, 2008 at 11:49 AM | Permalink

    45 (Andrew): “export on TSI”: well, in the land of the blind, the one-eyed is king :-)
    I’m no real expert on TSI. The ‘real’ experts disagree all over the map, as you have to be a real expert to do. The TSI-people will have to convince the rest of us about why THEIR particular version is correct or ‘better’. This process is ongoing, and such debate is scientifically healthy. Science is only truly dead when it is ‘settled’.

  52. Posted Mar 18, 2008 at 12:05 PM | Permalink

    27 (Leif):

    The problem I’m having [over and over ] is that some people have things backwards, their argument goes something like this: “Since the LIA and the MWP are obviously caused by the Sun [what else?], then obviously, the Sun must vary accordingly”. This is putting the cart before the horse, IMHO.

    Leif, Have you considered that changes in CR’s may produce anomalous cloud patterns (tropics versus polar), thus promoting wind pattern changes. These wind changes may have the effect of producing greater snowfall in winter or less ice melt in summer, and thus change albedo. Just a wild thought, but probably the kind of thinking required to produce the magnified climate effect due to small solar changes. Wind patterns are an underplayed character in Earth climate.

    Tom

  53. Posted Mar 18, 2008 at 12:18 PM | Permalink

    52 (me):

    Somewhere on CA, I recall seeing a graph of cloud anomolies, and there were slight pattern differences between solar min/max. I cannot find it, so maybe someone could post it again. Also, it has been shown, I believe, that there are more Nina’s in minimums. Nina’s are essentially caused by wind pattern changes. So, even though Leif states that TSI has not changed enough to cause climate fluctuations, we have seen significant cooling during this Nina. I tend to side with Leif on the TSI argument, and additionally believe Earth climate has some magnification effect. What the foundationsignal is that causes this climate slide is still unknown.

  54. Posted Mar 18, 2008 at 12:40 PM | Permalink

    52-53 (tucker): the problem with “may” and “have you considered” etc is that those are just too vague to be useful.

    At the 2008 SORCE meeting I heard the following talk:

    Exploring the Tropospheric Response to Solar Forcing
    David Rind [drind@giss.nasa.gov], NASA Goddard Institute for Space Studies; Judith Lean, NRL, Washington; and Patrick Lonergan and Jean Lerner, Columbia University.

    The investigation of the influence of solar cycles on the stratosphere/troposphere system has become increasingly sophisticated in the last decade, due to a combination of empirical analyses and modeling studies. However, there is still considerable uncertainty about the actual tropospheric response to changes in solar radiative output, and the potential mechanisms involved. At least three competing theories have been proposed: surface absorption of solar radiation changes altering tropospheric dynamics directly; surface absorption changes amplifying the natural modes of ocean-atmosphere coupling; and
    tropospheric responses driven from the stratosphere due to changes in stability or planetary wave propagation. While not necessarily mutually exclusive, these have some opposing interpretations and cause-effect relationships which we examine with an extensive set of climate model simulations.
    Spectrally-differentiated solar radiation from 1950-2005 is input to GISS Global Climate/Middle Atmosphere Model 3 in four different atmospheric configurations and in conjunction with differing oceanic conditions (some 1500 years of simulations). Simulations in which the ocean is not allowed to respond emphasize the solar “stratospheric forcing” of the troposphere, while simulations with the full ocean response illustrate the solar “surfacedriving” capability. Analyses of the model simulations address the following: Do the standard runs reproduce the observed stratospheric and tropospheric temperature changes? Do the ∆SSTs in the calculated ocean runs match those in the historical data? Do the circulation and precipitation changes match those claimed for solar forcing? Are the results dependent on the ozone change profile; or the model resolution; or the other climate forcings that have been occurring?

    —————

    After the talk I asked the speaker why he did not mention the CR influence that is so popular in the debate. His answer basically was that the proponents have not offered any physical mechanism that can be modeled [i.e. computed from other parameters of the system], so nothing meaningful could really be said about it, other than just some hand-waving.

    I tend to agree with his assessment. [I know that Rind's association with GISS may render him 'suspect' in some people's eyes, but that it not a valid excuse for not taking him seriously, so I don't want any comments related to this association]

  55. Posted Mar 18, 2008 at 1:39 PM | Permalink

    Leif – many thanks for being a working climate/solar scientist who is prepared to take the time to communicate your objective scientific views with blog readers.

    I was wondering what your take is on climate change during the Holocene. Personally, as a non-climate scientist, I believe that internal climate variability is driven by known climate cycles such as the Pacific Decadal Oscillation, the North Atlantic Oscillation, the El Nino/Southern Oscillation, and the North Pacific Oscillation (Tsonis et al), plus volcanoes. External variability is due to solar/galactic/cosmic factors. I can’t see much of a role left for the enhanced greenhouse effect.

    What’s your view?

  56. TrueSceptic
    Posted Mar 18, 2008 at 1:44 PM | Permalink

    Re 50: Perhaps because Steve M is keeping this on the rails?

  57. Posted Mar 18, 2008 at 2:01 PM | Permalink

    Re; Lockwood’s latest paper (was there another TGGWS for him to respond to!?) – he assumes that the flawed metric of global average temperature is meaningful and accurate. No one has challenged the McKitrick/Michaels paper on global average surface temperature. I like his uncertainty range of of 49–160%.

    If anyone wants the paper, I can send it – assuming that is legal!

    p.m.biggs@bham.ac.uk

  58. Posted Mar 18, 2008 at 2:19 PM | Permalink

    55 (Paul): I’ll keep an ‘open mind’ [sorry for the associative pain that may evoke :-) ]. I’ll add to your list: Land use [you decide if that is internal or external]. There are people [e.g. Ruddiman] arguing that land use associated with agriculture may have ‘staved off’ an ice age. Maybe, I don’t know.

  59. Andrew
    Posted Mar 18, 2008 at 2:52 PM | Permalink

    I find Ruddiman’s hypothesis far fetched, personally. I find it hard to imagine that anthropogenic global warming dates back to before 1970. But still, its a provocative idea, isn’t it? Just boggles the mind…

  60. Posted Mar 18, 2008 at 2:53 PM | Permalink

    Thanks Leif – I left out anthropogenic factors, including land use, as I was considering natural factors, aside from the ‘enhanced greenhouse effect.’

    If you are correct about the Sun – then it either has very little influence via solar changes, or the climate is even more sensitive to small solar changes than is currently believed – you suggested this as a possibility yourself in your early CA posts, iirc.

    If, as I suspect, climate sensitivity to CO2 and the enhanced greenhouse effect is small, then we are left with the Tsonis et al ‘climate shifts,’ in my humble opinion.

  61. Raven
    Posted Mar 18, 2008 at 2:58 PM | Permalink

    It is worth nothing that the sun could still be responsible for climate change but there could be a 1000+ lag between cause and the effect due to the oceans.

  62. Posted Mar 18, 2008 at 3:05 PM | Permalink

    59 (Andrew): The great Danish physicist Niels Bohr once rejected an idea because, as he said “it was not crazy enough”.

  63. Andrew
    Posted Mar 18, 2008 at 3:06 PM | Permalink

    Paul, given the evidence, I’d say that climate must be fairly sensitive to solar variability (at least, perhaps especially certain regions). I have speculated that Cosmic Rays vary separately from the Sun, which might make that more probable. I still don’t see a problem with that (Leif has stated how CRF can be modeled, but He hasn’t commented on what the results look like.) Other than the fact that it is sort of ad hoc, is there any reason to discount the possibility that CRF varies more than TSI? I don’t know. For example, some records of Cosmic Rays show trends and others don’t. Proxies do but they may be effected by climate and volcanoes. In any case, if the CRF LACC link were correct (Some regard this as a pretty big if, but I have high confidence in the experiments going on right now), then the forcing over the solar cycle is inconsistent with high sensitivities, regardless of the long term trend.

  64. Andrew
    Posted Mar 18, 2008 at 3:10 PM | Permalink

    Leif, good point. I don’t discount Ruddiman’s theory, but I don’t buy it just yet. Its “just crazy enough” to be possible, in my book. ;)

    Raven, watch out, ocean delay is proportional to climate sensitivity! Your talking about a sensitivity WAY above the most alarmist models.

  65. Posted Mar 18, 2008 at 3:22 PM | Permalink

    Well, we still haven’t found L&F’s unknown solar amplifier which, despite being unknown, isn’t currently relevant according to L&F.

  66. Posted Mar 18, 2008 at 3:26 PM | Permalink

    64 (everybody): don’t forget to quote a number and a name if responding to something so one can figure out what belongs to what without having to parse everything.

  67. Andrew
    Posted Mar 18, 2008 at 3:31 PM | Permalink

    Sorry, my last comment was for 62.

  68. jae
    Posted Mar 18, 2008 at 3:40 PM | Permalink

    Leif:

    44 (jae): that is correct; we distinguish between Solar Cosmic Rays [not called that anymore - now it's SEP, solar Energetic Particles] and Galactic Cosmic Rays, that come from a myriad of sources and therefore do not vary much from year to year.

    Are we sure there is so little variation? I recall Shaviv or someone writing that there are large variations associated with the sun and its planetary system moving through the “arms” of the Milky Way.

  69. cba
    Posted Mar 18, 2008 at 3:45 PM | Permalink

    Scafetta & West are providing an interesting result from their black box analysis, especially with much higher climate sensitivity like Wang 2005 TSI. This approach looks at the data and not at the mechanisms actually in play.

    Leif,
    I did a comparison of ACRIM versus SORCE for 2003-2007. Average difference was about 4.9x W/m^2 and st dev was something like 0.15 W/m^2. Data was aligned to +/- 1 day (corrected for missing days). Out of 136x W/m^2 that’s rather good correlation not really showing any serious drift over several years. Thats the ACRIM composite data done by Willson (originally from ACRIM3 for that section. At least some bit of that std dev variation seems to be associated with the date/time alignment and some of those serious short term transients.

  70. Raven
    Posted Mar 18, 2008 at 3:55 PM | Permalink

    Andrew says:

    Raven, watch out, ocean delay is proportional to climate sensitivity! Your talking about a sensitivity WAY above the most alarmist models.

    The sensitivity models are built on the assumption that nothing other than GHGs is capable of causing temperatures to rise and that all of the current temperature rise is a result of heat trapped by GHGs in the atmosphere today. A model where the climate has ‘seasons’ lasting 1000+ years would require a lower sensitivity to current GHG emissions to account for heat released from the oceans. Someone has even developed a theory that explains the ice ages by postulating 12000 year ocean cycles. These effects would reduce CO2 to the bit player that is predicted by the basic Stephan-Boltzmann equations.

  71. Posted Mar 18, 2008 at 3:57 PM | Permalink

    68 (jae): the spiral arm variation is on a time scale of hundred of thousands [or more] years. On a time scale of millennia, the flux is pretty constant [we think], unless a supernova goes off ‘nearby’, such as η Carinae.

    69 (cba): the issue is how well ACRIM3 can be ‘stitched’ to ACRIM2, etc. So between ACRIM3 alone and SORCE you may have very good correspondence [except for the 4.9 W/sqm difference].

  72. cba
    Posted Mar 18, 2008 at 4:00 PM | Permalink

    Leif,

    I think Shaviv and Svensmark have both written about long term periodic variations of CR from our location in the arm and perhaps on the orbit (considering there might be a situation over there through the galactic center which we can’t see that might alter things too). Also, there are intermediate time events which might alter the situation – like starburst formations (forming of open clusters due to a supernova or the forming of a large A star creating localized pressure waves doing likewise…).

    I think Svensmark coined the term ‘cosmoclimatology’ and has also written a pop press book titled Snowball Earth – which is likely the origins of this stuff. The only other competition I’m aware of is the Milankovich cycles which may also have additional effects, in addition to or instead of the CR long term variation. That is except for possibly very long term solar variations.

    BTW, how does your ‘flat line sun’ theory work with the Glassner cycle and potentially other longer cycles?

  73. Andrew
    Posted Mar 18, 2008 at 4:03 PM | Permalink

    jae, 68, we are talking to radically different time scales here. Year to year versus “mega year”.
    Raven, 70, one strong argument for low climate sensitivity is the rapid recovery of the climate after volcanic eruption, but with ocean delay of a thousand years, well, that would be impossible. I think we are refering to two different concepts here.

  74. cba
    Posted Mar 18, 2008 at 4:09 PM | Permalink

    Leif,

    My thought was more on how well does ACRIM self adjust (compensate) for deterioation on the long run – your suggestion of SORCE’s claim to fame over that of ACRIM and previous approaches. Obviously, no one can really test for the stitching very well. It does lead to the notion that Willson is doing a decent job keeping the ACRIM measurements on track at least recently. Keeping them on track and independently verifiable recently would suggest that his group knows the equipment and understands what it is doing over time. Knowing that would suggest that the stitching is going to be our best bet as to what we have in the way of a composite TSI for the last 30+ years. Somehow, I don’t think we’re going to do better than that for quite a few years as we collect newer data. I personally haven’t tried the PMOD, but it doesn’t look like it’s going to possibly be as as accurate.

  75. Posted Mar 18, 2008 at 4:11 PM | Permalink

    72 (cba): don’t forget a reference to a comment #. My reply to #72 would be #71. The time scale thing.

    Longer cycles: there is no evidence for longer ‘cycles’. The Gleissberg ‘cycle’ [length cited as anything from 77 to 110 years] may just be a ‘natural’ time scale for solar processes, just like the ‘decadal’ cycles in terrestrial weather/climate. Glassner cycle? Barry Glassner have talked about cycles of fear in relation to the business cycle, but I don’t think you are thinking of him.

  76. Posted Mar 18, 2008 at 4:16 PM | Permalink

    74 (cba): I’ll let the TSI-people squabble over their calibrations. They will eventually get there. In the meantime, we, the users of their TSI, will just have to do a sensitivity analysis to ensure that any result we get does not depend solely on the trend or exact value of TSI.

  77. cba
    Posted Mar 18, 2008 at 4:29 PM | Permalink

    Leif,

    re. eta C., it’s almost passe’ for the worry warts. There’s a new kid on the block with some totally unmemorable number – actually, it’s twins. There are two WR stars out there – big massive brutes and it seems we are viewing their system on end – suggesting that if one of them goes as a SN with GRB – it’s pointed at us. They’re also in the general vacinity of being the same distance as eta C if I recall properly. It also brings to mind the notion of if one goes – maybe it’ll become a double header.

    While these are all way too far off to just remove our atmosphere or fry us in our tracks, they could still possibly provide quite a damaging extinction level event.

  78. Posted Mar 18, 2008 at 4:33 PM | Permalink

    77 (cba): please reply in the style I’m using here. Your last post should have looked like this:

    71 (Leif): re. eta C., it’s almost passe’ for the worry warts. There’s a new kid on the block with some totally unmemorable number – actually, it’s twins. There are two….

    This make it possible to process the comments by computer and link the comments together in a meaningful way. It is not just me whining…

  79. Posted Mar 18, 2008 at 4:49 PM | Permalink

    77 (cba): not wanting to be pedantic, but the WR-twins would emit a gamma ray burst [and that would fry us] if directed straight at us. Cosmic ‘rays’ are not [electromagnetic] gamma rays, but charged particles [mostly protons] that would not reach us in a straight line but would spiral along the [unruly and turbulent] Galactic magnetic field and be dispersed a lot.

  80. Bill F
    Posted Mar 18, 2008 at 5:16 PM | Permalink

    #63 Andrew,

    Keep in mind that there is no such thing as a “GCR detector”. Most of the records we have distinguish between GCRs and SEPs by their energy level, and there is always contamination of the GCR record from solar derived particles that fall into the GCR energy range. Also keep in mind that the measurement of particles at the energy levels of GCRs is dependent on the location of the detector within the earth’s magnetic field. The ability of GCRs to reach the earth’s surface (or the lower atmosphere) will change as the earth’s magnetic field changes. Over the last 150 years or so, the earth’s overall field strength has declined by about 10%, which would have the effect of letting in more GCRs. But also remember that the earth’s field (like its temperature) has not changed the same everywhere on the planet. Most of the decline in dipole moment can be traced to changes around the large magnetic anomaly in the south atlantic ocean.

    The point is that there is more to the GCR story than just modulation by solar activity. There is alot of noise and localized affect that has to be teased out of an individual record before it can be compared to a global record of clouds/temperature to look for similar trends. I tend to believe that the physical mechanism will be proven out by Svensmark and the Cloud experiments, and I think they will end up showing that the strongest effect will be found over large areas of ocean (which will then drive a somewhat lagged climate response to long term changes in GCR flux). However, there is still alot to learn about GCRs and their modulation before they will be able to present a really convincing argument for the climate connection without having their datasets attacked as inadequate.

  81. Posted Mar 18, 2008 at 5:37 PM | Permalink

    80 (Bill): Everybody, please try to stick to my convention: “nbr (name):”. I do plan to process all of these posts by machine. The ones that follow my convention might be skipped or not linked to where they belong. Please, it is not hard.

  82. Posted Mar 18, 2008 at 5:39 PM | Permalink

    81 (me): grrr, “the ones that don’t…”

  83. Bill F
    Posted Mar 18, 2008 at 5:45 PM | Permalink

    81 (Leif):

    My apologies. I think you need to be more specific that you want EXACTLY that format. I thought your previous posts were directed at users who weren’t putting any number or name at the front of their posts. I thought I had done quite well by putting a number and name only to realize that I had still failed you. ;)

  84. Bill F
    Posted Mar 18, 2008 at 5:47 PM | Permalink

    83 (me): And apparently you had just done so while I was typing my original message…as Gilda Radner would have said on Saturday Night Live…NEVER MIND!

  85. Andrew
    Posted Mar 18, 2008 at 5:48 PM | Permalink

    80 (Bill): Thanks for the input. The data are tricky in that way, unfotunately.

    81 (Leif): Did I do it right? Just checking.

  86. Posted Mar 18, 2008 at 5:51 PM | Permalink

    85 (Andrew): yes, when I stop whining you are doing great. There is a lot of interesting comparative work that can be done with all these posts, if they are organized in a reasonable way.

  87. TrueSceptic
    Posted Mar 18, 2008 at 6:22 PM | Permalink

    80+ (various): It would be easier if this site used the quoting conventions used by many other forums. Hit “quote” and you automatically reference what you are replying to.

    Please tell me if this function is here and I can’t see it.

  88. TrueSceptic
    Posted Mar 18, 2008 at 6:23 PM | Permalink

    86 (Leif): Are you conducting a survey of some sort?

  89. Posted Mar 18, 2008 at 6:33 PM | Permalink

    88 (TrueSceptic): survey? No, but from my work as a scientist I have learned that with a very small amount of meta-data up front, the effort of doing data analysis later on [if the wish arises] will be greatly facilitated. Injecting meta-data into old data is a very expensive process. At some point, someone might say: “wouldn’t it be great if we had A, B, or C”, and often the sad answer is that we can’t have any of those because of the prohibitive price of meta-data.

  90. Posted Mar 18, 2008 at 6:37 PM | Permalink

    87 (TrueSceptic): “It would be easier if this site used the quoting conventions used by many other forums. Hit “quote” and you automatically reference what you are replying to.”

    And how does the system know what you are replying to? And does it you your name? Maybe it does, I don’t know. I hate to be hostage to an automatic system I don’t control :-)
    Anyway, please, humor me from now on, so we can get back on topic.

  91. cba
    Posted Mar 18, 2008 at 6:58 PM | Permalink

    79 (Leif):

    Just checked, I used GRB. They’re rather unaffected by EM fields and distance, unlike CR. They also travel perhaps 0.000001% faster too. If the WR twins both went off in our direction, I’m not sure anyone would bother looking to see if there were any rise in CR a few days or weeks later.

  92. Posted Mar 18, 2008 at 8:18 PM | Permalink

    91 (cba): just trying to keep you on topic :-)

    The ‘constancy’ of GCR on time scales of centuries is enshrined in the models describing cosmic ray propagation and modulation. If we didn’t have a heliosphere, the GCR intensity would increase 1-2 orders of magnitude. If GCRs already now have any climate effect, a 1-2 orders of magnitude increase might have very dramatic consequences, perhaps much larger than whatever changes did occur during the Maunder Minimum, MM. The absence of such dramatic changes might tell us that the solar wind did not ‘shut down’ during the MM. Another piece of evidence in this direction is:

    Did the Sun’s Prairie Ever Stop Burning?
    Foukal, P.; Eddy, J.
    Solar Physics, Volume 245, Issue 2, pp.247-249, 10/2007
    DOI: 10.1007/s11207-007-9057-8

    Abstract
    The presence of the red flash at total solar eclipses requires the existence of an extended chromosphere and therefore of a photospheric magnetic network that gives rise to spicules. We draw attention to the earliest historical reports of a red flash at the 1706 and 1715 eclipses, which therefore imply a substantial, widespread photospheric field during at least the last decade of the Maunder Minimum. Our finding is consistent with reports of a persistent photospheric field throughout the Maunder Minimum from analyses of 10Be radioisotope evidence. We note, however, that the last decade may not be representative of conditions throughout the roughly 1645-1715 extent of that prolonged activity minimum.

    ——

    In spite of the cautious disclaimer in the last sentence, one might note that it takes time for the solar dynamo to build a sizable magnetic field.

  93. cce
    Posted Mar 18, 2008 at 11:04 PM | Permalink

    50 (Leif)

    Re: PMOD vs ACRIM & Lockwood

    Hoyt was quoted earlier (19) as saying “the Lockwood paper is nonsense” (referring to the L+F paper from last year). Do you agree with this characterization?

  94. Posted Mar 19, 2008 at 4:13 AM | Permalink

    93 (cce): [don't forget the colon after (name), like (name):]

    the Lockwood paper is nonsense

    Both Lockwood and, especially, Frohlich are respected scientists and have the right to be wrong, especially Lockwood. To characterize their work as nonsense is the kind of language that I would not subscribe to, even if the paper made no sense to me.

  95. cba
    Posted Mar 19, 2008 at 6:45 AM | Permalink

    92 (Leif):

    Seems I saw a suggested number recently of as small as a 0.25% variation in TSI (undoubtedly with all the other ramifications associated which might have an effect) was what might have caused the Maunder Min.

    It’s pretty obvious that our short term yearly variation of 6.6% (90W/m^2 due to orbital distances) is not wiping us out despite the fact that our tilt is enough to drive higher lattitudes into winter and summer conditions during that time frame, totally freezing over/thawing out about 10% of the surface on this short term basis. Also, the earth spends about 90% – 95% of its time in hardcore iceage conditions in a fairly consistent periodic fashion. This is conceptually understandable from the basis of a cloud regulated albedo mechanism, an iris type effect, which would be short circuited by high reflectivity from snow cover.

    Consequently, it would seem the search for the sun/earth climate factor should be on that which causes clouds to be formed (or prevents them from forming). This could either be direct from solar effect, such as modulating the CR flux which will in some cases affect the % of clouds present – at least in some areas. Or, it might be even less obvious and associated with the creation of weather conditions enhancing or hindering cloud formation. Like the clouds, it’s probably compounded even more by the fact that there are multiple types, each with multiple effects and multiple net effects and that it’s just as likely that the ultimate result is superposition of several virtually inconsequential effects rather than one overriding one. And now it seems that the satellite cloud data is defective.

  96. Posted Mar 19, 2008 at 7:15 AM | Permalink

    95 (cba): “as small as a 0.25% variation in TSI”. I think it is even smaller, like half that. To make the problem more acute.

    “And now it seems that the satellite cloud data is defective”. If so [and I believe this is indeed the case; viewing angle and all that], then the correlations people have found between cloud cover and anything [such as CRs] must be spurious. Which means that there does not even seem to be evidence of a cloud-solar connection.

    This all leads to the question: does there have to be a solar connection? even when the evidence for it is fading…

  97. Pete
    Posted Mar 19, 2008 at 7:53 AM | Permalink

    95 (cba): “as small as a 0.25% variation in TSI”.

    NASA offered a solution to this based on their analysis of the Maunder Minimum. They appeared to even suggest what the ‘magnifying effect was’, but I don’t know why this research wasn’t followed-up, discredited, enhanced upon. Seems nothing was done afterwards!

    NASA on the MM

  98. Pete
    Posted Mar 19, 2008 at 7:55 AM | Permalink

    97(pete): Sorry, this article was based on research done in the following paper. Some famous names in that lot!

    Regional MM Climate

  99. Mike Davis
    Posted Mar 19, 2008 at 8:11 AM | Permalink

    97 (Pete):
    Based on methods and results of those papers I think they should remain in oblivion.

  100. Posted Mar 19, 2008 at 8:23 AM | Permalink

    99 (Mike D): The paper should remain in oblivion simply because the solar irradiance used (Bard, Hout&Schatten, Lean 1995) is obsolete and does not reflect current thinking and reconstructions.

  101. cba
    Posted Mar 19, 2008 at 8:27 AM | Permalink

    96 (Leif):

    Obviously there is some connection. Without the sun, we’d be well on our to double digit average temperatures, if not towards single digit degrees K. I think it’s fairly well accepted that the sun was pushing some 30% less energy out quite some time back. We’ve had life around perhaps from the first few hundred million years of earth’s existance (at least after things settled down a bit with that probable lunar formation mess). Considering at least some mainstream astrophysics types accept that the sun is variable over all time frames, changes of output in the order of 0.25% don’t seem all that far fetched when we observe changes over the short term in excess of 0.1% through a typical SS cycle. Considering the few cycles we’ve observed differ significantly (or by easily measurable amounts) in intensity, I find it hard to accept the concept that it’s not going to vary by more than 0.125%. That would seem to be less than what would be expected from fairly typical sunspot cycle variations.

    Also, you didn’t respond to my question on Glassner cycles, etc. Am I to take it that you do not think these might actually be real?

  102. Posted Mar 19, 2008 at 8:58 AM | Permalink

    101 (cba): About variations of TSI. One has to specify a time frame. From day to day, TSI can vary quite a lot [up to 0.35%, e.g. during October 2003], but there variations [from one day to the next] are so short-lived that it is hard to see how they affect climate. On a time scale of a year, the variation is of the order of 0.1%, and on a time scale of centuries I think it is 0.125%. On a time scale of billions of years, we do indeed have 30% variations or more [five billion years from now when the Sun becomes a red giant]. But I do not think we need to drag that old saw into the equation. So, the issue is solely the 0.25%, where does that number come from? It looks like Lean’s 2001 reconstruction that she doesn’t even believe in any more [she was a coauthor of Wang et al. that has a variation of 0.12%. Is it really necessary to go around in circles on this?

    About the Glassner cycles: I have no idea what you talking about, so educate me.

  103. Andrew
    Posted Mar 19, 2008 at 9:25 AM | Permalink

    100 (Leif) The mechanism proposed is what I find interesting. Wrong reconstruction or not, giving it consideration in models would change their output (though I thought that was obvious). Also in response your 96, there doesn’t have to be a connection, but it would leave me scratching my head trying to understand the evidence I am aware of. The cloud correlation may be spurious, but maybe we should wait and see when we have better data before we jump to conclusions. The experiments are still promising, so maybe the sattelites just happened to get the right answer.

  104. Posted Mar 19, 2008 at 9:35 AM | Permalink

    103 (Andrew):

    The cloud correlation may be spurious, but maybe we should wait and see when we have better data before we jump to conclusions

    This sounds like good advice to the people proposing the cloud-sun connection in the first place before they jumped to their conclusions.

  105. Pete
    Posted Mar 19, 2008 at 9:45 AM | Permalink

    103(Andrew): The mechanism proposed is what I find interesting.

    Yes Andrew. The important connection in this Paper was UV>Ozone>Planetary Waves>Various Climate Oscillations>Drivers of Regional Climate(including NH).

    Back to the UV. I thought Lean at the SORCE project was investigating the variation in the Sun’s UV Spectrum? I understood some studies showed this to be varying at 10 Times than that of the whole TSI Spectrum?

  106. Posted Mar 19, 2008 at 9:53 AM | Permalink

    195 (Pete):

    the variation in the Sun’s UV Spectrum?

    Yes, this varies a lot more than the whole TSI, and if you goto Xrays the variations are even greater: thousands of times more, but the energy involved for both the FUV and below is so minute that we are talking solar radiative forcing anymore. Then we are dealing with chemistry in the stratosphere etc. If you talk UV and below, you have disconnected yourself from TSI. Apples and Oranges.

  107. Pete
    Posted Mar 19, 2008 at 9:54 AM | Permalink

    104(Pete): Myself

    Yes, this is the Lean document from 2007. “Also, the solar ultraviolet radiation, which varies far more than the total solar irradiance, influences stratospheric chemistry and dynamics, which in turn appears to couple to the troposphere and provide an indirect forcing of surface temperatures [e.g., White, 2006] and climate [e.g., Geller, 2006].”

    Lean PDF

  108. Posted Mar 19, 2008 at 9:56 AM | Permalink

    106 (me): “we are NOT talking solar radiative forcing anymore.”

  109. Posted Mar 19, 2008 at 9:59 AM | Permalink

    107 (Pete): of course, all of that is well known but now you are not talking solar forcing in terms of W/sqm. Let’s not confuse the issues. If you want to talk about TSI don’t bring UV into the picture.

  110. Pete
    Posted Mar 19, 2008 at 10:00 AM | Permalink

    108(Leif): Sorry, I thought you were searching for any Solar>Climate connection, not just TSI? I think I’ve missed the point somewhere along the line, but I’m not sure where! I enjoy reading this thread though.

  111. Posted Mar 19, 2008 at 10:07 AM | Permalink

    110 (Pete): this particular topic started with #97. Of course, all the other usual suspects must be considered in the total picture, but sometimes focusing on one at a time is useful.

  112. Andrew
    Posted Mar 19, 2008 at 10:31 AM | Permalink

    104 (Leif) Fair enough. I think the results would validate such a link, personally, but I’m always willing to be wrong.

  113. TrueSceptic
    Posted Mar 19, 2008 at 2:43 PM | Permalink

    90 (Leif): It was just an aside. That is how most internet forums work. This is a blog that allows comments so perhaps I shouldn’t expect the same.

    End of aside. :)

  114. TrueSceptic
    Posted Mar 19, 2008 at 3:10 PM | Permalink

    101 (cba): Do you mean Gleissberg cycles (87 years, varying between 70 and 100, being a multiple of the 11-year cycle)?

  115. See - owe to Rich
    Posted Mar 19, 2008 at 3:43 PM | Permalink

    111 (Leif)

    When I read the link from #97, it spoke to me of ozone changes caused by UV changes. Therefore this isn’t TSI, but part of the spectrum which varies more.

    I firmly believe in one Sun, and in its connection to the LIA because temperatures decreased and sunspots decreased at the same time.

    But others believe in AGW because temperatures rose while CO2 concentrations rose (albeit only 25 years).

    Who is to say my religion is righter than theirs? (Hey, I’m not looking for a deafening silence here, defend me! Not you, Leif, you’re exempt.)

    Rich.

  116. cba
    Posted Mar 19, 2008 at 4:03 PM | Permalink

    102 (Leif):

    Sorry about the glassner cycle – dropped a bit or two somewhere. Let’s try Gleissberg cycle and toss in the Suess cycle and Hallstatt cycle as well.

  117. Posted Mar 19, 2008 at 7:18 PM | Permalink

    116 (cba): From my reply in #75″

    Longer cycles: there is no evidence for longer ‘cycles’. The Gleissberg ‘cycle’ [length cited as anything from 77 to 110 years] may just be a ‘natural’ time scale for solar processes, just like the ‘decadal’ cycles in terrestrial weather/climate. Suess cycles may be two Gleissberg cycles. This signal is reportedly seen in the 14C.
    Other cycles? Clilverd et al. have a whole slew of them: SPACE WEATHER, VOL. 4, S09005, doi:10.1029/2005SW000207, 2006; They say:
    “We use a model for sunspot number using low-frequency solar oscillations, with periods 22, 53, 88, 106, 213, and 420 years modulating the 11-year Schwabe cycle”.
    Ken Schatten has a good word for this: “Cyclomania”.

    McCracken [JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 112, A09106, doi:10.1029/2006JA012119, 2007] believes in even weirder stuff:
    “The cosmogenic cosmic ray data exhibit a ∼2300-year periodicity, and it is proposed that the steadily increasing HMF since the 15th century represents the first quarter cycle of an associated ∼2300-year periodicity in the HMF”.
    Needless to say we disagree with him see our comment here. BTW, our comment paper has been rejected by four reviewers after five months of deliberation. Apparently, there are powerful, dark forces one is not supposed to disagree with.

  118. Posted Mar 19, 2008 at 8:27 PM | Permalink

    118 (me):

    This just in:

    [1] Initial 14C activity in a late Holocene stalagmite (a14Cstal) from Attaho¨hle (Sauerland, Germany) varies synchronously with atmospheric 14C activity (a14Catm). This observation provides independent chronological support for the correlation between a solar forcing proxy (atmospheric 14C production rate, a14Cprod) and a climate proxy (d18Ostal) in this stalagmite, and eliminates the need to ‘tune’ the chronology. Monte Carlo randomizations in which 100,000 synthetic d18Ostal datasets were generated indicate that the correlation between a14Cprod and d18Ostal is statistically robust. The absence of a significant time lag between a14Cprod and d18Ostal implies rapid climate response at this site to variations in solar output. Positively correlated a14Cprod-d18Ostal indicates drier winter conditions during late Holocene solar minima. This supports atmospheric amplification models that predict reduced NAO indices during solar minima if kinetic enrichment of d18O during dry phases outweighs the
    reduction of d18O in rainfall that results from lower air temperatures during negative NAO phases.
    Citation: Jackson, A. S., F. McDermott, and A. Mangini (2008), Late Holocene climate oscillations and solar fluctuations from speleothem STAL-AH-1, Sauerland, Germany: A numerical perspective, Geophys. Res. Lett., 35, L06702, doi:10.1029/2007GL032689.

  119. David Archibald
    Posted Mar 19, 2008 at 10:13 PM | Permalink

    Re 115, Rich, you need a bit of cheering up. Try some Usoskin: http://cc.oulu.fi/~usoskin/personal/Sola2-PRL_published.pdf

    Here’s a quote from that paper: The most striking feature of the complete SN profile is the uniqueness of the steep rise of sunspot activity during the first half of the 20th century. Never during the 11 centuries prior to that was the Sun nearly as active. While the average value of the reconstructed SN between 850 and 1900 is about 30, it reaches values of 60 since 1900 and 76 since 1944. We conclude that the high level of solar activity since the 1940s is unique since the year 850.

    As Usoskin notes, the increase in solar activity was associated with an increase in the Earth’s temperature. Usoskin’s study was based on incontrovertible Be10 and C14 data.

    The Maunder Minimum had five solar cycles in a 70 year period, for an average of 14 years each. According to the temperature/solar cycle length relationship demonstrated by Friis-Christensen and Lassen and quantified by me at 0.7 degrees per year of solar cycle length, this would equate to a two degree decline for the mid-latitudes, which is about what happened in the Maunder Minimum. In terms of mode of action, I buy the UV-ozone-NAO theory from Drew Shindell in 98. I have seen a number of other papers outlining the same mechanism, a mechanism which satisfies Dr Svalgaard’s requirement for causality with only a minute change in TSI. GCR-clouds work too – there was a spike of Be10 late in the Maunder Minimum which was associated with a colder decade.

    Since the mechanism relies upon the NAO being negative during weak solar activity, one way to check it may be to look at positive versus negative NAO during the last 100 years of solar maxima and minima. Does anyone know if this has already been done? A couple of years ago the UK Met Office predicted a cold winter on the basis of a negative NAO, and were correct. It would also be good to track UV in the solar cycle 23 progression, if anyone can recommend a source of data.

    In terms of the progression of this cycle, the last 27 day period of the IMF averaged 3.3, a bit under Dr Svalgaard’s floor of 4 to 5. As we have possibly a year to go before minimum, I have lowered my projection of the low to 2 in terms of IMF.

  120. Posted Mar 20, 2008 at 3:29 AM | Permalink

    119 (David A):

    In terms of the progression of this cycle, the last 27 day period of the IMF averaged 3.3, a bit under Dr Svalgaard’s floor of 4 to 5. As we have possibly a year to go before minimum, I have lowered my projection of the low to 2 in terms of IMF.

    David, when you use the OMNI data to get the IMF: http://omniweb.gsfc.nasa.gov/form/dx1.html

    you have to pay attention. On the web-form you select variables to show:

    Bartels Rotation Number
    IMF Spacecraft ID
    Plasma Spacecraft ID
    # Fine Scale Points in IMF Avgs
    # Fine Scale Points in Plasma Avgs

    etc.

    It is very important to select the two above ones [Fine scale points...]. If you do, you get:

    Selected parameters:
    1 Bartels rotation number
    2 # of points in IMF averages
    3 # of points in Plasma averages
    4 Field magnitude average, nT
    5 SW Proton Density, N/cm^3
    6 SW Plasma Speed, km/s

    YEAR DOY HR 1 2 3 4 5 6
    2007 3 0 2367 27 27 5.1 3.8 467.
    2007 30 0 2368 27 27 4.5 5.1 437.
    2007 57 0 2369 27 27 4.8 5.8 456.
    2007 84 0 2370 27 27 4.4 6.9 421.
    2007 111 0 2371 27 27 4.1 5.5 418.
    2007 138 0 2372 27 27 5.2 5.2 450.
    2007 165 0 2373 27 27 4.4 6.4 438.
    2007 192 0 2374 27 27 4.2 5.3 437.
    2007 219 0 2375 27 27 4.4 5.7 435.
    2007 246 0 2376 27 27 4.2 4.9 449.
    2007 273 0 2377 27 27 4.1 4.8 422.
    2007 300 0 2378 27 27 4.6 4.6 433.
    2007 327 0 2379 27 27 4.1 5.7 447.
    2007 354 0 2380 27 27 4.4 5.1 480.
    2008 16 0 2381 27 27 4.6 4.3 499.
    2008 43 0 9999 1 3 3.3 2.3 638.

    then you can see that the 3.3 nT you saw in the last row, refers to only one day of data and not a whole rotation (27 days). Because of compression and rarefaction waves in the solar wind, the IMF is not steady but goes up and down accordingly. Only when you have a full rotation (27 days) does the number reflect the ‘global’ IMF. If you look back through the previous rotations since the beginning of 2007 shown above you’ll notice that the IMF indeed is between 4 and 5 as it should be [and with no downward trend].

    Your credibility falters with such sloppy work.

  121. Posted Mar 20, 2008 at 3:36 AM | Permalink

    120 (me): hmmm, I was trying to use the ‘Code” quicktag, but it didn’t do what I wanted [although the preview looked great]. Try once more:

    YEAR DOY HR 1 2 3 4 5 6
    2007 3 0 2367 27 27 5.1 3.8 467.
    2007 30 0 2368 27 27 4.5 5.1 437.
    2007 57 0 2369 27 27 4.8 5.8 456.
    2007 84 0 2370 27 27 4.4 6.9 421.
    2007 111 0 2371 27 27 4.1 5.5 418.
    2007 138 0 2372 27 27 5.2 5.2 450.
    2007 165 0 2373 27 27 4.4 6.4 438.
    2007 192 0 2374 27 27 4.2 5.3 437.
    2007 219 0 2375 27 27 4.4 5.7 435.
    2007 246 0 2376 27 27 4.2 4.9 449.
    2007 273 0 2377 27 27 4.1 4.8 422.
    2007 300 0 2378 27 27 4.6 4.6 433.
    2007 327 0 2379 27 27 4.1 5.7 447.
    2007 354 0 2380 27 27 4.4 5.1 480.
    2008 16 0 2381 27 27 4.6 4.3 499.
    2008 43 0 9999 1 3 3.3 2.3 638.

    NASA even tries to show you that the last rotation is not full by giving it a rotation number of 9999. The last full rotation, as you, can see, was 2381.

  122. Posted Mar 20, 2008 at 3:40 AM | Permalink

    121 (me): I’ll just experiment a bit, so the next few comments are to be ignored.

    <pre>
    code code code
    code code code
    </pre>

    test

  123. Posted Mar 20, 2008 at 3:43 AM | Permalink

    122 (me): I give up for now. Anybody knows how to get this to work?

  124. Posted Mar 20, 2008 at 3:54 AM | Permalink

    119 (David A): Although just quoting selected old papers is not the way to conduct a debate you may wish the contrast the conclusion of Usokin et al: “We conclude that the high level of solar activity since the 1940s is unique since the year 850″ with that of Muescheler et all [Quaternary Science Reviews vol 26, p.82, 2007]:

    The tree-ring 14C record and 10Be from Antarctica indicate that recent solar activity is high but not exceptional with respect to the last 1000 yr.

    Clearly, there is some [healthy] debate going on and good scientific conduct says not to selectively pick what fits your ideas without giving due consideration to other views.

  125. Posted Mar 20, 2008 at 4:43 AM | Permalink

    In interpreting 10Be data it must be remembered that 10Be is mostly produced at low to mid-latitudes: see here from whence it moves with the atmospheric circulation to the polar regions for deposition in the ice, thus ‘atmospheric circulation’ is also a factor that influences the 10Be concentration. That is, not only GCRs play into this, but the weather/climate too.

  126. Posted Mar 20, 2008 at 4:51 AM | Permalink

    117 (me,cba): Here is a report of the ultimate magnetic cycle [if you want to believe that]:

    Sun’s Magnetic Activity Varies In 100,000-Year Cycles
    Thanks to new calculations by a Dartmouth geochemist, scientists are now looking at the earth’s climate history in a new light.

    Mukul Sharma, Assistant Professor of Earth Sciences at Dartmouth, examined existing sets of geophysical data and noticed something remarkable: the sun’s magnetic activity is varying in 100,000-year cycles, a much longer time span than previously thought, and this solar activity, in turn, may likely cause the 100,000-year climate cycles on earth.

    This research helps scientists understand past climate trends and prepare for future ones.

    Published in the June 10 issue of Earth and Planetary Science Letters (Elsevier, volume 199, issues 3-4), Sharma’s study combined data on the varying production rates of beryllium 10, an isotope found on earth produced when high-energy galactic cosmic rays bombard our atmosphere, and data on the past variations in the earth’s magnetic field intensity.

    With this information, Sharma calculated variations in solar magnetic activity going back 200,000 years, and he noticed a pattern.

    Over the last 1 million years, the earth’s climate record has revealed a 100,000-year cycle oscillating between relatively cold and warm conditions, and Sharma’s data on the sun’s magnetic activity corresponded to the earth’s ice age history.

    “Surprisingly, it looks like solar activity is varying in longer time spans than we realized,” says Sharma. “We knew about the shorter cycles of solar activity, so maybe these are just little cycles within a larger cycle. Even more surprising is the fact that the glacial and interglacial periods on earth during the last 200,000 years appear to be strongly linked to solar activity.”

    Sharma’s calculations suggest that when the sun is magnetically more active, the earth experiences a warmer climate, and vice versa, when the sun is magnetically less active, there is a glacial period. Right now, the earth is in an interglacial period (in between ice ages) that began about 11,000 years ago, and as expected, this is also a time when the estimated solar activity appears to be high.

    Beryllium 10 is useful for studying the geology from hundreds of thousands of years ago mainly because it has a half-life of about one and a half million years.

    In addition, there are two key factors that have affected beryllium 10 production over the last 200,000 years: the earth’s magnetic field and the sun’s magnetic activity. When there are high-intensity solar magnetic storms, more charged particles are interacting with cosmic rays, and less beryllium 10 is produced. Likewise, the earth’s magnetic field changes the flux of cosmic rays into and out of the atmosphere.

    Since the production rate of beryllium 10 and earth’s magnetic field intensity are known for the last 200,000 years, Sharma could calculate solar magnetic activity for this time period.

    “I took sets of existing, independent data and made new comparisons and calculations,” says Sharma. Then he went a step further to make a connection with the history of ice ages by looking at oxygen isotopes in the oceans, which reveal the history of how much ice was at the poles and are therefore a measure of average global surface temperature.

    “I compared the estimated past variations in the solar activity with those of the oxygen isotopes in the ocean. Although there is a strong relationship between solar activity and oxygen isotopic variations, it is too early to say exactly what is the mechanism through which the sun is influencing the terrestrial climate.”

    One explanation of the 100,000-year cycle was offered by the Milankovitch Theory of Ice Ages in the 1940s, which suggested that the cyclical variations in the earth’s orbit around the sun result in the earth receiving varying amounts of solar radiation that, in turn, control the climate.

    This explanation is under dispute because the variations of the solar energy in relation to the changes in orbit are very small. Other current research focuses on past variations in the sun’s irradiance, or heat intensity (as opposed to the magnetic activity).

    Sharma notes that more analysis is needed to test his theory.

    “I’ve only looked at 200,000 years. My calculations need to be verified for a million years, for instance. Plus, regarding the current global warming debate, it still needs to be examined if the role of solar activity will exacerbate the rising temperatures that result from carbon dioxide buildup in the atmosphere.”

    This work was supported by Dartmouth College, the Max Planck Institute and by a grant from the National Science Foundation. – By Susan E. Knapp [06-Jun-2002]

  127. Andrew
    Posted Mar 20, 2008 at 5:36 AM | Permalink

    126 (Leif)

    it still needs to be examined if the role of solar activity will exacerbate the rising temperatures that result from carbon dioxide buildup in the atmosphere.

    Why did they find it necessary to insert that bit? Well at least they acknowledge debate…This is interesting, I think I’ve heard of it before. Hmm, why would magnetic activity of the sun/earth just happen to have the same or about the same periodicity as Milankovitch cycles? Seems odd to me.

  128. Posted Mar 20, 2008 at 5:56 AM | Permalink

    127 (Andrew):

    happen to have the same or about the same periodicity as Milankovitch cycles? Seems odd to me.

    And solar activity just happens to be low during the LIA too. Seems odd to me :-) There is a human tendency to pay attention to ‘oddities’ [helps survival in the jungle from whence we came]. I see hundreds of car license plates each day, yet don’t remember any except the “xxx 007″ I saw yesterday [being a Bond fan; both of the Bonds (page 4)]

    The problem with all these things is to separate the various causes, some with [approx.] similar periods and this is really hard.

  129. David Archibald
    Posted Mar 20, 2008 at 7:16 AM | Permalink

    Re 120, Dr Svalgaard, I am afraid you haven’t read deeply enough into the omni site. If you click the data availability button, you will see that 2008 042 is the last day of data for the IMF. So why did you select 2008 043? Here is the button: http://omniweb.gsfc.nasa.gov/html/ow_data.html#Availability
    And here is the result:

    YYYY DDD – YYYY DDD

    1963 331 – 2008 042 (IMF)
    1963 331 – 2008 042 (Plasma)
    1967 150 – 2005 365 (Energetic Proton Fluxes)
    1963 001 – 2008 031 (Kp index and Sunspot Number)
    1963 001 – 2003 365 (Dst indices)
    2004 001 – 2006 364 (Provisional Dst)
    1963 001 1988 182 (AE-index)
    1990 001 1994 365 (Provisional AE-index)

    So what is the number for the 27 day average to 2008 042. It is 3.3!

    Furthermore, Dr Svalgaard says that the IMF was not in downtrend during 2007. How do you explain the third chart in Svalgaard 2 post 80? It looks like a downtrend. By your own post 120, the IMF started 2007 at 5.1 and finished at 4.4. I resist the temptation to calculate the percentage. There were bigger numbers at the beginning of the year and smaller numbers towards the end – thus a downtrend. Where is it going to stop? By the way, some Canadians are worried sick about the low magnetic readings: http://www.popularmechanics.com/science/earth/4248062.html

    The relevant quote from that article is: “Recent magnetic field readings are as low as he’s ever seen, he says, and he’s worked with the instrument for more than 25 years. If the sun remains this quiet for another a year or two, it may indicate the star has entered a downturn that, if history is any precedent, could trigger a planetary cold spell that could bring massive snowfall and severe weather to the Northern Hemisphere.” The man is on message.

  130. Posted Mar 20, 2008 at 7:43 AM | Permalink

    129 (DA):

    The relevant quote from that article is: “Recent magnetic field readings are as low as he’s ever seen, he says, and he’s worked with the instrument for more than 25 years. If the sun remains this quiet for another a year or two, it may indicate the star has entered a downturn that, if history is any precedent, could trigger a planetary cold spell that could bring massive snowfall and severe weather to the Northern Hemisphere.” The man is on message.

    … but the above paragraph doesn’t say anything. Saying magnetic readings are the lowest in 25 years has what relevance? If he can prove through proxies that they are the lowest in 500 years, I’ll sit up straighter in my chair, but 25 years of data cannot possibly prove the point either way. As with AGW-themed articles, I cringe at the use of code words such as, “may”, “could”, “might”, and so forth. A scientific treatise should not include those words, as they bring conjecture into the facts and figures.

  131. Posted Mar 20, 2008 at 8:29 AM | Permalink

    129 (DavidA):

    Let’s look at the table that we get from the OMNI data a bit more closely:
    2008 16 0 2381 27 27 4.6 4.3 499.
    2008 43 0 9999 1 3 3.3 2.3 638.

    The first line says that 27-day rotation #2381 starts on day 16 of 2008, has 27 days worth of IMF data with an average of 4.6. The second line says that the next rotation starts 27 days later on day number 43, but that it only has 1 day of IMF data with value 3.3. Since this is less than a rotation, NASA tries to tell you by the designation 9999 that the 3.3 is not a valid rotational average, but, evidently, they failed.

    About trends: I said that there is no trend since the start of 2007. This means two things:
    1) the time interval is from Jan. 1, 2007 to the end of the data which seems to be near Feb. 12, 2008, so one should not take the 4.4 for the last rotation of 2007 as the last point for the trend, but rather the 4.6 for rotation 2381. And you don’t calculate trends by taking the difference between just the last and the first point. I could have selected rotation 2368 instead of 2367 [on the criterion that I would the last year's worth of data] and you would have had a starting value of 4.5 and and ending value of 4.6.
    2) the trend has to be statistically significant and this one is not. Furthermore, the two points above 5 nT were caused by CMEs that ejected extra closed flux and do not add to the open magnetic flux important for solar minimum conditions. You can see this from the following plot that also shows the sunspot number.

    Of course as we go from solar max towards solar min, the field goes down as the sunspot number declines. The important thing is that now when there are hardly any spots left, the IMF has stopped declining, has hit the floor and will stay there until new spots [if any :-) from cycle 24] will produce CMEs that will add closed flux to the IMF and drive the values up again.

    Re open/closed flux: There is no ‘open’ flux according to Maxwell’s equations, but we say that IMF flux is closed if we can identify where both ‘ends’ of the field line are on the Sun, and that the flux is ‘open’ if we cannot. The closed flux looks like magnetic bottles or tongues, whereas the open flux, e.g. in the polar regions don’t have any clear looping back to the other pole.

  132. Posted Mar 20, 2008 at 8:52 AM | Permalink

    130 (tucker): At least DavidA does not use weasel words like “may”, “could”, “might”, but boldly uses “must”, “will”, and “absolute certainty”. :-)

  133. Posted Mar 20, 2008 at 9:17 AM | Permalink

    Keeping track of the IMF: last 7 days
    If you want the last 24 hours replace 7d in the URL by 24d
    The white curve is total IMF B. The blue curve shows the direction of the field. The orange curve is the density, the yellow is the speed, and the green is the temperature. The red curve is the so-called ‘southward component’ that is important in feeding energy into the Magnetosphere.

  134. Posted Mar 20, 2008 at 9:20 AM | Permalink

    Another wonderful real-time plot is this one

  135. Posted Mar 20, 2008 at 9:57 AM | Permalink

    133 (me): 24h not 24d . You can also do 2h, 6h, and 3d, with the obvious result. A small warning: the data is estimated only and the OMNI site is still the place to get the final numbers.

  136. Jim Arndt
    Posted Mar 20, 2008 at 11:31 AM | Permalink

    Hi,

    Leif, here is the latest from NASA. SC24 in May 2008. Also I have a graph of the geomagnetic index and notice in 2005 there is a sharp drop and then a continuation at low level and less variability. I find this interesting and would like to know if it compares to others minimas. Also do you think that when large CME’s hit the atmosphere that the high energy particals may act to inhibit the formation of nuclei necessary for cloud formation. Meaning that it acts to break the nuclei apart.

  137. Posted Mar 20, 2008 at 12:32 PM | Permalink

    136 (Jim A): The sharp drop is quite common [e.g. in 1911-1912, 1922-1923, 1964-1965] (but not universal), and doesn’t mean much [it has to come down eventually - sometimes that is abrupt and sometimes not]. The large CMEs do not deliver enough energetic particles per se to do much of anything because they are short lived and in energy usually fall far short of Galactic Cosmic Rays. [On very rare occasions we get strong Solar Energetic Particles from major solar flares].

  138. Posted Mar 20, 2008 at 1:06 PM | Permalink

    136 (Jim A):

    here is the latest from NASA. SC24 in May 2008

    Now, where get you get that from? I’m on the NASA panel and this is news to me.

  139. Pete
    Posted Mar 20, 2008 at 1:31 PM | Permalink

    138 (Leif): Now, where get you get that from? I’m on the NASA panel and this is news to me.

    You have to fit the puzzle together here Leif! (No one told you!) The root link to the graph from Jim #136 is here WattsUp

    This links to the updated prediction data here

    …..which is plotted for us with May 2008 appearing to be the Min, as marked on the graph:

  140. Jim Arndt
    Posted Mar 20, 2008 at 1:39 PM | Permalink

    Hi,

    Leif, 138 – I believe Hathaway released this. Maybe of his own not sure. Here is the source. According to the post is was from a meeting just a few days ago.

    http://wattsupwiththat.wordpress.com/2008/03/20/new-solar-cycle-24-goalpost-established/

    http://solarscience.msfc.nasa.gov/images/ssn_predict.txt

  141. See - owe to Rich
    Posted Mar 20, 2008 at 1:54 PM | Permalink

    138 (Leif)

    Ah, you _were_ on the NASA Panel. Sorry, I shouldn’t be so wicked…

    127 (Andrew)

    You know, I don’t think Milankovitch cycles are 100000 years – Wikipedia says 21K, 26K, and 41K. A cycle which is supposedly 100K is Earth passing through the mean plane of the solar system, which supposedly means we accumulate more dust, which is excellent for cloud making and glacial cooling.

    Rich.

  142. Posted Mar 20, 2008 at 1:56 PM | Permalink

    139,140 (Pete,Jim A): looks like Hathaway’s own adaptation. He is getting a bit “uncomfortable” with the lack of SC24 upswing, but, hey, all that can change next month [not this one because STEREO already looks ‘around the corner’]

  143. Jim Arndt
    Posted Mar 20, 2008 at 2:12 PM | Permalink

    Hi,

    Leif, 142 – Since SC23 number are similar to SC13 and SC15, isn’t possible SC24 wont start until May 2009 and be in the low 60′s SSN. SC23 could possible go 13 to 14 years.
    See Clilverd forecast 42 SSN.

    http://users.telenet.be/j.janssens/SC24Clilverd.pdf

  144. Posted Mar 20, 2008 at 4:01 PM | Permalink

    143 (Jim A): first, respect the numbering convention: “142 (Leif):”
    Second, yes that is possible. The real question is: “is it plausible?” and here I’m less sure. The polar fields right now point to max at 70; should they weaken, the prediction might go lower, but past experience says that they don’t begin to weaken before the next cycle is well under way. So, I don’t think we’ll go all the way to Clilverd’s number, and I certainly do not buy his cyclomania leading him there.

  145. Jim Arndt
    Posted Mar 20, 2008 at 4:27 PM | Permalink

    144(Leif): Thanks for the reminder I tend to forget that. Yea I wish the cycles where so cut and dry, it would make life easy. I have found that the only correlation is there is no correlation.

    If I read it right then you say the polar fields will only weaken after the the cycle starts, is that correct? Hathaway says that SC25 will be weak because of the conveyor belt is slowing, has that changed or has it slowed faster than thought. I haven’t seen any new information on that recently. Could that cause SC23 to drag out longer.

  146. Posted Mar 20, 2008 at 5:07 PM | Permalink

    145 (Jim A): The polar fields weaken for two reasons [AFAIK]:

    1) magnetic fields are due to currents that decay with time [Ohmic dissipation into heat]. This is a slow process [on the Sun] and would not be very important for a single cycle, but during a Grand minimum it might matter. I’ll not go into the math [it is not correct, anyway, IMHO], but traditional wisdom holds that the decay time for a sunspot is of the order of 1000 years and that of a ‘general’ field longer than the age of the Sun.

    2) opposite polarity is brought by the ‘conveyor’ belt up from the sunspot zones and cancels out the existing polar fields. This is thought to be the process by which they reverse. This process in turn depends on the speed of the belt and on the existence of opposite flux [i.e. cycle 24 spots]. No SC24 spots and it doesn’t matter how fast or slow the conveyor belt is. The length of SC23 has almost nothing to do with any of this. SC23 just fizzles on its own account. SC24 magnetic fields are already generated [we think] but are sitting at depth and are in the process of rising to the surface. When they reach the surface they form spots and we have SC24 and the length of SC23 is then just determined by when SC24 starts. This in turn depends [we think] on how strong the new magnetic flux at depth is, how it is organized [into thin strands or huge flux tubes or whatever], and how deep they are. None of these things are known, which, of course, does not keep people from making strong statements to the contrary [just as in the climate debate].
    As the fields reach the surface they need to assemble into a spot. Here is a link to a movie showing how this happens. There is one little thing wrong with the info you get at the link, namely that “We’ve never seen anything quite like it”. This has been known for over a hundred years, but has been overshadowed by the theoretical “knowledge” that sunspots form when a huge magnetic tube breaks through the surface and you see two big spots with a big loop connecting them. This is what people would like to see because it makes ‘sense’. It is however not what happens. My good friend Ken Schatten has perhaps already explained what really happens. Here is his percolation theory.

  147. Posted Mar 20, 2008 at 5:35 PM | Permalink

    146 (me): When you watch the movie, concentrate on the left-hand portion. After a while you’ll see a lot of little black AND white elements pop up. These have opposite polarities. Watch them neatly assemble themselves into a large white spot moving right and a black spot moving left. Near the end of the movie there is another outbreak. Once a spot has reached a large size it begins to disintegrate again: watch the little black and white elements move away from the spot.

  148. Pete
    Posted Mar 20, 2008 at 5:44 PM | Permalink

    145(Jim A): “Hathaway says that SC25 will be weak because of the conveyor belt is slowing, has that changed or has it slowed faster than thought. I haven’t seen any new information on that recently. Could that cause SC23 to drag out longer.”

    I was also wondering about this recently, and I think Leif forgot to answer :) In the report Hathaway published in May 2006 he said the Southern Conveyor was down to 0.35 m/s and the Northern down to 0.75 m/s, and I quote “We’ve never seen speeds so low” They have been “around 1 m/s since the 19th Century”. They usually turn on a “40 Year Circuit”.

    Link To Hathaway Conveyor here

    I am also curious, how he measured these Conveyors?, and what is the current status? Are they still slow, or unchanged, or faster!?

  149. maksimovich
    Posted Mar 20, 2008 at 6:07 PM | Permalink

    re 136(Jim)

    Also do you think that when large CME’s hit the atmosphere that the high energy particals may act to inhibit the formation of nuclei necessary for cloud formation. Meaning that it acts to break the nuclei apart.

    A more qualitative description would be Solar proton events(SPE).Jackman (2007) has an interesting perspective.

    Solar eruptions sometimes produce protons, which impact the Earth’s atmosphere.These solar proton events (SPEs) generally last a few days and produce high energy particles that precipitate into the Earth’s atmosphere. The protons cause ionization and dissociation processes that ultimately lead to an enhancement of odd-hydrogen and odd-nitrogen in the polar cap regions (>60 geomagnetic latitude). We have used the Whole Atmosphere Community Climate Model (WACCM3) to study the atmospheric impact of SPEs over the period 1963–2005. The very largest SPEs were found to be the most important and caused atmospheric effects that lasted several months to 10 years after the events. We present the short- and medium-term (days to a few months)
    atmospheric influence of the four largest SPEs in the past 45 years (August 1972; October 1989; July 2000; and October–November 2003) as computed by WACCM3 and observed by satellite instruments. The polar effects can be summarized as follows:

    1) Mesospheric NOx (NO+NO2) increased by over 50 ppbv and mesospheric ozone
    decreased by over 30% during these very large SPEs;
    2) upper stratospheric and lower mesospheric NOx increased by over 10 ppbv and was transported during polar night down to the middle stratosphere in a few weeks;
    3) mid- to upper stratospheric ozone decreased over 20%; and
    4) enhancements of HNO3, HOCl, ClO, ClONO2, and N2O5 were indirectly caused by the very large SPEs, although the model results suggest impacts at higher altitudes than indicated by the measurements for the October–November 2003 SPE period.

    An interesting observation is Table 1 shows the magnitude of the fifteen largest individual SPEs, in terms of the computed middle atmospheric NOy production, during the past 45 years.Note that eight of these event periods occurred during the current solar cycle.

    These events perturb the atmospheric system for reasonable periods,this is a complication when quantifying(measuring) the persistent regime(which is already far from photochemical equilibrium ie in a continuous state of self reorganization )

  150. Posted Mar 20, 2008 at 6:11 PM | Permalink

    148 (Pete):
    1) that SC25 will be low “because the conveyor belt has slowed” is on the assumption that I do not share that the dynamo is deep and that the ‘turn-around’ is 20 or more years [two cycles].
    2) “We’ve never seen speeds so low”, around the turn of the 20th Century, the cycles were also very small, yet, the speed was not, apparently. The speed is measured by the drift of magnetic flux [spots] at the surface which is much larger, about 20 m/sec. Under the assumption that the return flow takes place so deep in the sun that the density is 10 times higher, the speed down there will be ten times as slow [it's the same mass moving around]. That is where he gets the 1 m/sec or lower.
    3) “They usually turn on a “40 Year Circuit”, well, I thought Dikpati said 17-22 years, but what is factor of two among friends…
    4) Current status? Nobody knows, because Hathaway has not taken the trouble to measure the speed again; maybe because “it is off the charts”? :-) Also, Dikpati et al. have not tried to obtain newer data [since 2002 or 2003] for their prediction. In both cases I feel that the people involved are a bit uneasy about rocking the boat with new data. [we see same attitude in climate research where the ancient Hoyt & Schatten and Lean TSI reconstructions are still being used today when they fit some pet idea].

  151. Posted Mar 20, 2008 at 6:38 PM | Permalink

    149 (maks):
    1st, try to use the standard quoting scheme “nbr (name):”

    2nd, And let’s not confuse CMEs and Major Solar Flares. They have nothing to do with each other, other than both being accidental effects [that may or may not both occur] of a rapid reorganization of the solar magnetic field [very often, but not always, near an active region].
    3rd, the 10 year duration of the effects of an SPE I have to check, I didn’t think so.

    Ed Cliver and I pointed out in our cycle 23 prediction paper [2005] that with a low cycle 23 “Average space weather might be ‘‘milder’ with decreased solar activity, but the extreme events that dominate technological effects are not expected to disappear. In fact, they may become more common. Two of the eight strongest storms in the last ~150 years occurred during solar cycle 14 (Rmax = 64) [Cliver and Svalgaard, 2004], while three of the five largest 30 MeV solar energetic proton events since 1859 [McCracken et al., 2001] occurred during cycle 13 (Rmax = 88). McCracken [Space Weather, 2007] proposes that only if solar activity is really low, do we get the monster events.

    Bottom line is that big events do not seem to follow solar activity as much as one intuitively [and automatically] might presume.

  152. Posted Mar 20, 2008 at 6:58 PM | Permalink

    151 (me): I must be slipping :-) The prediction paper was, of course, for cycle 24, not cycle 23. There is an interesting thing about 23 though that is relevant to posts #151 and #149, namely that the IMF for cycle 23 matches that for cycle 13 very closely. The sunspot numbers for the too cycles are also close [after correction for the ~20% jump that Waldmeier introduced in 1947 :-) ]

  153. maksimovich
    Posted Mar 20, 2008 at 8:26 PM | Permalink

    151(LS):

    The “bottom line” is you either did not read the bottom paragraph,or indeed understand the importance.

    That is nothing unusual,understanding of non-linear systems in competition is not “intuitive” in western science,in Russian mathematical science it has been for 100 years.

    Ruzmaikin succinctly states this in a recent paper.

    Linear and non-linear systems respond differently to external forcing. A classical example of a linear system response is the Hooke’s law of elasticity that states that the amount by which a material body is deformed is linearly proportional to the force causing the deformation. Earlier climate change studies used this linear approximation to evaluate the sensitivity of the global temperature change caused by external forcing. However the response of non-linear systems to external forcing is conceptually different; the issue is not a magnitude (sensitivity) of the response. Non-linear systems have internally defined preferred states (called attractors in mathematics) and variabilities driven by residence in the states and transitions between them. The question is what is the effect of an external forcing: change of the states, residence times or something else? Answer to this question is critical to our understanding of climate change.

    Based on the model studies mentioned above we can formulate the following, updated conjecture of the climate system response to external forcing: external effects, such as solar, the QBO and anthropogenic influences, weakly affect the climate patterns and their mean residence times but increase a probability of occurrence of long residences. In other words, under solar or anthropogenic influence the changes in mean climate values, such as the global temperature, are less important than increased duration of certain climate patterns associated say with cold conditions in some regions and warm conditions in the other regions

    There is a very important message here,understanding these philosophical attributes is a key to understanding what are the natural variations,what are aperiodic,and what is anthropic.

    Indeed in photochemical reactive-diffusive systems the “rules of the game” are “counterintuitive” to what you are stating to some commentators.”For example, some chemical reactions are provoked only by light of frequency higher than a certain threshold, light of frequency lower than the threshold, no matter how intense, does not initiate the reaction.”

    This is an important “axiom” well understood at the molecular level,and here the “diophantine approximation industry” does not cut the mustard.

  154. Posted Mar 20, 2008 at 8:50 PM | Permalink

    153 (maks): Of course I agree with you and my good friend Alexander Ruzmaikin on the ‘linearity’-problem in Western Science.
    And I maybe not have made myself clear. I said “Bottom line is that big events do not seem to follow solar activity as much as one intuitively [and automatically] might presume.” What I meant to say was that I and my colleagues think that it is clear that big events do not necessarily follow linearly from big cycles, but that we actually have a hard time convincing most ‘Western’ scientists of that because they automatically think so. Having said that, your post might have to cross a rather large gulf to be comprehensible to most of the readers here.

  155. maksimovich
    Posted Mar 20, 2008 at 9:42 PM | Permalink

    154(LS):Agreed,sometimes intent ‘gets lost in translation”

    As a suggestion some commentators when confounded by some mathematical terminology,or phrase especially with non-linear systems, rather then wikipedia they try scholarpedia,where often the original authors can be found, with applications that tend to allow a more simplistic understanding(and limitations especially in the microstate).

  156. Patrick M.
    Posted Mar 21, 2008 at 7:37 AM | Permalink

    re 118 (Leif):

    Do you have a link to the paper?

  157. Posted Mar 21, 2008 at 7:57 AM | Permalink

    156 (Patrick M): yes, it is pay-walled, but here it is.

  158. Posted Mar 21, 2008 at 8:12 AM | Permalink

    157 (me): The main Figure is this:

    (a) a14Cprod (blue) vs. δ18Ostal (green) as a time-series for the period 2.36 to 2.86 ka; (r2 = 0.563). (b) Moving the δ18Ostal data backwards in time by 50 years resulted in a maximum r2 value of 0.609.

    The main contribution to the correlation comes from the peak at 2.7 ka BP. I’m not terribly impressed.

  159. See - owe to Rich
    Posted Mar 22, 2008 at 3:55 AM | Permalink

    Svalgaard #3 685 (Archibald) et seq

    David,

    I have somewhat belatedly perused your paper linked from #685, though as I expected it is an updating of your ones from last year. You could regard me as someone who is “copying your methods” , as you will see from my article here, but I don’t currently have any plans to make a peer-reviewed paper out of it.

    Though I support the cycle-length theory (and have promised Leif I would research any possible connection with cycle sizes), in my article you will find that I have criticized your temperature predictions from the assumed linear relationship. Before I explain that further here, I want to ask you a question about the graphs you show for New England sites, for example Hanover NH. What smoothing filter did you apply to the solar cycle length data for these examples? Was it the 1+2+2+2+1 quoted in Butler & Johnson, and if so how is that filter aligned to the temperature data being fitted? (I previously tried, but failed, to reverse engineer this from the B&J paper.)

    If a 1+2+2+2+1 filter is being applied, then please observe the following. Suppose the slope is measured at 0.7 degrees per year of cycle, as you claim for some examples. Then this means that

    T = a + sum_{i=1}^5 0.7b(i)(11-c(i))/8

    where T is predicted temperature, a is a constant (intercept), b(i) = 1,2,2,2,1, and c(i) is the corresponding cycle length. The reason for the 8 is to normalize by the sum of the b(i).

    Now, let us assume that your estimate for the length of Cycle 23, which you have resolutely stuck by on 13 years for some time, and which I copied in my article, is correct. The previous cycle lengths were 9.7, 10.3, 11.6, 10.7, 10.0. Then to predict Cycle 24 temperatures we have to apply this formula, and we have to know how to align it. My reading of Lassen & Friis-Christensen is that the alignment would require us to know the lengths of Cycles 24 and 25 (and 26?), but let’s not go there.

    Suppose it really did align with the end of the filter pointing at Cycle 23. Then we get

    T(24) = a + 0.7(-2.0+2.6+1.4-1.2+0.3)/8 = a + 0.10,
    T(23) = a + 0.7*( 1.3+1.4-1.2+0.6+1.0)/8 = a + 0.27

    This predicts a cooling from Cycle 23 to 24, but only by 0.17C, much much less than your prediction of 2.2C for Hanover.

    If I have used the wrong filter or alignment, then please advise and we can recalculate. I urge you to look into this and revise future versions of your paper in this respect, if I am proved correct. In respect of global cooling, we wouldn’t want to be alarmist, would we :-)

    Thank you for cheering me up with the Usoskin article at #119 (modulo Leif’s cold water).

    Rich.

  160. David Archibald
    Posted Mar 22, 2008 at 5:49 AM | Permalink

    Re 159, Rich, thankyou for having a look at the Hanover, NH graph. What I did in that graph was plot the length of a solar cycle against the average temperature of the following solar cycle. There was no filtering. There is no need for filtering that I can see. I included three US sites to provide local relevance for the New York conference. They are the three longest series that I have found to date in the US that provide a statistically significant signal. I have now worked through Europe. Sites closer to the Arctic Circle provide a better response than those further away. In graphical presentation, dispersion from the linear trend is mostly on the low side, which means volcanoes perhaps.

    Anybody can repeat what I have done on Hanover – it is simple enough for highs school students to do.

    By the way, the one year average decrease in solar cycle length in the 20th century relative to the 19th century explains all of that warming of the 20th century by the relationship demonstrated by the Hanover data.

    In terms of 20th century warming, why didn’t you use the last 30 years of satellite data, which shows no warming over that period. As shown by Hansen’s fiddling of the Peruvian data etc, the Hansen temperature series has no credibility. It would be better to splice the satellite data to a pre-1978 data set. I think that would show the 1930s as warmer than present.

    On the length of Solar Cycle 23, I simply reverse engineered it from Clilverd’s estimate of the amplitude. He has a slightly higher amplitude for Solar Cycle 25. Two consecutive cycles of an amplitude of around 45 is a repeat of the Dalton Minimum – solar cycles 5 and 6. What was the length of Solar Cycle 4? It was 13.6 years. Hathaway found a weak, inverse relationship between solar cycle length and the amplitude of the following cycle, so there is a physical basis in the relationship.

    Clilverd’s estimate is within the error bar of Schatten’s estimate, so wavelet does not preclude solar dynamo theory. Solar dynamo theory is a simple ratio affair. I applied it to polar faculae (the poor man’s magnetometer but more up to date) and came up with a Solar Cycle 24 amplitude of 45. The general shape of the solar cycles in the 20th century is similar to the general shape of them in the 18th century.

    The beautiful thing about the solar cycle length/temperature relationship is that you don’t have to know how it works to use it. Every day’s delay in the onset of Solar Cycle 24 will lower the temperature of Hanover, NH by 0.002 degrees. It is a freight train, it is unstoppable, remorseless and most dreadful in its certainty.

    As my posting of graphs has been fraught with difficulty, I will discuss your CO2 sensitivity separately.

  161. David Archibald
    Posted Mar 22, 2008 at 6:15 AM | Permalink

    Re 159, Rich, the eminent climate scientists, Lindzen, Kininmonth, and Bill Gray, say that the warming will be about half the Stefan-Boltzman figure. Based on Idso’s observations from Nature, I derived about the same. The IPCC need positive water vapour feedback and get up to 6.4 degrees. I imagine that there would have been a special dispensation for the scientist who sold his soul for that figure.

    As I said in my presentation, as the whole AGW theory hinges on whether or not it is positive or negative feedback, perhaps this should be examined by someone. Well, somebody did, and it was Roy Spencer. Hopefully above there should be the last slide from his presentation in New York. Using data from the Aqua satellite, Roy Spencer derived that the warming will be about 0.5 degrees. The paper was published in GRL and written up on CA last September, I think.

    The significance of it is that up to now, the IPCC modellers have been guessing, and justifying that guessing on the lack of hard, real world data. Roy Spencer’s work on the data from the Aqua satellite provides that hard, real world data. The whole, rotten IPCC edifice comes tumbling down.

    In the real world, the CO2 warming effect is a fraction of solar variability, and lost in the noise of the system. It is extremely minute in the decadal scale of your modelling.

  162. Posted Mar 22, 2008 at 8:06 AM | Permalink

    160 (David A):

    I applied it to polar faculae (the poor man’s magnetometer but more up to date) and came up with a Solar Cycle 24 amplitude of 45.

    There is no need to be the poor man. High-quality polar field measurements are available since 1976 on the Wilcox Solar Observatory website here. The Data is updated about every ten days [last time March 4th 2008]. One little caveat: the data for 1976 and 1977 should be increased by ~20% as they were contaminated by scattered light].

  163. Posted Mar 22, 2008 at 10:07 AM | Permalink

    162 (me): David, if you use the real polar fields from Wilcox Solar Observatory, what maximum for SC24 do you then “come up with”?

  164. duwayne
    Posted Mar 22, 2008 at 3:38 PM | Permalink

    I’ve seen references to a forecast by the Russian scientist Khabibullo Abdusamatov of a significant decrease in Solar Activity starting within the next 5 years. Has this been discussed somewhere in this very lengthy string?

  165. Posted Mar 22, 2008 at 4:52 PM | Permalink

    164 (duwayne): Abdusamatov is of the growing school that thinks that solar activity is decreasing. I [and others] happen to think so too, but that does not mean that all the rest Abdusamatov claims is automatically endorsed by us. Few details have surfaced, except that Abdusamatov has advised Putin not to sign various treaties, so it is hard to form an informed opinion and assessment.

  166. Posted Mar 22, 2008 at 4:56 PM | Permalink

    164 (duwayne) Abdusamatov seems to be associated with John Casey and the dubious SSRC which is not a good sign.

  167. David Archibald
    Posted Mar 22, 2008 at 7:47 PM | Permalink

    Re 163, Dr Svalgaard, the answer is 60.

  168. Posted Mar 22, 2008 at 8:12 PM | Permalink

    167 (David A): 1: please respect the standard numbering scheme: “nbr (name):”.
    2: could you tell us the procedure or reasoning leading to 60?
    3: Would 60 or something else make any difference compared to the prediction of 45?

  169. Erl Happ
    Posted Mar 23, 2008 at 7:24 AM | Permalink

    Leif
    Thanks for your continued support of this discussion. Good to see Svalgaard 4 racing away and I look forward to Svalgaard 5.

    In 1976 Wilcox Svalgaard and Scherrer published a paper in the Journal of Atmospheric Sciences entitled ‘On the reality of a Sun-Weather effect’ outlining a connection between the effect of the ‘solar magnetic sector structure as extended away from the sun by the solar wind’ and the strength of a ‘vorticity index’ relating to the area of all the low pressure systems in the northern hemisphere during winter. I understand that an increase in solar wind related to a fall in the vorticity index of about 10% in winter.

    The paper concludes with the recommendation that ‘serious efforts towards understanding the physical mechanism should be made”.

    Does that paper stand scrutiny today? Has the mechanism been identified?

    In my role as armchair theorist I hypothesise that there is an inverse relation between the vorticity index of low pressure systems outside the tropics and the vigour of the low pressure systems within the tropics (20°N to 20°S). This would imply a connection between the solar wind and tropical warming events.

    I further hypothesise that the warming event will be influenced by the location and intensity of the ionospheric equatorial anomaly as it changes over time.

    What other notions might one entertain? I look forward to your thoughts on the matter some thirty-two years after publication.

  170. Posted Mar 23, 2008 at 8:21 AM | Permalink

    169 (Erl):
    The vorticity area effect [VAI] looked pretty solid thirty years ago. Later work showed that the amplitude of the effect varied over time. This is usually perceived as a negative (c.f. the water level in Lake Victoria or frequency of Indian Monsoons) and mainstream science quietly ‘forgot’ about the effect [euphemism for 'ignored']. Brian Tinsley belongs to a hardy band of supporters who has kept the subject alive [even if on life-support]. Here are a couple of fairly recent papers on the VAI:

    Are stratospheric aerosols the missing link between tropospheric vorticity and Earth transits of the heliospheric current sheet?
    Kirkland, Matt W.; Tinsley, Brian A.; Hoeksema, J. Todd
    Journal of Geophysical Research, Volume 101, Issue D23, p. 29689-29700, 1996.
    DOI: 10.1029/96JD01554
    Abstract
    Evidence has accumulated for the past two decades demonstrating a correlation between Earth transits of the heliospheric current sheet (HCS) and changes in winter tropospheric vorticity. These correlations persisted for a few years following the Agung and El Chichón volcanic eruptions, but were significantly weaker at other times. This suggests that the missing link in a physical mechanism explaining the correlation may involve volcanic aerosols and their effect on cloud microphysics, via atmospheric electricity. An analysis of 500-mbar northern hemispheric vorticity for the 1991-1994 winter periods following the Pinatubo eruption shows a similar correlation between tropospheric vorticity and Earth transits of the HCS, supporting the previous interpretation.

    Atmospheric Dynamical Responses to Solar Wind Variations on the Day-to-Day Timescale.
    Tinsley, B. A. [Tinsley@UTDallas.edu]
    American Geophysical Union, Fall Meeting 2001, abstract #A11A-0031, 2001

    Abstract
    In the early 1970s it was shown by John Wilcox and associates and verified by Colin Hines that the strength of winter storms across the northern hemisphere decreases at times of solar wind ‘sector boundary’ crossings. These are now known as heliospheric current sheet (HCS) crossings, and correspond to the extension of the coronal streamer belt passing over the Earth, with reductions in solar wind velocity by about 10%. The strength of winter storms is objectively evaluated by the vorticity area index (VAI) calculated from gridded geopotential height data sets. It was shown by Tinsley, Hoeksema, Baker and Kirkland in the mid-1990s that this VAI response (the Wilcox effect) tracks the decrease in MeV electron flux precipitating from the magnetosphere, with a lag of less than a day. The MeV electron flux is strongly correlated with solar wind velocity, and together with its associated X-ray Bremsstrahlung modulates the stratospheric vertical column resistance. For winters when the column resistance is unusually high because of a high mixing ratio of H2SO4 from volcanic eruptions, the ionosphere-earth current density Jz is modulated by these stratospheric resistance variations. The winter storm response can be understood in terms of a general theory (Tinsley, Space Sci. Rev., 94, 231-258, 2000), and it involves changes in cloud microphysics and precipitation from the storm systems at mid-high geomagnetic latitudes, due to electroscavenging by cloud droplets. The electroscavenging rate tracks the Jz changes. There is a similar effect from reductions in tropospheric resistance associated with changes in cosmic ray flux during magnetic storms, first noticed in
    the 1960s by Walter Orr Roberts and associates. The Roberts and Wilcox effects are part of a more general influence of solar activity affecting Jz and clouds, which is part of an even more general influence of electroscavenging on clouds. The precipitation changes associated with cosmic ray and Jz changes have been evaluated by Kniveton and Todd (GRL 28, 1527-1530 and 3279). In winter storms the VAI response arises from a redistribution of vorticity within the storm because of diabatic heating changes. The storm vorticity changes have longer term dynamical and climatic consequences.

    ——-

    I don’t know what my own view on this would be today, having not followed it closely enough. But my involvement with this work might help dispel the myth that I’m a ‘solar denier’ :-)

  171. Posted Mar 23, 2008 at 8:49 AM | Permalink

    167,168 (me,David A):
    Credibility is often closely associated with openness in data and methods. I’ll here show how I arrive at my prediction [Rmax = 72] for SC24. We assume that the polar fields are the seeds for the next cycle. Of course, one does not just take the polar fields on any given day, but over a reasonably long period [~3 years] before and around solar minimum. As a measure of the strength of the solar polar fields we take the absolute value of the difference between the North and South polar fields. This has two advantages: 1) minimizes zero-level errors, and 2) removes the annual effect of the Sun’s axis’ apparent ‘tipping’ to and fro by 7 degrees:

    Also assuming that if there is no seed there will be no spots we include the origin [0,0] and calibrate the relation to get Rmax = 0.65 PF, which for the present value [111] of the PF yields Rmax24 = 72.

    Any ‘error bars’ will not be in the measurements [where they are practically zero], but in the assumptions [where we don't know what they are]. Often +/- 10% is used for [other] predictions so if we assume that we are no worse or better, we might go with that and end with 72+/-8.

  172. John Lang
    Posted Mar 23, 2008 at 9:28 AM | Permalink

    Leif, so far most of these threads have focussed on sunspots and the solar cycle. But there are other variations in solar output over time.

    Judith Lean’s reconstruction of total solar irradiance show an increase in solar energy reaching Earth of about 3 W/m2 since the Maunder Minimum and around 2 W/m2 since 1900. (As an aside it seems solar irradiance has been falling since about the year 2000 and certainly in the latter stages of cycle 23.)

    Given the Earth is a sphere, we take 1/4 of those changes or a total increase in solar energy of 0.75 W/m2 since 1700 and 0.5 W/m2 since 1900.

    The range of Earth’s temperature sensitivity to changes in W/m2 is between 0.3C to 1.1C per 1 W/m2. So just changes in the Sun’s energy reaching Earth since 1900 could be responsible for up to 0.6C change in temperatures since 1900.

    I wonder if you could comment on this.

  173. Posted Mar 23, 2008 at 9:58 AM | Permalink

    172 (John L):

    Judith Lean’s reconstruction of total solar irradiance show an increase in solar energy reaching Earth of about 3 W/m2 since the Maunder Minimum and around 2 W/m2 since 1900.

    John, the reason we have focused on sunspots and the sun’s magnetic field is that those where what Lean and others were basing their reconstructions on. So, the two sides of solar output are strongly linked. Note, that Lean herself [with Wang, 2005] has published later reconstructions where the change since the MM was only 1 W/sqm and since 1900 only 0.5 W/sqm, effectively halving the increase you calculate. The change since 1900 [and MM for that matter] is probably only half, again, of the Wang et al. [incl. Lean] number, yielding only a 0.08-0.30 degree temp increase [for the range of sensitivity you quote]. The reason for this smaller rise since 1900 is the same as for the MM: a reassessment of the sun’s magnetic field. So, there you have the reason for the obsessive focus on sunspots and the like.

    I have shown it before, but here are several TSI-reconstructions:

    The one I put most trust in is the pink curve by Dora Preminger [and my own red curve :-) ]

  174. Posted Mar 24, 2008 at 9:30 AM | Permalink

    These are exciting days for solar science. Not a time to dig into entrenched positions, but rather a time to watch closely for a chance to advance the state of the science. Recent ocean data suggesting an unexplained ocean cooling–coupled to a difficult to explain solar downturn–should cause all scientists involved in climate to put on a bit of humility and quiet expectation while watching for new clues.

    Thomas Kuhn’s observation that older scientists are too set in the old theories to make many new contributions to the science, seems to be operating here. When a scientist working in one field (solar physics) falls back on data from other fields as soon as the data stops going his way, that suggests something deeper and more emotional tying him to a fixed POV.

  175. Mark T.
    Posted Mar 24, 2008 at 10:02 AM | Permalink

    Interesting that a lack of activity is seen as “exciting.” :)

    Don’t get me wrong, I agree on a scientific basis that this lack of activity is exciting in the sense that there’s a lot of new ground to be discovered. There’s simply some humor in the apparent irony (hmmm, incongruity?).

    Mark

  176. See - owe to Rich
    Posted Mar 24, 2008 at 10:41 AM | Permalink

    David, here are a few comments on your responses.

    Re 159, Rich, thankyou for having a look at the Hanover, NH graph. What I did in that graph was plot the length of a solar cycle against the average temperature of the following solar cycle. There was no filtering. There is no need for filtering that I can see. I included three US sites to provide local relevance for the New York conference. They are the three longest series that I have found to date in the US that provide a statistically significant signal. I have now worked through Europe. Sites closer to the Arctic Circle provide a better response than those further away. In graphical presentation, dispersion from the linear trend is mostly on the low side, which means volcanoes perhaps.

    - Thanks for clarifying that; I’m pleased to see there isn’t a filter, especially one using future data. It may be worth you emphasizing that in your papers/presentations, as it differs from Friis-Christensen&Lassen and Butler&Johnson. I am a bit concerned that if you are only showing series which provide a statistically significant signal, then you will get accused of cherry-picking.

    Anybody can repeat what I have done on Hanover – it is simple enough for highs school students to do.

    - Yes; so would you be willing to share your data? To reciprocate, here is what I used for Armagh:
    ymax = c(1760, 1770, 1779, 1788, 1802, 1814, 1827, 1838, 1849, 1860, 1870, 1883, 1894, 1907, 1917, 1928, 1937, 1947, 1957, 1968, 1979, 1989, 2000, 2013)
    armcyc = c(NA,NA,NA,NA,8.336364, 8.309091, NA, 8.636364, 9.436364, 9.054545, 9.109091, 8.436364, 8.618182, 8.918182, 8.972727, 9.027273, 9.336364, 9.436364, 9.363636, 9.127273, 9.118182, 9.272727, 9.990909)

    These values are derived from published data, averaged over 11 years centred on ymax.

    By the way, the one year average decrease in solar cycle length in the 20th century relative to the 19th century explains all of that warming of the 20th century by the relationship demonstrated by the Hanover data.
    In terms of 20th century warming, why didn’t you use the last 30 years of satellite data, which shows no warming over that period. As shown by Hansen’s fiddling of the Peruvian data etc, the Hansen temperature series has no credibility. It would be better to splice the satellite data to a pre-1978 data set. I think that would show the 1930s as warmer than present.

    - a. I don’t like splicing data (how does one know one spliced it right?) and b. I’m not convinced it would help that much, as the UAH data (for example) shows a decadal trend of 0.14C, which contradicts your “no warming” statement.

    The beautiful thing about the solar cycle length/temperature relationship is that you don’t have to know how it works to use it. Every day’s delay in the onset of Solar Cycle 24 will lower the temperature of Hanover, NH by 0.002 degrees. It is a freight train, it is unstoppable, remorseless and most dreadful in its certainty.

    - Well, you don’t have to know how a television works to use it. The amount of cooling you state is not certain, because you have not published error bars on your slopes (BTW you should call them “slopes” and not “correlations”). Since your correlation (or R^2) is not unity, there will be error bars. The other thing is, that if we don’t know how it works, we can’t be sure that (because of some weird real explanation) it won’t stop working (though personally I’m fairly convinced).

    Rich.

  177. jae
    Posted Mar 24, 2008 at 11:02 AM | Permalink

    Leif: Are there significant changes in the spectral distribution of wavelengths between solar minima and maxima? (Apologies if this has been discussed; I can’t keep up with the hundreds of comments on these threads).

  178. jae
    Posted Mar 24, 2008 at 11:03 AM | Permalink

    Oh, I meant at Earth’s surface. I now there are UV changes at very high levels.

  179. Posted Mar 24, 2008 at 11:20 AM | Permalink

    178 (jae): UV at the Earth’s surface. The UV that varies a lot with the solar cycle is absorbed high in the atmosphere and does not reach the surface. There might be readers that can give you a better [or dissenting] answer.

  180. Posted Mar 24, 2008 at 12:17 PM | Permalink

    175 (Mark T):

    Interesting that a lack of activity is seen as “exciting.”

    Now, if the lack would be so marked that we are entering another Maunder-type minimum, that would indeed be extremely exciting.
    [Not that I think we will, but solar activity will be low].

  181. Dennis Wingo
    Posted Mar 24, 2008 at 2:06 PM | Permalink

    Leif

    I wanted to throw you a small curve ball to see what you think. Here is a paper that discusses solar variations much larger than anything that has been discussed here. It is in the paleoclimatology area of study so it is of interest here as well.

    http://www.cosis.net/abstracts/EGU05/01910/EGU05-J-01910.pdf

    I don’t have a strong opinion here, just wondering if you had heard of this and had any comment.

    By the way I also think that a Maunder minimum would be very exciting scientifically speaking.

  182. Posted Mar 24, 2008 at 2:32 PM | Permalink

    181 (Dennis): I’m not impressed. The Hickey paper only refers to Nov. 1978-May 1979, and we all know that the 0.4% are day-to-day variations caused by big sunspots, that have compensating brightenings around them. There was a big one in Oct. 2003, too. These have no or little climate impact. The Stuiver & Braziunas paper is an analysis of the 14C record that may or may not have anything directly to do with solar irradiance, and so on.

  183. John Lang
    Posted Mar 24, 2008 at 3:30 PM | Permalink

    Looks like the Sun has finally got going again.

    A few big sunspot areas (Cycle 23) have appeared and the Stereo Behind image shows another one is coming.

    We really don’t want any Maunder Minimums; history can keep its Little Ice Age. There’s too many people of the planet for another one of those right now.

    http://sohowww.nascom.nasa.gov/data/realtime/mdi_igr/512/

    http://stereo-ssc.nascom.nasa.gov/browse/

  184. Posted Mar 24, 2008 at 5:12 PM | Permalink

    184 (kim): yes, this is quite normal, that there are short-lived flare-ups on the way down. Not to worry…

  185. Raven
    Posted Mar 24, 2008 at 5:57 PM | Permalink

    182 (Leif) – Does this group of Cycle 23 spots make it mathematically impossible for the minimum to be March 2008 or earlier?

  186. Jim Arndt
    Posted Mar 24, 2008 at 6:21 PM | Permalink

    185, (Leif): Interesting theory going around that the suns magnetic activity is effecting bee populations. Since the bees use magnetic compasses on cloudy days and they think they are getting lost because of the sudden change in the geomagnetic index. Just a theory though.

  187. Posted Mar 24, 2008 at 6:23 PM | Permalink

    186 (Raven): No, if tomorrow and every day from now on we have a lot of spots (both SC23 and later SC24) so that after another six months we have a hundred spots, then minimum would probably have been last October. But this is not very likely. The way it is going is quite typical, and shows that minimum is still some months away.

  188. Posted Mar 24, 2008 at 6:27 PM | Permalink

    187 (Jim A): solar cycle 23 is just like cycle 13 (107 years ago). The bees survived that one (and millions of other ones before that) as they will this one.

  189. Chris Knight
    Posted Mar 24, 2008 at 6:36 PM | Permalink

    The bees in the UK are grateful for the long 2007 season, and are currently enjoying the long 2008 lie-in due to the delayed spring here. Buzz you in April.

  190. David Archibald
    Posted Mar 24, 2008 at 10:15 PM | Permalink

    Re 168, Dr Svalgaard, I think you answered 2 in 171, and, yes, 45 isn’t much different in the scheme of things from 60.

    Re 176, Rich, my guess is that at least 95% of temperature series do not show a signal because they are too far from the Arctic Circle. That doesn’t make the ones that do any less valid.

    I didn’t centre on ymax. I used the average of the following solar cycle, however long that was. I made up a template in Excel and put each series through it.

    I have said this before, UAH and the other one are flat. Right at the moment they are in the middle of a range that is about 1.5 degrees wide. That is flat. Since you like error bars, it would be instructive to quote the error bars on that one.

    Re your last sentence, that is why I asked at the beginning of my presentation if we lived in a special time in which the laws of physics and Nature have gone haywire. If you have answered yes to that, you might as well give up on everything. You won’t, so the freight train is coming. It’s plaintive whistle can be heard across the plain.

    Both the Butler and Johnson paper and Roy Spencer’s presentation at the New York climate conference are hosted on Warwick Hughes’ blog: http://www.warwickhughes.com/agri/armagh_temp_history.pdf and

    http://www.warwickhughes.com/agri/Recent-Evidence-Reduced-Sensitivity-NYC-3-4-08.ppt

  191. Posted Mar 25, 2008 at 1:20 AM | Permalink

    190 (David A):

    Re 168, Dr Svalgaard, I think you answered 2 in 171, and, yes, 45 isn’t much different in the scheme of things from 60.

    I didn’t centre on ymax. I used the average of the following solar cycle, however long that was. I made up a template in Excel and put each series through it.

    1st: use “168 (Svalgaard):” and not “Re 168, Dr Svalgaard…”
    2nd: you still didn’t explain how you got 60. From what you state, the 60 seems to be the average of the following cycle, not “ymax”. I showed my Excel, now show yours…

    David, these details are important and your credibility stands or falls with getting them right, or at least being able to explain them.

  192. Pete
    Posted Mar 25, 2008 at 3:52 AM | Permalink

    190(David A):

    David, the B&J paper covers the period up to 1992. Purely to eliminate any suggestion that Solar/Temp collelations break-down after the mid-80′s, such as one of the more recent accusations from the F&L Papers as an example, and another example was the CH4 programme The Great Global Warming Swindle, which also used graphs which stopped in the 80′s.

    Have you thought about putting a little extra work into extending the Armagh data to the present day, to see if the plot still has a solid correlation with Cycle Length and Temperature, as shown in the paper up to 1992? Although I appreciate the suggested lag of 5-8 Years, it should be possible, at the least, to extend the graph to 2001?

  193. kim
    Posted Mar 25, 2008 at 7:34 AM | Permalink

    192(Pete): Can we re-visit your graph in comment #454 of the Svalgaard #2 thread? I understand it is speculative.
    ==================================================

  194. David Archibald
    Posted Mar 25, 2008 at 8:26 AM | Permalink

    Re 192, Pete, oh ye of little faith! I have just come back from visiting friends and have two bottles of shiraz in me so we will keep this short. You know what the problem with GISS is? They know that the modern warm period started in 1900 so most of their data series start in 1880, so you only see warming, even though there is data from prior to 1880. There was a time when gentlemen did climate science so stuff was freely available. So from the time that gentlemen were running business there are some data sets that have data from the beginning up to about 1980. Luckily, a friend of mine sent me a bunch of long term European data series.

    You want Armagh, I will give you better than Armagh. Behold the CET, the longest running temperature series on the planet, from 1659 to 2004. The correlation is beautiful. Yes, I could work through the rest and splice on another solar cycle or two from GISS, but it won’t change the outcome. I moved on to mode of action. So you can see I have thought about putting in a little extra work and I don’t like the idea at all. The professors are wanting to progress the cancer stuff, so my climate contribution will fall off somewhat. The chimes strike midnight and I must embrace Morpheus.

  195. Posted Mar 25, 2008 at 8:44 AM | Permalink

    194 (David A): The graph looks impressive, but most of the correlation comes from a single point at the right with a solar cycle length of 15. Referring back to your own table on page 25 of your “Solar Cycle 24: Implications for the United States” paper, there is no cycle since 1660s with such a length. So what value can anyone attach to your graph?

  196. Stephen Richards
    Posted Mar 25, 2008 at 9:34 AM | Permalink

    Lief
    Don’t be so hard on David. After two bottles of red I wouldn’t be able to raise a finger to the keyboard.

    The point on the right seems to stem from the natural distribution of solar length ie. Gaussian? Hence a couple of very short and one very long. If David were to remove the 10% high and the 10% low (sorry short and long) would that be more to your satisfaction?

  197. Posted Mar 25, 2008 at 9:55 AM | Permalink

    194 (David A):

    I have read-off the points on the graph [the best I could] and do not get the same regression line [which could be due to small errors in reading off the numbers], although R2 is not much lower. Omitting the dubious point at length=15 yrs [as there is no cycle with such a length], shows a much smaller [and now not significant] R2 [the red number], indicating that most of the correlation comes from that single point. Also, your graph claims to show the temperatures for the following cycle, although the other graphs of correlations between length and temperature in your paper(s) use the current cycle [as far as one can tell]. It is this kind of “moving target” or inaccuracy that totally destroys any value one might see in the papers.

  198. Posted Mar 25, 2008 at 10:17 AM | Permalink

    196 (Stephen):

    If David were to remove the 10% high and the 10% low (sorry short and long) would that be more to your satisfaction?

    No, because of all the other errors. All I require is just some scientific honesty and disclosure of methods and data, like don’t use 15 yrs when no cycle has that length, or tell us what data leads to the 15 yrs, etc. Omitting 10% from already flawed data is not going to make much difference. In my comment #197 I show the effect of omitting the highest x point [red line and R2 number] and the highest and lowest y points [light blue line and R2 number]. Doing this also destroys the correlation, but I’m basically against such cherry picking a posteriori.

  199. Stephen Richards
    Posted Mar 25, 2008 at 1:14 PM | Permalink

    Thanks Lief

    You answered my question brilliantly. You did exactly what I expected and your result was as I expected. I have always had bad feeling about scatter graphs which are then liearly regressed.

  200. Posted Mar 25, 2008 at 1:43 PM | Permalink

    199 (Stephen): Next time, observe the ‘style’ we have for quoting comments :-)
    Linear regression of a ‘scatter’ graph is not so bad if the correlation is good, but, I agree, is dubious when the correlation is weak.

  201. See - owe to Rich
    Posted Mar 25, 2008 at 3:18 PM | Permalink

    192 (Pete)

    Re Armagh, please see my analysis here, which goes up to 2004. I would prefer comments on it to be over there (to keep stuff together).

    190 (David)

    Thanks, but I still look forward to you sharing your Hanover (and perhaps CET) data, like I did (Leif also).

    Rich.

  202. Posted Mar 25, 2008 at 3:49 PM | Permalink

    201 (Rich): after wading through the login-process, I have this question:
    You say that you have about 12 or 13 degrees of freedom. I take that to mean about 1 per cycle. But your smoothing model decreases the number of freedoms as the “x-values” are no longer independent. Maybe you have only about 6 degrees of freedom [or less]…

    And for David: I suggest a listing with columns for each station like this:

    cycle# length temperature

    and some metadata: Do the temps refer to this cycle, the next, or… Are the lengths from min to min, max to max, smoothed [which model?]. A link to the provenance of the temp data.

  203. Raven
    Posted Mar 25, 2008 at 6:56 PM | Permalink

    Here is an interesting one sided analysis of PMOD vs. ACRIM: http://www.skepticalscience.com/Determining-the-long-term-solar-trend.html

    Leif, do you have an opinion on the two datasets?

  204. Posted Mar 25, 2008 at 7:48 PM | Permalink

    203 (raven): I expressed my opinion in Svalgaard #3 comment#721. I repeat the main points here:

    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.

    Here is one of his Figures:

    Maybe his last point is where I would hang my hat.

    P.S. I know everybody involved in this debate and have discussed the whole thing with them. They still have to convince me one way or the other.

  205. Raven
    Posted Mar 25, 2008 at 8:19 PM | Permalink

    Leif (204): Thanks. For the record I did try to search and see if you had said something earlier. However, the search facilities on this blog are limited and your post did not have the keywords PMOD or ACRIM in it. Perhaps it would be worth having a FAQ with links to the appropriate posts.

  206. David Archibald
    Posted Mar 26, 2008 at 4:06 AM | Permalink

    Re 195, Dr Svalgaard, I realised there was a problem with the Maunder Minimum so I went back to original sources to determine what the solar cycles were. This paper on Japanese cedars provided the data: http://212.191.70.140/29-ICRC/PAPERS/SH34/jap-masuda-K-abs1-sh34-oral.pdf

    There were 5 solar cycles in 70 years for an average of 14 years each. One of them was 15 years long. So you thought that you were quoting my own research at me, but I had already moved on, in teamspeak.

    Re 201, Rich, you can get Hanover from GISS. You don’t need me, all you have to do is make the effort.

  207. Pete
    Posted Mar 26, 2008 at 4:14 AM | Permalink

    193(kim): “Can we re-visit your graph in comment #454 of the Svalgaard #2 thread? I understand it is speculative.”

    Why not. Let’s bring it within the frame of the Armagh data – which appears to be the current topic. Just a recap, the plot I did was for Leif TSI series, and taking into account the Length of Cycle as well as Amplitude. It was also projected forwards to Cycle 24 based on the same pattern of Cycle 14 at Leif’s request.

    So below, we have this plotted in Blue. Very very similar profile to that of the Cycle Length graph shown in the Armagh paper from Butler. Armagh Annual data in yellow. 6 month smoothing in Black. (ignore flat plots after 2008 and before 1790).

    Taking into account the major Volcanic eruption around 1880 which sharply dropped the temperature, I have to say the general trend from 1797 to around 1950 is fairly well matched.

    After 1950? Well, Solar appeared to Peak around 1965. Temperatures had already started falling around 1950. Was this the Sulphate/Aerosol effect proposed by scientists to explain some of the 60′s and 70′s cooling? Who knows. So Solar rises again towards a Peak around 1990. Temperature rise around 1 Cycle behind. Warming restarted 1980.

    Now, with a slowly decling Solar since around 1990, we might expect the temperature to respond around 2001? Well this was the point it started to go flat certainly. We’re almost 2 Cycles behind the Peak now, and I’d want to see an imminent drastic reduction in the temperature if the general ‘trend’ is to hold true.

    Now, other papers from Armagh already state they see a Cyclic pattern in their temperatures on a 7-8, 20-23, 30-33 Year period. It’s here in this paper: Trends and Cycles from 2007. They make a link with their series and the North Atlantic Oscillation (NAO), and also mention the Hale Cycle. So, is this the Buffer mechanism between Solar>Ocean>Armagh Temperatures? May explain the 1 Cycle lag that is proposed.

    Here’s a link to the data I used to fill in the Armagh data to 2004: Armagh Series

    Here’s a link to other publications from Armagh Observatory: Archive

    Finally, here’s the graph:

  208. Posted Mar 26, 2008 at 4:49 AM | Permalink

    206 (David A): so there should be several of 14 years of length, yet your graph shows none. Please submit the list that I suggested in #204.

  209. pochas
    Posted Mar 26, 2008 at 2:30 PM | Permalink

    171 (Leif):

    Eventually a stronger cycle will have to succeed a weaker one. Does your method predict when this will happen?

  210. See - owe to Rich
    Posted Mar 26, 2008 at 3:32 PM | Permalink

    202 (Leif): I answered your interesting comment over on the CA Forum.

    206 (David): Sure I can get Hanover from GISS, though I generally try to avoid touching GISS with a bargepole. I just thought we could help each other out here with digested data – I give you Armagh in a nice ‘R’ format, you give me Hanover. But it’s no big deal. There was an Englishman, an Australian, and a Danish-American in a bar ometer one day, …complete the joke…

    Rich.

  211. Posted Mar 26, 2008 at 5:25 PM | Permalink

    209 (pochas): the polar field method looks at the measured polar fields at the end of a cycle [more precisely the last 3 years of the cycle]. If the polar fields there are strong, the next cycle will be large, and if the polar fields are weak [as they now are], the next cycle will be small. No built-in statistical voodoo.

    I don’t think we at the present can predict more than one cycle ahead with confidence; and not even ONE cycle if we are too early in the cycle. Let’s say that it is now 2010 and you asked if I could predict cycle 25. My answer would be no. By 2020, perhaps, we would be able to predict cycle 25.

  212. David Archibald
    Posted Mar 26, 2008 at 8:10 PM | Permalink

    Re 208, Dr Svalgaard, to paraphrase Chairman Mao, take Japanese cedars as the key. When you do that, the following solar cycles are generated:

    Year of Cycle
    Minimum Length
    1645
    1659 14
    1671 12
    1684 13
    1699 15
    1712 13

    There was one cycle of 14 years, but it ended before the beginning of the CET record.

    Re 210, Rich, your aversion to accessing data from that dark place, GISS, is understandable. If you email me at david.archibald @ westnet.com.au I will send you an Excel file entitled “A File for Rich” which contains the Hanover, Providence and Portland data. This is what I get:

    Station Period Same Cycle One Cycle Lag
    Rsq. Correl. Rsq. Correl.

    Providence, Rhode Island1835 – 2005 0.229 -0.478 0.38 -0.619
    Hanover, New Hampshire 1835 – 2005 0.033 -0.181 0.535 -0.731
    Portland, Maine 1837 – 2005 0.145 -0.381 0.486 -0.697

    Continuing on the theme of sharing joy, a professor from Helsinki sent me an email a couple of days ago which said in part,” I must say your argumentation is among the most convincing and beautifully distilled down to bare facts I have come upon in the present mess of IPCC-dominated “science” (to which most of our local politicians vigorously subscribe).” So I sent him the Helsinki 1829 – 2007 data set.

  213. Posted Mar 26, 2008 at 9:38 PM | Permalink

    212 (David A): well, here is the original (real data) from the Cedars:

    Masuda-san’s data are the black dots. The open dots are from Minze and Stuiver. It is very difficult to find the cycles. They had to filter the data with a bandpass of 10-18 years to get something out of the data. You can compare with McCracken & Beer [2007] incontrovertible 10Be data for the same time period:

    From your wording ["ended"] I infer that the minimum year you give is the ending minima.
    Using that, the 10Be data gives:
    1636
    1648 12
    1661 13
    1674 13
    1684? 10
    1697 13
    1708 11
    and then for several cycles, the minimum are lost in the noise. Note that the 14C and 10Be fluxes are inverted relative to solar activity, so minima appear as maxima in the curces.

    The main point is that for these early cycles the lengths are very ill defined and hardly useful to base any conclusions on. Also, explain why on the CET plot you use the following cycle while on the other plots you use the current cycle. It is probably because you have discovered that it “works” better with one cycle lag, although your paper didn’t say so. In which case the size of cycle 24 has no bearing on the coming ‘cooling’, if any.

    And to quote Mao: “To regard everything as positive is to see only the good and not the bad, and to tolerate only praise and no criticism”.

  214. Andrew
    Posted Mar 26, 2008 at 9:48 PM | Permalink

    213 (Leif) I’m surprised no one has mentioned this yet, but Basil Copeland and Anthony Watts have found a 22 year cycle in dT, which apparently corresponds to the Hale cycle:

    This isn’t another cycle that you include as “cyclomania” is it?

    The number is arbitrary becuase I didn’t want to break the system.

  215. Posted Mar 26, 2008 at 10:12 PM | Permalink

    214 (Andrew): no number is needed if you start a new sub-topic. In fact, it is the absence of a number that signals a new sub-topic.
    In the 1970s there was much talk about a 22-year cycle in drought in the Western U.S., but interest somehow waned.
    To my knowledge there is no solar parameter that shows a 22-year cycle. The Hale-cycle doesn’t count as that is just a polarity shift, the energy output [that goes with the square of the magnetic field] does not have a 22-year period. There is a true [weak] 22-year cycle in geomagnetic activity caused by the fact that for 11 years [from one max to the next] the suns magnetic dipole is aligned with Earths while for the next 11 years [again from max to max] they are anti-aligned. You can find the reason for this on Page 51 [53 of the PDF] of my [old] paper here.

  216. Raven
    Posted Mar 26, 2008 at 10:34 PM | Permalink

    215 (Leif) – Could the polarity shift affect the way GCRs that hit earth? For example, if incoming GCR flux varies spatially then a change in magnetic field could direct a different group of GCRs towards earth. This could result in a change in GCR flux over a short time scale.

  217. Posted Mar 26, 2008 at 10:51 PM | Permalink

    216 (Raven): The GCR modulation is slightly different between solar cycles. Go back to #11. Note, how every other maximum is sharply peaked while the ones in between have a broad peak. The difference is caused by slightly different drifts of the GCRs in the solar system depending on the polarity of the poles [so again from max to max], but the total flux is not affected much, just the shape of the solar cycle modulation. This effect [which is on a long time scale - years] is among the few that are well understood.

    Some Russian researcher claims that the agitation of inmates of lunatic asylums increases when the polarity changes as the heliospheric current sheet sweeps by the Earth every week or two, so who am I to say that the weather/climate might not be similarly affected? :-)

  218. David Archibald
    Posted Mar 27, 2008 at 3:27 AM | Permalink

    Re 213, Dr Svalgaard, there’s the rub. Despite of the importance of the Maunder Minimum to our understanding of climate, seemingly there is no reference text on the solar cycles of that period. Perhaps I could write the definitive paper. The Japanese researchers in their paper say: “However, the frequency analysis of the carbon-14 record has shown significantly the period of about 13-15 years together with the period of about 23-29 years, which suggest the persistent cyclic magnetic reversals of the Sun through the Maunder Minimum with the period several years longer than that of recent solar activity.”

    Re 214, Andrew, don’t worry, there are a multitude of papers showing a Hale cycle effect on climate. This is just one of them: Hale Cyclicity of Solar Activity and Its Relation to Climate Variability

    http://www.springerlink.com/content/j7l094tn80hvv178/

    Relevant passages include “The wavelet analysis of annual series of conifer tree rings for the period 1458–1975 has revealed climatic oscillations with periods of 20–25 years. Especially strong climatic variations in comparison with the recent ones were found to occur during the Maunder minimum epoch when the period of oscillations increased from 22–23 years to 24–29 years.”

  219. Pete
    Posted Mar 27, 2008 at 5:23 AM | Permalink

    It should also be noted that Anthony Watts (and Basil Copeland) have just extracted a 22 Year Cycle in their recent article “Solar Fingerprint Found”

    Just to add to my posting above on Armagh v Solar Length/Amplitude. This suprised me a little. I did two sorts in Excel. The Top one is Long Solar Length > Short. The Bottom one is Low Temperature to High.

    Interesting hey. I never expected the trends would be so obvious. Not perfect, but something mathematical is trying to jump out. Worth more serious study. (I am but an amateur!)

  220. Pete
    Posted Mar 27, 2008 at 5:33 AM | Permalink

    219(Pete): Myself.

    Ignore the Year scale on those graphs, as this is meaningless after the Sorting. Also, you obviously would not expect a perfect Solar match with temps, as they are also being driven by such factors as ENSO, Volcanoes, Ocean Current, and Local Synoptics at this site.

    Also, some of those point which appear to deviate massively are generally just One Year out of this series which is from 1798 to 2004. So 206 Years and 10-20 points or so are out, less than 10%.

  221. Posted Mar 27, 2008 at 10:36 AM | Permalink

    220 (Pete): for the folks explain how you calculate your solar index.

  222. Pete
    Posted Mar 27, 2008 at 11:16 AM | Permalink

    221(Leif): “for the folks explain how you calculate your solar index”

    Leif, basically the Solar Index is Cycle Length, but I also take into account amplitude, so it is an Index that is based on the relationship between the Length and the Max Amplitude. The idea being it is possibly a better indicator of the ‘Energy’ being delivered per Cycle Pulse.

    Now, I have really spooked myself out this afternoon, and I need to go and have a reality check. Some of you may have noticed in the Armagh graph above that at Long Cycles the temperature trended very well, but towards the end the relationship broke down somewhat as the Cycles shortened.

    Shall we move onto Hadley now? This has made me sit back somewhat as I don’t know what it means what-so-ever. The Top graph below is Hadley v’s Solar Index. It is less impressive than Armagh, but there are times it is good, such as around 1880 to 1950.

    Bottom graph has startled me out of my Chair to be honest. Again, I sorted Long Cycles to Short Cycles. Again a fairly good trend for the first 1/3. Again a good match of Global Temperature trend at Long Cycles only.

    Have you seen the last 2/3 of that graph! As Solar Length reduces the Amplitude of the temperature swings decay away, and not only that the temperature has a frequency of around 22 Years, except for the 2nd larger Cycle which appears reversed – so it is 44 Years, but would be 22 if that -ve hump was upside down. Weird.

    How did this 22 Year ‘beat’ in temperature get extracted when I sorted my data by Cycle Length? Is there a systematic error in my Spreadsheet somewhere? If there is I ain’t found it yet. The Yellow Hadley plot is from pure Hadley Temperature Data – it has only been % scaled to fit on the graph with the Solar Index.

  223. Posted Mar 27, 2008 at 12:16 PM | Permalink

    222 (Pete): your explanation is not good enough. You can give a better one. A formula, algorithm, etc. Enough that somebody else can calculate the number.

  224. Pete
    Posted Mar 27, 2008 at 12:43 PM | Permalink

    223(Leif): “your explanation is not good enough. You can give a better one. A formula, algorithm, etc. Enough that somebody else can calculate the number.”

    Didn’t think I gave an explanation, hoping someone else would do that! May be I was talking to myself out loud.

    Actually, I realise now where the 22 Year freq. comes from, it is now obvious that this is very near the Length of the average Cycle that I am sorting against! So I can account for that. The clustering of temperatures to produce that Wave when I put it though the Filter must be a coincidence. Looks nice, but as the Cycle Lengths are out of order chronologically, it must be a coincidence that it looks like a pattern.

    There you go, I debunked my own results – after carefully re-examining my method!

  225. Pete
    Posted Mar 27, 2008 at 1:00 PM | Permalink

    223(Leif): ““your explanation is not good enough. You can give a better one. A formula, algorithm, etc. Enough that somebody else can calculate the number.”

    Ahhh! You mean the Index I used for Solar Length/Amplitude. I thought I explained that back in Svalgaard #2? You explained it back to me quite nicely as the area under the curves down to a TSI floor of 1365.5. Add up the TSI bar from this floor to the height for each year of the Solar Cycle and then Divide by the Length to get an Index which is then the relationship between Amplitude and Length. Quite straighforward really. Didn’t I explain this?

  226. Pete
    Posted Mar 27, 2008 at 1:07 PM | Permalink

    This is what you get by following the above method. I rounded the Cycle Lengths to whole years, you might get more accurate results by using to one decimal place. I meant to update for this but never did.

    TSI YEAR &
    (AREA) CYCLE NUMBER

    3.74 1760-1
    4.97 1769-2
    4.63 1779-3
    6.66 1790-4
    1.94 1803-5
    1.53 1816-6
    3.26 1828-7
    5.43 1838-8
    5.72 1849-9
    4.61 1861-10
    5.06 1872-11
    3.18 1883-12
    3.51 1894-13
    2.7 1906-14
    3.28 1916-15
    3.28 1927-16
    4.55 1938-17
    4.75 1947-18
    5.95 1958-19
    4.34 1969-20
    5.44 1980-21
    4.99 1990-22
    4.59 2001-23
    2.69 2013-24

  227. Andrew
    Posted Mar 27, 2008 at 1:10 PM | Permalink

    Okay, I’ve been revisting some old stuff, and I was hoping someone could help me understand this graphic:

    Attributed to Doug Hoyt, which shows “the ratio of the umbra area to that of the whole spot (U/W) derived from Greenwich Observatory data”

    Comments, anyone?

  228. Posted Mar 27, 2008 at 1:34 PM | Permalink

    225 (Pete): yes, but it was not clear that you used that one [had you just said: "as explained in #2"]…

  229. Posted Mar 27, 2008 at 1:38 PM | Permalink

    225 (Pete): yes, but it was not clear that you used that one [had you just said: "as explained in #2"]…

    227 (Andrew): the explanation is quite clear. Trouble is: nobody knows what it means, that is: what is the physics or conditions determining this.

  230. Andrew
    Posted Mar 27, 2008 at 3:06 PM | Permalink

    229 (Leif) Thanks, I think my curiosity will be satisfied with that. Another question: some one has just pointed out to me the proximate start and end dates of the “big” minimums:

    Oort (1010-1050)
    Wolf (1280-1340)
    Spörer (1415-1534)
    Maunder (1645-1715)
    Dalton (1790-1840)

    A simple linear fit and extrapolation of the start dates reveals that the next “big” minimum should have started around 2005. This seems pretty regular to me. What do you make of it?

  231. Posted Mar 27, 2008 at 7:12 PM | Permalink

    230-233 (many): I’ll comment on what I found at the link:

    the warming of the late 20th century appears to be no more “unusual” than the warming during the 1920′s and 1930′s. Both appear to have their origin in a solar cycle phenomenon in which the sinusoidal pattern…

    this is stated without any justification than appearance and is not science.

    Note the Judith Lean trace (brown) is significantly out of family with the others. I don’t recall where I read she had retracted the data. I’ll see if I can find it

    the Lean curve is old. Scientists never retract anything unless it was based on flawed data or methods and “they were found out”. If a paper was based on the best data available at the time and the analysis method is otherwise sound, there is nothing to retract. Later data and better methods often lead to new papers with possibly different results. That is called progress.

    I have high regard for Dr. Svalgaard’s encyclopaedic grasp of all details solar, but I cannot condone regarding his opinions and theories as mainstream. Try his old collaborators, Schatten and Hoyt if orthodox belief is your central concern.

    I’m not sure ‘condone’ is the right word here. The mainstream often lags a bit behind the leading edge. And H&S are certainly not mainstream solar physics anymore, although they may still be ‘mainstream’ in climate ‘research’ because they provide a convenient increase where there otherwise would be none.

    Now that the global warming community has bullied Judith Lean into reconstructing her solar reconstructions at least three times now and the latest numbers show hardly any solar variation whatsoever,

    I hate to use the word, but this is just pure nonsense. Judith is a VERY good scientist and represents some of the best work in this field. But, again, new data and ideas emerge all the time and reconstructions have to be updated as needed, and she is doing a fine job in the right direction :-)

    Dr. Gerhard Loebert (10:52:17) :
    A Compilation of the Arguments that Irrefutably Prove that Climate Change is driven by Solar Activity and not by CO2 Emission [...] While the rotational velocity of the Earth and the thermonuclear energy conversion process on the Sun react simultaneously to the passage of a vacuum density wave, a time span of 6 years is needed for the energy to be transported from the core of the Sun to the Earth’s atmosphere and for the latter’s reaction time.

    This fellow is a stark contrast to Lean. He is not even wrong. This is gibberish. Just to comment on the last assertion: it takes between 17,000 and 50,000,000 [my own favorite number is 200,000] years for solar energy [in terms of photons - neutrinos make it in couple of seconds] to travel from the core to the surface.

  232. Raven
    Posted Mar 27, 2008 at 10:22 PM | Permalink

    jae (236): Do a Google search on ‘Dr. Gerhard Loebert’ you will find nothing but spam posted to other people’s blogs. That does not mean he is wrong – but it is not a good sign.

  233. Jeff C.
    Posted Mar 27, 2008 at 11:01 PM | Permalink

    Leif (231): “the Lean curve is old. Scientists never retract anything unless it was based on flawed data or methods and “they were found out”. If a paper was based on the best data available at the time and the analysis method is otherwise sound, there is nothing to retract. Later data and better methods often lead to new papers with possibly different results. That is called progress.”

    Thanks for making the distinction. I made the comment over on Anthony’s blog and it was poor choice or words. The point I was trying to make was that later TSI work in which she contributed more closely matched that of other solar scientists. I certainly didn’t mean to denigrate her reputation and must learn to be more careful with my words.

    As a long time lurker on this thread I have really enjoyed the discussion and learned a great deal. Thanks for all the time you have devoted to this effort.

  234. Stargazer
    Posted Mar 28, 2008 at 2:21 AM | Permalink

    230 (Andrew):

    someone on solarcycle24.com proposed that a ‘new. minimum should be called the ‘Gore mimimum’

    I like irony.

    Oort (1010-1050)
    Wolf (1280-1340)
    Spörer (1415-1534)
    Maunder (1645-1715)
    Dalton (1790-1840)
    Gore (2005-

    what do you all think !

  235. JohnB
    Posted Mar 28, 2008 at 2:59 AM | Permalink

    #234 (Stargazer).

    I understand the irony but I’d be reluctant to immortalise the name – I’d prefer to stick to a naming convention which recognises those who may, through their work in science, have shed light on the event. On this basis I would see the coming minimum named either the Landscheit Minimum or the Fairbridge Minimum.

  236. Posted Mar 28, 2008 at 6:33 AM | Permalink

    235 (John B): I agree that Gore is not deserving, but neither are the other two you mention. Their ‘predictions’ are based on tidal influences, and incidentally with different periods, 166 yrs and 178 yrs, versus the 194 yr period the the minima given in #234. I have elsewhere suggested the ‘Eddy’ minimum to honor the man who brought the idea of grand minima into the mainstream solar science. Maunder was widely ignored until ~1975.

  237. Posted Mar 28, 2008 at 8:32 AM | Permalink

    Folks, it seems that some comments 230-233 were deleted. I don’t know how and why. Steve M? at any rate they were not too important. Just screws up the numbering scheme. This should teach me not to fuss about the numbering…

    Steve: I deleted a few comments late last night promoting a fringe author. Editorially, I didn’t think that a discussion of this particular author was something that I wanted to devote space at the blog to. I didn’t get the impression that you wanted to bother either.

  238. JohnB
    Posted Mar 28, 2008 at 10:04 AM | Permalink

    #236 Leif. Thanks for your comment and incidentally many many thanks for your continued contribution on this site which is hugely appreciated. I nominated Landscheit and Fairbridge because both researched the impact of barycentric motion and both forecast as a result that we would now be entering a period of cooling. I won’t expand any further because Steve doesn’t allow this particular area of science and as the site owner I respect his view. The following paper by Chervatova in some ways summarises their work:

    http://www.ann-geophys.net/18/399/2000/angeo-18-399-2000.pdf

  239. Posted Mar 28, 2008 at 10:53 AM | Permalink

    238 (JohnB): being right for the wrong reason doesn’t count. And Steve is correct to exclude the stuff that is not science.

  240. JohnB
    Posted Mar 28, 2008 at 11:13 AM | Permalink

    #239 Leif. I respect your opinion. Such absolute certainty can bring great comfort. My lack of knowledge means I am condemned to view their work with an open mind.

  241. Richard Sharpe
    Posted Mar 28, 2008 at 11:24 AM | Permalink

    JohnB, I tend to side with Leif. Anyone can observe correlations and so forth. Explaining the underlying mechanism is much more important.

    Furthermore, it is not a sign of an open mind to believe that something was achieved through magic when in reality it was high tech (some will understand the reference, no doubt). Also, beware of cargo-cult science.

  242. Posted Mar 28, 2008 at 11:51 AM | Permalink

    240 (JohnB): there is no such thing in science as absolute certainty. Theories that are accepted as part of the current paradigm form and interlocking whole. There are always [a very small number of] observations or ideas that do not fit within the paradigm. These are put on the back burner and may become important later [they mostly don't]. The problem with all the tidal/barycentric ideas is that the energy is not available to do anything. There are other star systems with giant planets very close to the star and on those there are such effects on stellar activity, but the mechanism just doesn’t work on the sun. Keeping an open mind is not enough, you also have to apply some sense.

  243. JohnB
    Posted Mar 28, 2008 at 12:04 PM | Permalink

    #241 Richard. My #240 was a sincere comment. I am uncomfortable when I hear the work of Fairbridge or Landscheit dismissed because there is no detailed explanation of the mechanism by which their observations might be realised. That same view rejected the work of Mendel when his work was first published, it wasn’t until long after his death that his contribution was recognised. He observed what was happening. He had no explanation of the detailed mechanism by which it happened.

    If you read their work, and Chervatova’s, it is hard to dismiss out of hand. In many ways I hope they are wrong. The world is ill prepared for another Maunder mininmum.

    I view their work with an open mind.

  244. Posted Mar 28, 2008 at 2:12 PM | Permalink

    243 (JohnB): There are many signs of low solar activity coming. If that happens, it is thus not a vindication of L or F or A. So when you have an open mind is that about low activity or about the viability of their particular ‘mechanisms’?

  245. Raven
    Posted Mar 28, 2008 at 5:47 PM | Permalink

    244 (Leif): Given the shear number of predictions I would be surprised if random chance does not deliver at least one ‘prophet’ who was bang on. That said, Landscheit seems to have a track record that is hard to dismiss as dumb luck. It is possible that he is on to something even if he did not understand what it is. OTOH, someone had a model that predicted outcomes of presidential elections based on the outcome of a football game – a model that was 100% accurate over 20+ elections until 2004.

  246. John Lang
    Posted Mar 28, 2008 at 6:51 PM | Permalink

    There have been a lot of great posts on these threads.

    It seems to me the only solid conclusions that came from them are that the Sun is a variable star but it does not vary that much at all. However, the total possible extent of those variations and the potential effects on Earth’s climate are still unknown but still expected to be small (compared to the ice age cycles for example). It is also difficult to predict those variations but we are making substantial progress.

    Given the ice age cycles the Earth has experienced have clearly much more impact on the climate than solar effects, perhaps the Milankovitch cycles should be looked at more closely instead.

  247. Andrew
    Posted Mar 28, 2008 at 8:06 PM | Permalink

    246 (John Lang) I’m still not convinced that little variation in TSI necessarily discounts related effects as drivers. At any rate, your statement assumes that any change must be due to radiative forcing and discounts the possibility of internal variability (Lindzen considers this the main competing explanation for GW) for instance the idea of shifts in oscilations of the oceans (PDO, etc.):

    http://www.worldclimatereport.com/index.php/2007/08/17/climate-change-chaos/

    Anyway, it seems to me that some warming is likely due to soot, which I suspect more than offsets the cooling tendency of other pollutants.

  248. JohnB
    Posted Mar 29, 2008 at 3:22 AM | Permalink

    #244 Leif. Do I have an open mind about low solar activity or the viability of Fairbridge and Landscheit’s work? Both I guess Leif.

    What I read them to say is that the solar system has mass. It can be calculated. That mass is held together around the largest single concentration of the solar system’s mass, which is the sun. Because the mass is not evenly distributed, the centre of gravity for the solar system is not at the centre of the sun. Because the distribution of the mass of the solar system changes with time, the centre of gravity moves with time.

    This seems a reasonable position to take.

    They then observe that the change in distribution of the solar system’s mass is cyclical, and the timing of those cycles correlates with “recent” periods of cooling on earth.

    Again, seems reasonable to me.

    Chervatova explicitly plots the change in distribution of solar system mass and illustrates how that change in distribution correlates with the Wolf, Sporer, Maunder, Dalton, and potentially the cooling period we are about to enter.

    http://www.ann-geophys.net/18/399/2000/angeo-18-399-2000.pdf

    Correlation isn’t causation, accepted.

    For me it’s interesting that David Hathaway made his comment in May 2006 about the Solar “conveyor belt” having slowed to a record low. “It’s off the bottom of the charts… this has important repercussions for future solar activity.” He used this to forecast a very low Cycle 25. Could this be connected to the Cof G movement? Is this part of the “mechanics” behind Fairbridge and Landscheit’s observations?

    Wouldn’t it be reasonable to infer that possibly the reciprocal effect of gravity from those parts of the solar systems mass which are outside of the Sun could have an effect on the solar conveyor? The response to this is usually that the individual units of solar system mass outside the sun are too small to have such an effect. Perhaps that’s true, but during the Wolf, Sporer, Maunder and Dalton minima the mass of the solar system concentrated more to one side of the sun. It’s in a similar position now.

    As an aside I came across this NASA paper by Ching-cheh Hung which mentions the effect of non-solar mass on solar tides and solar activity albeit with reference to much smaller cycles.

    http://ntrs.nasa.gov/search.jsp?R=141113&id=6&qs=No%3D10%26N%3D4294888830

    #241 Richard. I’m not sure I understand your second paragraph. Magic?

  249. Philip Mulholland
    Posted Mar 29, 2008 at 4:00 AM | Permalink

    JohnB #248

    From Arthur C. Clarke’s three laws:-

    ¶“Any sufficiently advanced technology is indistinguishable from magic.”

    The New York Times

  250. Posted Mar 29, 2008 at 6:29 AM | Permalink

    245,250 (many): we have been down that road before and there is no need to rehash all the old tired arguments. ‘Predicting’ low solar activity from past behavior can be done simply by noting that activity was low for a few cycles following the years 1700, 1800, 1900, and now 2000:

    My 9-year old grandson remarked that when he saw the above graph from the sunspot center in Brussels. Prediction without viable mechanism doesn’t count. And I, for one, do not buy into the notion that “the beautiful thing is that we don’t need to know how it works, just look at the graphs”.

  251. Steve McIntyre
    Posted Mar 29, 2008 at 7:24 AM | Permalink

    #248. JohnB, no more Landscheit or barycentric planets please. Discuss them elsewhere please.

  252. John Finn
    Posted Mar 29, 2008 at 7:55 AM | Permalink

    Just a few comments on the solar cycle length (SCL)/temperature link. I suspect that David Archibald’s predictions of drastic imminent cooling are way off the mark – even at his preferred set of locations (i.e. Hanover, Armagh, De Bilt etc). In fact, if I were an advocate of AGW, I would promote David’s findings as the mainstream sceptic viewpoint. Then, when the predicted cooling fails to materialise – Bingo! the Sun’s activity is lower – temperatures should be much lower – the fact that they’re not is all down to CO2.

    In his papers, David refers to a plot by Butler and Johnson which shows the relationship between SCL and the ~200 year temperature record at Armagh from which he concludes temperatures will decline by 0.5 degrees for each year of SC length. I’m not sure if this is relative to the average SCL of ~10.7 years or the length of the previous cycle. Either way, with the current cycle likely to be 12+ years, this still represents a significant cooling.

    Now bear with me here, my current internet access is limited to a couple of hours a week at the local library, so much of what I write here is from memory, but I believe it to be broadly correct.

    Butler and Johnson plot the 11 year mean temperature centred on the years of solar maximum AND solar minimum. My interpretation of this is that the temperature for the 1996 solar min (SCL ~9.6 yr) is the mean temperature for the period 1991-2001. Similarly the temperature for the 2000 solar max is the mean temperature for the 1995-2005 period. This is the last point (max or min) for which we have full data. However, we do have some years of data which will contribute to the next data point, i.e. the SC23/SC24 solar minimum which according to David (and others) is due in 2009. The B&J temperature for 2009 should be the mean temperature for the period 2004-2014, so we already have 4 full years of data (i.e. 2004-2007).

    Unfortunately, I can’t find any current data for Armagh , but the UK Met Office do hold up-to-date records for Northern Ireland as a whole. For those who may not be aware, Armagh is the county town of Co. Armagh which is situated in Northern Ireland. Now before David Archibald or anyone else objects to the use of the NI record on the grounds of UH contamination or whatever, I’ve checked the 2 records and they are extremely similar – though, since 1975 (to 2002) , Armagh shows the STRONGER warming trend. In fact, I can’t find anywhere in NI that has a higher rate of recent warming than Armagh.

    So – what does the NI record (begins 1914) tell us about the past 4 years. Well – 2007 appears to be the warmest year on record, 2006 the 2nd warmest and 2005 the 3rd warmest. And just for good measure 2004 is the 7th warmest. Remember – if my interpretation is correct – the temperature for 2009 (mean 2004-2014) is supposed to be at least one degree lower ) than the temperature for 1996 (mean 1991-2001). I’ve just seen something written by David which suggests 1.6 deg lower, but let’s be generous. I’m perfectly willing to accept a one degree fall as confirmation of the theory.

    The mean NI temperature for 1991-2001 is 8.96, so we’re looking for a mean temperature (2004-2014) of ~7.96. The current mean temperature for 2004-2007 is 9.61 which means that the next 7 years (including 2008) would need to have an average temperature of 7.02 deg for a one degree fall to be achieved. If I tell you that the lowest temperature in any one year is 7.66 (in 1979) you can judge for yourselves the likelihood of this happening. It’s going to take a significantly colder few years just for 2009 to be as cool as 1996. For all the hype about the recent cooling (which isn’t going to continue) temperatures are only low relative to the recent warm period.

    In a nutshell: anyone who places too much reliance on David’s findings is likely heading for disappointment.

    But – (there is a but) if there is a relationship between SCL and temperature – it’s always struck me that simply plotting raw temperature against SCL assumes that temperature is only dependant on the one relevant solar cycle, whereas intuitively, at least, I would expect previous cycles to have some (less-weighted) influence. For example, following the completion of a sequence of 3 solar cycles of lengths 13,13,13 yr respectively, I would expect temperatures to be lower than at the end of a 10,10,13 yr sequence. If SC lengths do occur in clusters then this could possibly account for the steep gradient David has found with some of his regression slopes, i.e. many of the data points are the result of a cumulative effect (both warm and cold).

    All this is probably trivial to many readers on this site who deal with all sorts of complex filtering and suchlike, but I was just wondering (as I’ve never seen it done) if there’s any value in plotting the temperature change over each cycle (rather than the actual temperature) v the SCL. It just seems to me that this provide more of a direct measure of the effect of each cycle. It might also show that David’s 1.6 degree fall may eventually come to pass – but not for another 30 years or so perhaps.

  253. John Finn
    Posted Mar 29, 2008 at 7:56 AM | Permalink

    213 (Leif) :

    I was interested by the McCracken & Beer plot. First impressions suggest a very strong (inverse) relationship between the 10Be data and the CET data over the same period. In particular, CE temperatures hit a deep trough in the 1690s before rapidly rising around 2 degrees by the mid 1730s.

    Well – it interested me, at least.

  254. Posted Mar 29, 2008 at 9:51 AM | Permalink

    253 (JohnF): The 10Be-record is not well understood. 10Be is created by cosmic rays at low and mid-latitudes, then transported by atmospheric circulation to the polar regions and removed from the atmosphere by rain and snow, so 10Be is not independent of climate nor of volcanic activity [aerosols]. So trying to correlate every wiggle with weather/climate is somewhat dubious. The gross features are likely correct, namely that cosmic ray activity was higher during the LIA because solar activity was reduced.

  255. Posted Mar 29, 2008 at 11:27 AM | Permalink

    Re# 252. I think JohnF’s point : “…If SC lengths do appear in cluster…” may have touched on a perinent area. Perhaps the inter-play of climatic feedbacks tend to temperature stablity centred on a median global temperature only disturbed by a prolonged period when one of the factors effecting climate is changed. Hence there may be some merit in arguing that a single solar cycle of 13 years will not be sufficient to disturb the global temperature stability but a sequence of three 13 year cycles may be sufficinet to do so.

    Just thinking out loud but I think John has made an interesting point.

  256. Posted Mar 29, 2008 at 12:17 PM | Permalink

    255 (Martin): The emphasis on length is misplace, IMHO, statistically, long cycles are weaker, so three long cycles in a row simply means that solar activity [spots, TSI, CR modulation, whatever] was low during some forty years and that may be what is important, if anything.

  257. JohnB
    Posted Mar 29, 2008 at 12:39 PM | Permalink

    #251 Steve. Understood. Apologies.

    John

  258. Posted Mar 29, 2008 at 12:50 PM | Permalink

    Re#256 Thanks again Lief for your quick response and for pointing out what is really the obvious. Just another thought though: in a media world dominated by short-termism, how can you persuade the wider debate to start thinking in terms of forty year time spans? It’s a problem I don’t know the answer to.

  259. Posted Mar 29, 2008 at 1:28 PM | Permalink

    258 (Martin): 40-year spans; I once worked for the phone company. They have a requirement that all new systems must be compatible with existing equipment 40 years back. The Wilcox Solar Observatory is now in its fourth decade. Mount Wilson Observatory have observed the sun “on every clear day since 1912″. Several geomagnetic observatories have observed the Earth’s magnetic field for more than a century. Weather records extends even further back. Things are not so bleak. Bad policies expire with the bad people that made them. Each new generation can generally pick up the pieces and make their world. Call me an optimist on this.

  260. David Archibald
    Posted Mar 29, 2008 at 8:21 PM | Permalink

    Re 250, Dr Svalgaard, you say you do not buy into not knowing how it works, therefore you accept that the solar cycle length/temperature relationship is valid? Are you going to put more effort into figuring out how that works than figuring out why the convective zone rotates faster than the radiative zone, the mechanism for which you are quite happy to remain blissfully ignorant of? In a world in which solar physicists do not know why the convective zone rotates faster than the radiative zone, why should the laity be denied the correlations between solar activity and climate that they can derive?

    Re 252, John Finn, I find the length of your analysis of my paper quite encouraging. It is like the RealClimate writeup of my previous paper on 4th October, 2007. It means that I am pushing the right buttons and getting traction. Let’s recap on theory. Solar Cycle 22 was 9.6 years long, which is short, and therefore the following decade should be hotter than normal, and it was, at least up to 2005 odd. This sentence of yours: “For all the hype about the recent cooling (which isn’t going to continue)” suggests that your analysis contains a lot of hoping and wishing and dreaming. Anyway, thanks again for the encouragement.

  261. Posted Mar 29, 2008 at 8:37 PM | Permalink

    With due respect: What about cycle 23 strange behaviour?. It seems getting a new peak right now with spots 987,988 and 989. How long will it last?

  262. Posted Mar 29, 2008 at 9:28 PM | Permalink

    260 (DavidA): Your first sentence: talking about double negative… Regardless, your statement does you a well-deserved disservice.

    About solar rotation: this is not such a mystery.
    See “Differential rotation and meridional flow in the solar convection zone and beneath”, by Kitchatinov, L. L.; Rüdiger, G.: Astronomische Nachrichten, Vol.326, Issue 6, p.379-385, 2005, DOI:10.1002/ansa.200510368

    Abstract
    The influence of the basic rotation on anisotropic and inhomogeneous turbulence is discussed in the context of differential rotation theory. An improved representation for the original turbulence leads to a Λ-effect which complies with the results of 3D numerical simulations. The resulting rotation law and meridional flow agree well with both the surface observations (partialΩ/partial r
    with helioseismic measurements:

    looks pretty good to me, and we know how it works.

    261 (Adolfo): cycle 23 is not unusual. “Last gasp” spots near minimum are quite normal. These spots and their magnetic fields [as organized features] will be gone in a few weeks.

  263. Posted Mar 29, 2008 at 9:31 PM | Permalink

    262 (me): and here is the theoretical calculation:

    looks pretty good to me, and we know how it works.

  264. Posted Mar 30, 2008 at 12:36 PM | Permalink

    In #11 I showed the run of cosmic ray flux for the past several cycles (fun to compare the last cycle with Belov’s prediction in #15 – prediction is hard). I noted the alternating peaked and rounded maxima [which are understood in terms of drifts and polar fields]. Every other flux maximum [at solar minimum] can thus be compared. Below I show this for the minima around 1965 [blue], 1986 [pink], and now 2008 [yellow]. The smooth curves are simple parabolas fitted to the fluxes. The curves have been shifted [to within 0.5 year] to ‘line up’ at minimum:

    Several things to notice:
    1) at every minimum [so far] the flux is close to the same within a couple of percent.
    2) it seems hard to push the minimum for 2008 out too much, say towards the end of 2009.
    3) it seems unlikely that the flux will increase substantially, say to above 10,500 [off the graph], and especially not to 12,000 as some people have predicted. [note that by coincidence the count is almost the percentage times a hundred. CR fluxes are often quoted in percent of the maximum values - at solar minimum]. This may have some implications for people claiming that cosmic rays influence climate.

  265. VG
    Posted Mar 30, 2008 at 8:53 PM | Permalink

    Leif I was wondering if you would care to comment on this analysis

    http://wattsupwiththat.wordpress.com/2008/03/30/evidence-of-a-significant-solar-imprint-in-annual-globally-averaged-temperature-trends-part-2/

  266. Posted Mar 30, 2008 at 10:04 PM | Permalink

    Evidence of a Significant Solar Imprint in Annual Globally Averaged Temperature Trends

    published here

    11 and 22 year cycles show up nicely in HadCRUT surface temperature data when filtered properly. Comments welcome.

  267. Posted Mar 31, 2008 at 4:18 AM | Permalink

    266 (Anthony): I don’t like the word ‘properly’. Do some sensitivity tests:
    1) try another series, GISS?
    2) add some random noise, signal still there?
    see how much noise you have to add to make the signal go away.
    3) filter the series backwards in time.
    4) make a superposed epoch analysis with solar minimum as key times

  268. Posted Mar 31, 2008 at 6:41 AM | Permalink

    267 (me): and why analyze the ‘rate of change’? If the rate of change is cyclic so is the function: dy/dt = sin(at), then y = c – 1/a cos(at). So you should see the signal in y as well. If you don’t ….

  269. Posted Mar 31, 2008 at 8:29 AM | Permalink

    Thanks Leif, I appreciate the suggestion we’ll see how well it stands up to those sorts of tests.

    I understand your concern over “properly”. When I used the word, I was not suggesting that other methods were “imporoper”. I was only saying that we’d applied the HP filter properly to the task at hand. I could have used a better word or phrase.

    Such are blog comments, sometimes they get banged out in haste.

  270. Posted Mar 31, 2008 at 8:35 AM | Permalink

    269 (Anthony): I was thinking of a much more serious misuse of ‘proper’, namely as the ‘proper’ way of looking for this particular ‘signal’. If you have a series that contains all frequencies [white noise] and band-pass filter it with any frequency band, you will found signal at precisely the frequency you are looking at. No?

  271. kim
    Posted Mar 31, 2008 at 9:01 AM | Permalink

    Basil C. is suggesting a cosmic ray mechanism in the comments to that post.
    =================================================

  272. Pete
    Posted Mar 31, 2008 at 10:01 AM | Permalink

    270(Leif): “If you have a series that contains all frequencies [white noise] and band-pass filter it with any frequency band, you will found signal at precisely the frequency you are looking at. No?”

    Leif, but why would the filter extract the frequency of the Solar Cycle from the White Noise, and not any other frequency? Do you think this is a coincidence or a function of the Filter being used?

  273. Jeff C.
    Posted Mar 31, 2008 at 10:15 AM | Permalink

    Leif (268): “and why analyze the ‘rate of change’? If the rate of change is cyclic so is the function: dy/dt = sin(at), then y = c – 1/a cos(at). So you should see the signal in y as well. If you don’t ….”

    This is just my opinion, but it appears to me that the same phenomena can be seen in the Y axis but it is much less apparent. There may be a good reason for this. Anthony, Ross McKitrick and others have argued that the surface temperature record has an exagerated positive slope over time due non-climactic influences (UHI, equipment changes, poor siting issues, etc.). Unfortunately, since the satellite record only extends back to 1979, it is the best we have. By analyzing the rate of change, one can in essense de-embed the potential non-climactic slope (or a potential climactic influence of other that solar origin). A slow rate of increase vs. a fast rate of increase is very apparent in the plots. If the suggested non-climactic influences were not there, the raw temperature data might show positive and negative swings in the Y-axis. With it there, the warming slows as opposed to actually cooling.

    I’m sure there are other reasons for analyzing dT/dt, but that’s what makes sense to me.

  274. Posted Mar 31, 2008 at 12:36 PM | Permalink

    272 (Pete):

    why would the filter extract the frequency of the Solar Cycle from the White Noise, and not any other frequency?

    Because of this quote from the paper:

    For the layman reader, this equation is much like a tunable bandpass filter used in radio communications, where lambda (λ) is the tuning knob used to determine the what band of frequencies are passed and which are excluded

  275. Posted Mar 31, 2008 at 12:49 PM | Permalink

    273 (Jeff): The standard way of dealing with this is to subtract an 11-year running mean, that will take care of most non-solar cycle related variation. And has the advantage of being understandable to all. Not for you, but for Anthony: in the usual description of the HP-filter it is suggested to use lambda = 100 for yearly data, 1600 for quarterly data, 14400 for monthly data. If I was analysing the data sitting on Mars where a year is 687 Earth days, what value of lambda to use?

  276. Posted Mar 31, 2008 at 12:56 PM | Permalink

    275 (Anthony): what difference does it make what the unit of time is? The series is a set of numbers indexed by the integers 1, 2, 3, …, N. Any cycles [and e.g. how many there are] will be independent of if those integers represent centuries, years, or microseconds, no?

  277. Chris Knight
    Posted Mar 31, 2008 at 12:59 PM | Permalink

    If real, the same signature should be apparent in all temperature records/proxies, not just global syntheses.

  278. Anthony Watts
    Posted Mar 31, 2008 at 1:13 PM | Permalink

    RE270, like a radio, if you tune it to a range of frequencies (band) if there is a signal there you’ll hear it. If there is no signal, noise. Also like in a radio, you can pull weak signals from the noise using a discriminator. The question really is, is the signal strong enough to be discriminated from the noise? I think it is. One of our commenters at my blog used an FFT method to resolve a similar pattern to what we found.

    see also this paper, which purports to recognze a Hale cycle effect in surface data:

    TRENDS AND CYCLES IN LONG IRISH METEOROLOGICAL SERIES
    C.J. Butler, A. Garcý´a-Sua´rez and E. Palle´

    ABSTRACT
    We have analysed the trends in four long meteorological time series from Armagh Observatory and compared them with series available from other Irish sites. We find that although maximum and minimum temperatures have risen in line with global averages, minima have risen faster than maxima, thereby reducing the daily temperature range. The total number of hours of bright sunshine has fallen since 1885 at the four sites studied, which is consistent with both a rise in cloudiness and the fall in the daily temperature range. Over the past century, soil temperatures at both 30cm and 100cm depths have risen twice as fast as air temperature. Wavelet analysis has found significant cycles with periods of 7-
    8 years, 20 – 23 years and 30 – 33 years in the seasonal and annual meteorological series from Armagh. Some of these cycles are clearly linked to the North Atlantic Oscillation.

    “…..Wavelet analysis has shown that some of the same periodicities seen previously in rainfall and temperature are also present in sunshine and soil temperature for Armagh. The 6 – 8 year period, which is clearly seen in at least one season in all four meteorological series, is most likely associated with a similar periodicity in the North Atlantic Oscillation. However, it is evident from a comparison of Figs 6 and 7 that the periodicities are often more clearly seen in the wavelet plots for Armagh meteorological data than in those for the NAO Index and SOI, where they form a broad spectrum of periodicities rather than a sharply defined peak. The implication could be drawn that the NAO, rather than controlling the meteorological series directly, is itself controlled by a common forcing mechanism. The physical nature of the driving force behind these oscillations remains unclear; however, the fact that one of them occurs near 22 years, the period of the solar magnetic (Hale) cycle, suggests that, in this case at least, solar activity may be involved. While these periodicities come and go, it may eventually be possible to use them in making long-term estimates of future conditions, such as total summer rainfall, once the driving mechanism is more clearly understood.”

    http://www.arm.ac.uk/preprints/2007/494.pdf

    RE275 I’m not sure I understand your question as phrased.

  279. Andrew
    Posted Mar 31, 2008 at 1:13 PM | Permalink

    277 (Chris Knight): I have to disagree. There are forcings and internal variability effects which have greater impacts on some parts of the Globe than others. For instance, temperatures at the coasts are often those most strongly influenced by ocean circulation changes. Regional scales are effected by way to many variables for just one effect to pop out.

  280. cba
    Posted Mar 31, 2008 at 1:40 PM | Permalink

    270 (Leif):

    If one runs an FFT on the data then the natural frequencies measurable by the sample collection, anything with 2 sample points per waveform or greater (as long as there is a full waveform worth), will come out with phase information (if you don’t ‘window’ the input data). There may be more modern and sophisticated approaches using wavelets etc. but this should work and it will work on white noise or even an impulse – which gives uniform frequency spread and uniform phase if the sequence is begun with it. There will be no filter factor involved in the data. In fact, one can use the FFT to filter frequency domain and return a filtered time sequence with an inverse transform.

  281. See - owe to Rich
    Posted Mar 31, 2008 at 2:04 PM | Permalink

    252 (Finn)

    Please see my study here. I think it answers most of your questions, but if you have more then please do post a reply over there.

    Basically, though I would not have done that work without Archibald’s papers and slides, I also tend to think that his rates of cooling are exaggerated. Your faith in NI temperature records may overlook that many sites will be subject to UHI. Nevertheless, the fact that the Armagh temperatures 1995-2004 are so high does disturb me. I think we have to consider cooling relative to that. My graph predicts a cooling of 0.5 degrees for Cycle 24 (2009-2019) assuming maximum in 2013.

    Rich.

  282. Pete
    Posted Mar 31, 2008 at 2:31 PM | Permalink

    274(Leif): “lambda Tuning”

    Leif, I’ve been experimenting with Anthony’s method using the Hadley Monthly data, so the recommended Lambda is 14,400 according to the author of the HP Filter. I also retuned at random to 10,000 / 6,400 / 25,000, so a bit of up and a bit of down. But the extraction is more or less the same no matter what Lambda is. It was not until I started reducing Lambda to below 1,000, and then it started tuning in more the curvature of the Monthly Series, as one might expect.

    Is this still what you would expect?

  283. Posted Mar 31, 2008 at 2:53 PM | Permalink

    278 (Anthony): To clarify: I sample a sine wave on the interval 0-62.83 at interval of 0.1. Add some random noise plus a trend over the whole interval. This gives me a series of 628 points numbered 1 through 628, containing exactly 10 waves. Now, I use HP. What lambda to use? not 100 as this is not yearly data, not 14400 as this is not monthly data.

  284. Philip Mulholland
    Posted Mar 31, 2008 at 2:57 PM | Permalink

    282(Pete): Lambda sensitivity

    At some small value of lambda, the HP filter returns the input data. Try setting lambda very small (e.g. 0.001) this lambda returns the input values for the data I am working on. Then try setting lambda = 1 and observe how the smoothing process works on the source data by plotting the raw and filtered data over a narrow time range.

  285. Posted Mar 31, 2008 at 3:02 PM | Permalink

    282 (Pete): I don’t know what I would expect. I know what I see and that is no solar cycle, 11 or 22 yr.

    280 (cba): am I supposed to disagree somehow? What you suggest is sensible and so should suffice for any purpose, even Anthony’s. So why doesn’t he do just that?

  286. Tomas S
    Posted Mar 31, 2008 at 3:13 PM | Permalink

    Interesting post at Atmoz blog where he gets a very nice correlating between sunspots and the satellite temp record. (is this OT or is it ok in this thread?)

  287. Tomas S
    Posted Mar 31, 2008 at 3:14 PM | Permalink

    I meant correlation of course.

  288. Posted Mar 31, 2008 at 3:35 PM | Permalink

    287 (Tomas): quite alright, and no need to correct minor mistkaes.

  289. John M
    Posted Mar 31, 2008 at 3:39 PM | Permalink

    re 286

    Note that Atmoz labels it “humor”.

    Beware the fool (April or otherwise) in sheep’s clothing.

  290. Tomas S
    Posted Mar 31, 2008 at 3:39 PM | Permalink

    Hmm, seems like that post is joke i fell for though. (1st of April you know)

    My only excuse is that i just quickly scanned the blog post and thought it interesting. :)

  291. AlanB
    Posted Mar 31, 2008 at 3:46 PM | Permalink

    Tomass = Atmoz?

  292. Anthony Watts
    Posted Mar 31, 2008 at 4:58 PM | Permalink

    RE285 “What you suggest is sensible and so should suffice for any purpose, even Anthony’s. So why doesn’t he do just that?”

    Actually one of our commenters has done just that (FFT) and reproduced essentially the same curve we have.

    “I’ve replicated your results using pure Fourier analysis with the (growing like Topsy!) C++ toolset I mentioned earlier.”

    Here is the raw HADCRUT3VGL monthly data with a low-pass filter which removes everything above the 15′th harmonic in frequency space – i.e., everything shorter than (150/15 = 10) years:

    Though it is smoothed a bit more than our example.

  293. Anthony Watts
    Posted Mar 31, 2008 at 5:02 PM | Permalink

    RE 283 I’ll play with that a bit

  294. Pete
    Posted Mar 31, 2008 at 5:35 PM | Permalink

    285(Leif): “I don’t know what I would expect. I know what I see and that is no solar cycle, 11 or 22 yr.”

    You’re correct. When zooming in 1950 to 2008, and using your TSI series, I can’t see it either.

  295. Posted Mar 31, 2008 at 5:43 PM | Permalink

    RE Leif (#275), Pete (#282), although the HP (Hodrick-Prescott) filter is widely used in economics, it is also sharply criticized for generating the appearance of cycles in data with no such patterns. Cogley and Nason (J. Economic Dynamics and Control 1993) show that it can generate cycles from a random walk, with the length of the cycles determined by the tuning paramter. HP recommended 1600 with quarterly data and 14400 with monthly data in order to generate business “cycles” that average about 8 years. Sure enough, when you filter US GDP etc with the HP filter using this tuning parameter, it shows regular 8 year cycles. The same trick should work for stock prices.

    If you wanted a highly persistent monthly temperature series to exhibit 11 year “solar cycles” whether or not they are present in the data, for example, it should be sufficient to use an HP filter with lambda approximately (11/8)*14400, or 19800.

  296. steven mosher
    Posted Mar 31, 2008 at 5:46 PM | Permalink

    RE 291. If Atmoz came on as Tomas S, I would give him double quatloos for being exceedingly clever. especially since leif mentioned filtering the series ass end first

  297. cba
    Posted Mar 31, 2008 at 5:47 PM | Permalink

    285 (Leif):

    Actually, I figured you’d probably say it would work BUT there was some newer slightly improved way of doing … (perhaps) using a wavelet approach …… – or something along that lines.

    As for someone else using / not using FFTs – I’m only guessing but it would be most likely because they didn’t know about it or how it might be available to them. Such things are perhaps a bit tricky, residing more in DSP than in statistical analysis.

    I was able to apply the FFT approach to some data using excel’s FFT add-in as an experiment the other day. It’s not a difficult thing to do. Most of my experience with them is programming them into realtime analysis applications and I have had real good luck using them for that application – which is unrelated to data analysis of this nature.

    I’ve also seen them used for filter applications rather than the usual processing intensive FIR filters etc. with very good results as well. It’s enough to make one wonder why bother creating DSP filters like this when it’s so easy in the frequency domain and evidently less processing intensive.

    I’ve also been an occaisional ‘user’ of wavelet technology and have witnessed some things that may even be superior performance but again, it’s not data analysis such as this.

  298. kim
    Posted Mar 31, 2008 at 5:53 PM | Permalink

    294(Pete) Eyeballing that there is a sense that the lag warming is longer than the lag cooling. Or am I daydreaming?
    ======================

  299. Andrew
    Posted Mar 31, 2008 at 6:32 PM | Permalink

    On the general subject of finding solar cycles in the temperature data, here’s my effort, looking for an eleven year cycle: First, I started with the annual anomalies
    Next, I separated eleven year periods into columns. Then, I did this over again. Then I sorted the numbers up, and down. Then I put the columns back, creating two new “anomaly” series. I averaged these. Then, I subtracted that series from the original. Here you go:

    Something is trying to pop out, but I think El Nino and volcanoes are obscuring it. So next I’ll try removing El Nino.

  300. Jan Pompe
    Posted Mar 31, 2008 at 6:33 PM | Permalink

    kim(298) Take a look at the slopes. Take into account El Nino which is stored internal energy being outed in 98 and the volcano blowing it’s top in the early 90s.

    The rate of cooling is sightly slower but the solar cycle is short compared to the relaxation time and the upturn comes sooner this should change with lengthening cycles. Time will tell.

  301. Posted Mar 31, 2008 at 6:48 PM | Permalink

    Pete, here are a couple of tired plots from the hurricane threads which vaguely resemble the curves in #294:

    and

  302. Posted Mar 31, 2008 at 7:28 PM | Permalink

    294 (Pete): we have reasonably good temperatures and sunspot numbers from 1850 onwards, so we should not zoom in on shorter pieces of the data. If there is a connection it must be there all the time.

  303. Anthony Watts
    Posted Mar 31, 2008 at 7:47 PM | Permalink

    RE295, Hu, we are using lambda of 100 and 7 rather than the suggested numbers (1600 with quarterly data and 14400 monthly).

  304. Basil
    Posted Mar 31, 2008 at 8:02 PM | Permalink

    Re #295

    Isn’t the concern with finding cycles where there are none in the detrended series, i.e. the cyclic component?

  305. David Archibald
    Posted Mar 31, 2008 at 8:22 PM | Permalink

    Steve: snip – please talk about this issue somewhere else.

  306. Basil
    Posted Mar 31, 2008 at 8:31 PM | Permalink

    #283

    “To clarify: I sample a sine wave on the interval 0-62.83 at interval of 0.1. Add some random noise plus a trend over the whole interval. This gives me a series of 628 points numbered 1 through 628, containing exactly 10 waves. Now, I use HP. What lambda to use? not 100 as this is not yearly data, not 14400 as this is not monthly data.”

    What is the objective? To see if HP can extract the original cycle from the noise and trend you’ve added to the signal? HP was designed for time series data, and the lambda’s are based on the frequency of time series data. I’m not sure how appropriate it is to use HP for what you describe.

  307. Raven
    Posted Mar 31, 2008 at 9:26 PM | Permalink

    This discussion of using complex statistical analysis to extract signals from data is starting to sound a lot like BCPs and PC1. I think I agree with Dr. Svalgaard – if the signal really exists in the data it should be possible to extract it using easier to understand techiniques.

  308. Erl Happ
    Posted Mar 31, 2008 at 9:50 PM | Permalink

    Back in Svalgaard 1/295 Stephen Richards said: ‘if the sun is a major facture in climate change then the climates sensitivity to the irradiance of our sun must be high, ie highly sensitive.

    And in 297 Leif was quick to agree. This post focuses on the question of amplification and the Earths response to changes in TSI.

    The satellite record of temperature in the troposphere is arguably the most reliable temperature series we have. The tropics is where irradiance is most intense on a surface area basis and the oceans are the part of the Earth system that is most capable of storing energy and responding to changes in irradiance. So, I have chosen to represent the Earths temperature in terms of the temperature of the atmosphere above the oceans, in the tropics.

    The satellite record of solar irradiance posted above (Svalgaard 4/204) is the best record that we have of changes in solar irradiance. We know that the temperature of the atmosphere responds to changes in irradiance, especially so in the stratosphere and the ionosphere. In the mid to lower troposphere the range of temperature change, on a monthly basis exceeds that at the surface and this suggests that part of that larger range is due to direct heating by elements of TSI. The other possible cause for this expanded range of temperature in the atmosphere is that the release of the latent heat of condensation warms the atmosphere while evaporation cools the surface. In this scenario all atmospheric heating (in the troposphere at least) might be due to surface heating.

    The date aligned graphs posted above show that an increase in solar irradiance is very clearly, (with several puzzling exceptions identified by solid rather than dashed rectangles), directly associated with a contemporaneous increase in temperature in the lower to mid troposphere over the oceans. This degree of symmetry could not be due to chance. Yes, temperature in the atmosphere seems to respond to solar irradiance. This is no big surprise. However, it is equally apparent that there is no response of temperature to the increase in TSI over the solar cycle as a whole. The response occurs over much shorter periods. This tells us that the Earth has built in resistance to temperature increase caused by increasing TSI. A lot of people looking for signals from the solar cycle in temperature data could note this.

    By my calculation the increase in irradiance over the period of the 1997-8 El Nino event at the upswing of solar cycle 23, from 1366.2 to 1366.6 watts per square metre, could of itself produce an increase in atmospheric temperature of 0.08°C. The actual temperature increase was in fact some 17.5 times greater at 1.4°C. This represents amplification of a high order.

    However, the greater increase in irradiance in 2002 from 1367.0 to 1367.8 watts per square metre produced a temperature gain of just 0.2°C. That is very close to the 0.17°K potential gain from the energy due to the irradiance alone. In this instance there is very little amplification. One might ask: Why does the amplification factor change?

    I hypothesise the atmospheric column in the tropics is conditionally sensitive (at various different elevations according to wave length of the radiation) to sunspot activity/TSI/ionising radiation and non-ionising but heating short wave solar radiation. I suggest that the effect of sunspot activity is to heat the atmosphere, make it less dense, displace it pole-wards, and in so doing promote the penetration of radiant energy in tropical regions. In penetrating the troposphere elements of this radiant energy raises the temperature of the air, reduces dew point, reduces relative humidity and therefore cloud cover. Cloud is the major factor responsible for attenuation of solar radiation and it is the change in cloud cover that provides the amplification. If the atmosphere was entirely heated from the surface change in cloud cover would be much less and large amplifications of TSI would be impossible.

    Since energy gain at the surface promotes evaporation (energy gain is almost completely resolved in increased evaporation rather than surface temperature increase in the tropical oceans) an increase in atmospheric transparency due to reduced cloud cover relies upon the change in irradiance being fast enough, and with sufficient persistence, to promote a net drying of the atmosphere against the countervailing force of rising humidity tending to restore cloud cover. Slowly achieved peaks in irradiance will not do the trick, no matter how high the peak. (See B in the figure above and consider this instance and the increase in irradiance over the entire solar cycle as illustrative of this point). The key to the strength of surface temperature response is therefore the rate of gain in TSI by comparison with the advance in evaporation. Nowhere in the solar cycle is this rate of gain in irradiance faster than in the upswing from solar minimum and this is where we commonly see the strongest El Nino events, a point that I have made in numerous posts on this thread. The strength of the gain in temperature at this time is facilitated by the coolness of the atmosphere prior to the upswing in irradiance. A prolonged La Nina cooling event prior to the upswing in irradiance (a frequent occurrence at the approach to solar minimum) establishes the pre-condition for rapid dehydration of the atmosphere because the absolute amount of moisture present in the atmosphere will be at its lowest point at this time. The atmosphere is pre-dried so to speak. Always the increase in irradiance carries the seed for the destruction of the amplification potential of future changes in irradiance. The amplification factor depends upon recent history i.e. the condition of the atmosphere prior to the event and the length of time since the last heating event.

    Here we are witnessing the application of heat causing warming and we would be very surprised if it did not happen. Does one need a correlation coefficient to validate the observation or is the coincidence of the periods of temperature advance with increases in irradiance sufficient? Given the importance of ‘rate of gain in irradiance’, and ‘pre-existing state of the atmosphere’ is a correlation coefficient an appropriate test of the relationship? I think not. We know that when heat is applied the object warms.

    Is the mechanism invalidated if a rise in irradiance occurs without an increase in temperature? In the figure above “A” denotes a period of strong increase in irradiance while temperature actually falls. The increase in cloud cover in this period is documented in my paper in figure 8 at http://www.happs.com.au/downloaders/Cloud_temp_tropo.pdf
    Similarly, the lack of amplification of temperature in 2002 due to the increase in humidity associated with the spike in irradiance in 2001 reinforces the point. Increases in irradiance are not always associated with temperature increase in the atmosphere.

    In further support of this concept I offer the following explanation: Consider that the tropopause is a zone where the sun (as a heating agent for the atmosphere) competes with the tendency for temperature to fall with altitude (due to the enhanced escape of long and short wave energy at diminished atmospheric density). Hence, temperatures fall with increasing altitude in the troposphere. But they rise with altitude in the stratosphere due to photo-disassociation of oxygen. The tropopause is the boundary between these two regimes. The energy gain above, at, and below the tropopause depends upon the ever changing emanations from the sun. Some of that energy must penetrate below the tropopause to cause the lapse rate to fall to zero at that point. Logically therefore the upper troposphere is heated directly by the sun. Since this is well above the cloud zone where latent heat is released the agent of atmospheric heating can only be irradiance in one form or another.

    Temperatures in the troposphere exhibit a greater range on a monthly basis than close to the surface as shown in figure 3 of the paper mentioned above. The middle and upper troposphere is heated jointly by the Sun and by the Earth. This statement is heresy but there is sense in it.

    Importantly, change in the balance between ENSO cooling and ENSO warming between solar cycles is a mechanism that must be taken into account in determining the origin of the current warming trend. In the paper cited above and previous posts I have shown that solar cycles 19, 20 and 21 were ENSO negative (cooling) and cycles 21, 22 and so far 23, ENSO positive (heating). Thus, the trend for temperatures to increase since 1976 owes a lot to solar influences.

    The historical record of temperature for European cities like London and St Petersburg reveals that low amplitude solar cycles (with little sunspot activity and small changes in irradiance) can produce periods of heating and cooling greater in intensity than high amplitude solar cycles. See also the diagram at my post Svalgaard 2/255. This suggests to me that changes in the upper atmosphere due to magnetic /solar wind/ionising radiation are also important in promoting ENSO heating events along with changes in those components of TSI that can simply heat the troposphere by virtue of the energetic but non-ionising components in TSI. Hence, although my past efforts to suggest how these effects might work have been seen to be un-physical, I persist with this suggestion. I see all energetic components of solar radiation tending to work together, diminishing atmospheric albedo and raising surface temperature. Exaggerated ENSO behaviour during weak solar cycles I regard as a critical observation that points to the importance of the mechanisms that influence the upper atmosphere. Solar irradiance, a function of sunspot number, changes little in low amplitude solar cycles. Some other aspect of solar activity like the solar wind interacting with the ionosphere and the neutral atmosphere must be responsible.

    The three heating events marked “C” are unrelated to changes in irradiance. The temperature of the atmosphere rises before irradiance increases or even as irradiance falls. This should raise eyebrows. Apparently heating occurs without the application of heat! The common factor is that these events occur at the low point of the solar cycle when the relatively invariable high speed wind of polar origin is the most important determinant of the character of the solar wind and the episodic flux from eruptive activity is least. My explanation is speculative. It may be possible that changes in the BZ component of the solar wind emanating from the little eruptive activity that does persist at solar minimum enables a penetration of the Earth’s magnetosphere by the powerful stream that emanates from the suns polar regions. Small changes in geomagnetic activity at this time (geomagnetic activity being merely an indicator of changes in the character of the solar wind rather than a force in itself) seem to be associated with large atmospheric temperature responses. Here, the solar wind rather than changes in irradiance is implicated in temperature change via change in the atmosphere’s albedo. This temperature change is disconnected from changes in irradiance in these “C” type events.

    Why some sunspot cycles should be ENSO positive and some ENSO negative could perhaps be related to the relative degree of ‘lumpiness’ in the occurrence of sunspot activity (the solar wind/total irradiance) as it affects the degree of amplification achieved or else the pattern of the solar wind. Since solar cycle 20 we have seen big swings in geomagnetic activity within cycles.

    In estimating the influence of El Nino dominant solar cycles on surface temperatures one should take account of further amplification due to the humidification of the stratosphere reducing the concentration of ozone and particulate matter. One would expect more sunshine at the surface of the Earth after a cycle where heating episodes were dominant. When temperatures in the stratosphere fall, one should expect atmospheric and surface warming.

    Thus, I maintain that it is the response of the atmosphere to emanations from the sun that is the driver for temperature change and heat storage processes on this watery planet Earth. Changes in TSI over time will be geo-effective but the degree of amplification is variable and the response likely to be different between one solar cycle and the next and at different points within each solar cycle. Where the warm water from the tropics travels markedly influences local temperatures at high latitudes. It is at high latitudes in the northern hemisphere that the recent warming has occurred. Change in carbon dioxide can not explain this partial warming. Changes in solar irradiance and the solar wind can explain it.

    TSI data source http://lasp.colorado.edu/sorce/news/2008ScienceMeeting/doc/Session1/S1_05_Dewitte.pdf

    Temperature data source: http://vortex.nsstc.uah.edu/data/msu/t2lt/uahncdc.lt

    Evidence of extra sunshine in England since 1980 http://www.timesonline.co.uk/tol/news/uk/article696586.ece

  309. kim
    Posted Mar 31, 2008 at 10:40 PM | Permalink

    302(Leif) Might not somewhat erratic oceanic oscillations mask some of the temperature correlation with the clocklike solar cycles? Eruptions should be easy to remove from the temperature data, oscillations not so easy.

    308(Erl) Water phase changes have the power to give a diversity of effects. I’m on my 5th reading of your post, and still learning.
    ======================================================

  310. Tomas S
    Posted Apr 1, 2008 at 12:23 AM | Permalink

    291(AlanB):

    Lol, no. I’m just a regular lurker here. Funny coincidence though.

  311. Posted Apr 1, 2008 at 12:36 AM | Permalink

    306 (Basil): time series are all in your mind. You can convert my example to a time series if you sample the sine wave in time, e.g. one sample every 0.1 month, then you get the full series after 62.8 months, containing 10 full cycles, each lasting ~six months.

  312. Posted Apr 1, 2008 at 1:03 AM | Permalink

    295 (Hu): Yes, it is clear that the various choices [100, 1600, 14400] for lambda aims at the same frequency [cycles per year]. Note that 1600=100*4^2 [quarterly] and 14400=100*12^2 [monthly]. This is no different from band-pass filtering 14C or 10Be data with an 8 to 12 year band and then claiming that you recover a 10-year cycle.

  313. AlanB
    Posted Apr 1, 2008 at 2:27 AM | Permalink

    Tomas (310) I am very sorry. Glad that you are real!

  314. Posted Apr 1, 2008 at 4:58 AM | Permalink

    And I’ll have to basically agree with Tamino’s analysis of the Watts correlation here.

  315. Posted Apr 1, 2008 at 5:06 AM | Permalink

    308 (Erl): most of the comment is beyond me and I’m not inclined to judge. There are some inaccuracies re solar physics, e.g.:

    Since solar cycle 20 we have seen big swings in geomagnetic activity within cycles.

    While not false, there are ‘big’ swings all the time, not just since cycle 20. E.g. in 1930 and 1952.

  316. Posted Apr 1, 2008 at 5:22 AM | Permalink

    In general, when people talk about the ‘peak’ of a solar cycle, they often forget that said peak is often very ill-defined. Here is what cycle 14 looked like:
    For Erl and others who match up detailed wiggles, cycle 14 provides ample material.

  317. Basil
    Posted Apr 1, 2008 at 6:27 AM | Permalink

    #314

    Leif, exactly what part of Tamino’s criticisms do you agree with?

    Tamino makes essentially two “valid” points, which should be taken for their merit, but not be used to close a blind eye to what we are seeing here. He actually agrees with what we did all the way through the derivation of the first differences plotted in our Figure 4, which we then compared to the peaks of the solar cycle. He then says we didn’t take all the peaks we see in the solar cycle. And that’s true. In Figure 4, there is a curious pattern of two smaller peaks around the peak of solar cycle 14. That’s certainly an anomaly that is worth investigating more carefully, but it doesn’t invalidate the rest of what Figure 4 shows, which is remarkably close correspondence.

    Perhaps I could put it this way? Anthony and I have to explain why 2 out of 16 peaks in our Figure 4 do not seem to match the Schwabe cycle periodicity we see in the results. Would it be too much, then, to expect Tamino (or Leif) to explain why 14 of 16 did?

    Then there’s the matter of our Figure 2. With higher order filtering, the odd numbered solar cycles are essentially filtered out, but we are still left with the a correlation to the odd numbered solar cycles, suggesting Hale cycle periodicity in the instrumental record. Again, if the basic technique is valid, and unless he recants Tamino basically agrees with the use of HP filtering here and how we derived Figure 4, we are seeing Hale cycle periodicities.

    Is this really so hard to imagine — that we might find something like this in the instrumental record? Actually, it has already been done, as noted in two largely overlooked or ignored papers which you can find here:

    http://scholar.google.com/scholar?q=brunetti+%22the+solar+signal%22&hl=en&lr=&btnG=Search

    Our analysis allows us to go further though, and not merely report the finding of solar cycle periodicities in the instrumental record, but a remarkable pattern of variable amplitude between even and odd numbered solar cycles that coheres remarkably well with current thinking about solar physics as to the role of magnetic pole reversal on GCR flux.

    As I said, Tamino makes two “valid” points. The other is that there may be an error in Figure 5 (nothing major, though), and that we cannot read too much, if anything, into the reported correlation coefficient.

    But honestly, if this is all that we’ve done “wrong,” the work stands up well to the worst that can be said about it.

    Basil

  318. Erl Happ
    Posted Apr 1, 2008 at 6:35 AM | Permalink

    315 (Leif) When one matches wiggles in temperature with irradiance there is no question as to cause and effect. There is no need to worry about correlations. What one has to worry about is why the relationship is not linear, seamless and perfect. Work that out and you have an answer to the sun/climate connection.

    If the wiggles are at ENSO time intervals there is little point in looking for a whole of solar cycle signal in the data. At the very least you should be able to conclude that the ENSO wiggle is due to changes in irradiance and that the response is non linear due to inteference from other variables….except obviously round solar minimum. The data shows that quite unambiguously.

    I would be more than happy with a simple statement of the sort: Yes, there could be something in this that is worth investigating. How how can we test these propositions?

    I don’t have temperature and irradiance data for solar cycle 14. Would that I had. As soon as one starts dealing with proxies like sunspots the queston as to causation rears its ugly head, then even perfect correlations get one no-where.

  319. Posted Apr 1, 2008 at 6:52 AM | Permalink

    317 (Erl): The irradiance follows the sunspot number closely enough. There were certainly a similar sequence of peaks in irradiance as in SSN during cycle 14 and we do have monthly temperatures from then, so the data is there. About the direct link between irradiance and temperature: if a few tenths of a watt is significant, then the 90 watt difference between January and July should be even more so.

  320. kim
    Posted Apr 1, 2008 at 7:24 AM | Permalink

    314 (Leif) What do you think of Tamino’s last graph in your link. Granted, it doesn’t reach significance. How about the slope?
    ======================================

  321. jae
    Posted Apr 1, 2008 at 7:32 AM | Permalink

    318 (Leif):

    if a few tenths of a watt is significant, then the 90 watt difference between January and July should be even more so.

    Is there that much change in the tropics?

  322. jae
    Posted Apr 1, 2008 at 7:33 AM | Permalink

    Erl: Do you think the cosmic ray theories have any validity?

  323. kim
    Posted Apr 1, 2008 at 7:35 AM | Permalink

    318 (Leif) Those few tenths of a watt difference are acting absolutely, the 90 watt differences relatively, since there are Januaries and Julies in each hemisphere. Maybe continentality has something to do with the magnification of effect.
    ==============================================================

  324. kim
    Posted Apr 1, 2008 at 7:37 AM | Permalink

    321 (jae) Clouds certainly seem to have a lot to do with the amplifying effect. Whether it is cosmic rays or more direct action of radiation on the aerospheres I’m less sure.
    ============================================

  325. jae
    Posted Apr 1, 2008 at 7:40 AM | Permalink

    323, Kim: I agree. It seems to be all about clouds (as Sam might say). Now, what causes clouds?

  326. Posted Apr 1, 2008 at 7:43 AM | Permalink

    319 (kim): since I think that the [temp] cycles are kind of spurious it makes little sense to correlate them with anything, but in any case, as Tammy points out, the slope goes the wrong way. I don’t agree with some of his rhetoric so shall not comment on that.

  327. kim
    Posted Apr 1, 2008 at 7:49 AM | Permalink

    325 (Leif) Granted we may be adding insignificance to spuriousness, but isn’t the slope as expected? I may just be exposing vast ignorance, for which I am not ashamed, but shouldn’t longer solar cycles correlate with shorter warming cycles?
    ============================

  328. Basil
    Posted Apr 1, 2008 at 7:54 AM | Permalink

    #311

    Leif, with that explanation, the default method of calculating lambda would be 2x2x100=400. That said, lambda is just a heuristic that determines, in effect, how much weight to give short term cycles versus long term cycles. Since temperature is likely to be the result of several, if not many, cyclic processes modulating each other in time, there is probably no one “right” value to use. It will depend, rather, on what kind of cycles one is looking for.

    On that, I do think it needs to be clear to everyone that we are using the results somewhat differently than the normal use in econometrics. For all the discussion of 11 or 22 year periodicities, we’re not deriving those from the cyclic component of HP filtering. In economic time series, that’s where the interest lies, and where the concern is in seeing cycles that are not there: in the cyclic component. We’re looking at the “trend” (smoothed) trend component, and the first differences, or time derivatives, of the trend component. And what we are seeing, after removing the cyclic component, is a residual cycle that when viewed in terms of its time derivative, shows a pattern of 11 and 22 year periodicity. And because we are looking at the residual cycle, and not the cyclic component itself, the results are relatively robust as to the value of lambda. With low order filtering, we see a pattern of 11 year periods, but with differing amplitudes. As the order of filtering increases, the lower amplitude peaks, corresponding to even numbered cycles, are effectively filtered out, but the higher amplitude peaks, corresponding to odd numbered cycles remains.

    Now when we get around to investigating the cyclic component, I might expect to find correlations with climate indices such as ENSO/MEI, etc. And there, the value of lambda may play a more critical role in “correctly” extracting the short term cycles. There, the common criticisms of HP filtering as applied to economic time series will be something to pay more attention to. But for smoothing the data, HP is effectively just a high pass filter, and lamda is what determines the cut-off point.

    Basil

  329. Posted Apr 1, 2008 at 8:28 AM | Permalink

    319 (kim):

    Those few tenths of a watt difference are acting absolutely, the 90 watt differences relatively, since there are Januaries and Julies in each hemisphere.

    no, there is only one January [and one July] and in that single one January both hemispheres receive 90 W/sqm more than they do in July, so the hemispheres do not cancel each other out. [sigh]

  330. kim
    Posted Apr 1, 2008 at 8:45 AM | Permalink

    329 (Leif) Touche. I should have said there are Summers and Winters in each hemisphere, instead, which cancel. That leaves the few tenths of a watt which isn’t cancelled to exert its effect. I do appreciate your patience.
    ===================================================================

  331. kim
    Posted Apr 1, 2008 at 8:49 AM | Permalink

    323 (kim) And again, can continentality contribute to the amplifying of that few tenths of a percent difference?
    ======================================================

  332. Basil
    Posted Apr 1, 2008 at 8:54 AM | Permalink

    On the length of the solar cycle, and Tammy’s graph…

    The length of the solar cycle is only crudely related to warming. Two solar cycles of equal length, and the same number of sunspots, will have very different warming impacts depending on the magnetic polarity of the cycle, whether it is parallel. A solar cycle with equivalent length to another, can have more sunspots, but produce less warming, if the phase is antiparallel. If there is a correlation between the length of the solar cycle and warming, it will be because parallel phases show some tendency to be longer. But the length of the solar cycle is not the causal factor in play, it is the magnetic polarity of the solar cycle.

    Those whose aim is to discredit what we’ve done (like Tamino), but who fail to place it in the fuller context of the physical mechanism we’ve cited to explain what we’re seeing, are just engaged in special pleading.

    Attempts to correlate “warming” (however measured) with sunspot numbers or the length of the solar cycle will forever remain limited to rough (and incomplete) correlations until they can formulate a model that captures the variation in hysteresis between even and odd numbered solar cycles. For an indication of the challenge involved, see Figure 3 in Mavromichalaki, et al, here:

    http://articles.adsabs.harvard.edu/cgi-bin/nph-iarticle_query?1996Ap%26SS.246….7M&defaultprint=YES&filetype=.pdf

  333. Erl Happ
    Posted Apr 1, 2008 at 8:54 AM | Permalink

    321 (Jae) The cosmic ray mechanism is driven by the solar wind (protons and electrons from the corona) that sweeps GC rays away by virtue of its strong magnetic signal. If it works we should see cold at solar minimum when the sun has few eruptions and consequently there are more GC rays around to promote cloud condensation. But, this is precisely the time when there is relatively little FUV to ionise the neutral atmosphere and little highly magnetised solar wind to open up a window in the tropics so the troposphere should be close to dew point and have lots of cloud at this time anyway. Secondly, I wonder if there is actually a shortage of cloud condensation nuclei in the atmosphere.

    I back the more direct notion of atmospheric heating and cooling by solar irradiance and the solar wind mechanism to explain changes in cloud cover.

    Some studies have shown that during Forbush events when the solar wind banishes GC rays cloud cover falls and there is less rain. I suggest that the lack of GC rays has nothing to do with it. It’s direct heating of the atmosphere via spurts in irradiance and the solar wind mechanism that does the trick. The lack of GC rays is co-incidental.

    Besides all that, look at the evidence. Is it consistently hot at solar maximum when there are few GC rays and cool at solar minimum? No. The pattern does not fit. The pattern of heating and cooling is ENSO.

    As to whether there is sufficient ionising radiation to heat the troposphere (a point that Leif has consistently disputed) consider this extract from Wikipedia:

    After the discovery of radioactivity by Henri Becquerel in 1896, it was generally believed that atmospheric electricity (ionization of the air) was caused only by radiation from radioactive elements in the ground or the radioactive gases (isotopes of radon) they produce. Measurements of ionization rates at increasing heights above the ground during the decade from 1900 to 1910 showed a decrease that could be explained as due to absorption of the ionizing radiation by the intervening air. Then, in 1912, Victor Hess carried three Wulf electrometers (a device to measure the rate of ion production inside a hermetically sealed container) to an altitude of 5300 meters in a free balloon flight. He found the ionization rate increased approximately fourfold over the rate at ground level. He concluded “The results of my observation are best explained by the assumption that a radiation of very great penetrating power enters our atmosphere from above.” In 1913-14, Werner Kolhörster confirmed Victor Hess’ results by measuring the increased ionization rate at an altitude of 9 km. Hess received the Nobel Prize in Physics in 1936 for his discovery of what came to be called “cosmic rays”.

    Cosmic radiation “high-speed protons and atomic nuclei”
    The Earth, and all living things on it, are constantly bombarded by radiation from outer space. This radiation primarily consists of positively charged ions from protons to iron nuclei derived from the sun and from other sources outside our solar system. This radiation interacts with atoms in the atmosphere to create secondary radiation, including X-rays, muons, protons, alpha particles, pions, electrons, and neutrons. The immediate dose from cosmic radiation is largely from muons, neutrons, and electrons, and this dose varies in different parts of the world based largely on the geomagnetic field and altitude. This radiation is much more intense in the upper troposphere, around 10 km altitude, and is thus of particular concern for airline crews and frequent passengers, who spend many hours per year in this environment. Here, the radiation exposure is not primarily due to the cosmic ray interaction with the thin atmosphere, but with the dense fuselage of the aircraft, causing relatively high background radiation in the cabin while the aircraft is at high altitude.

    It seems that there is plenty of heating potential in the troposphere from emissions from the sun both in terms of energetic non ionising radiation and also ionising radiation. Both forms vary with sunspot activity. Depending upon density and ionic populations the solar wind comes into play but it works directly with irradiance over all of the solar cycle except close to solar minimum. At minimum it seems to be seperately influential producing heating in the absence of increases in irradiance.

    Bottom line is that we dont need GCR’s to explain the changes in cloud cover. The pattern of irradiance change and solar wind variation is sufficient and the variations coincide with observed increases in temperature in the tropical troposphere. What else is needed?

  334. kim
    Posted Apr 1, 2008 at 9:08 AM | Permalink

    333 (Erl) There is no particular reason that both mechanisms, or more, may not be contributing to the amplification. I, too, once questioned the lack of cloud forming nuclei and was told that there are often episodes of supersaturation, greater than 100% humidity, that the cosmic ray generated cfn’s can work on. Can this be so?
    =======================

  335. Posted Apr 1, 2008 at 9:18 AM | Permalink

    Erl Happ says:
    April 1st, 2008 at 8:54 am

    I wonder if there is actually a shortage of cloud condensation nuclei in the atmosphere.

    http://earthobservatory.nasa.gov/Newsroom/NewImages/images.php3?img_id=16907

    seems to illustrate that there are areas starved of aerosols — there are other images. The ‘blue deserts’, which are spreading, will be very short of CCNs. I have seen a proposal to remove sulphur from bunker fuel. That should really help.

    JF

  336. jae
    Posted Apr 1, 2008 at 9:24 AM | Permalink

    Leif, 329:

    no, there is only one January [and one July] and in that single one January both hemispheres receive 90 W/sqm more than they do in July, so the hemispheres do not cancel each other out. [sigh]

    ???

  337. Andrew
    Posted Apr 1, 2008 at 9:26 AM | Permalink

    333 (Erl)

    Besides all that, look at the evidence. Is it consistently hot at solar maximum when there are few GC rays and cool at solar minimum? No. The pattern does not fit. The pattern of heating and cooling is ENSO.

    The pattern is “masked” by ENSO. See here:

    http://www.spacecenter.dk/publications/scientific-report-series/Scient_No._3.pdf/view

    If the process involved were direct heating, Climate would be impossibly sensitive. I’m not buying what your selling here.

    Aside from all that, the issues you take with the CRF->LACC theory are ill founded.

  338. jae
    Posted Apr 1, 2008 at 9:26 AM | Permalink

    I wonder how much of the cooling that results from major volcanic eruptions is due to the formation of more cloud nuclei (as opposed to reflection, scattering, etc).

  339. AlanB
    Posted Apr 1, 2008 at 9:43 AM | Permalink

    336 Jae

    In modern times, Earth’s perihelion occurs around January 3, and the aphelion around July 4 (for other eras, see precession and Milankovitch cycles). The changing Earth-Sun distance results in an increase of about 6.9% in solar energy reaching the Earth at perihelion relative to aphelion. Since the southern hemisphere is tilted toward the Sun at about the same time that the Earth reaches the closest approach to the Sun, the southern hemisphere receives slightly more energy from the Sun than does the northern over the course of a year. However, this effect is much less significant than the total energy change due to the axial tilt, and most of the excess energy is absorbed by the higher proportion of water in the southern hemisphere

    from Wiki-Earth

  340. Andrew
    Posted Apr 1, 2008 at 9:55 AM | Permalink

    333 (Erl): The full reason why CRF is cuasing the cloud changes and not TSI:

    However, variability in low cloud amount (LCA) correlates equally well with total solar
    irradiance (Figure 10), and cannot be uniquely ascribed to a single mechanism when
    using globally averaged data. This has lead to suggestions that the cosmic ray-low cloud
    link is a result of a tropospheric circulation response to TSI or solar UV (Kristjansson et
    al., 2002). One way to distinguish between these two processes is to utilize the property
    that cosmic rays arriving at Earth are modulated by the geomagnetic field whereas solar
    irradiance is not. Recent observational evidence indicates that the solar cycle amplitude
    in LCA, over the period 1984–2000, increases polewards and possesses a similar
    latitudinal dependence to that found in cosmic ray induced ionization (CRII) of the
    troposphere (Figure 13). This supports a physical mechanism involving cosmic rays
    rather than TSI or solar UV.

    If you really want to know about this stuff, this is the fullest treatment I’ve seen yet:

    http://www.spacecenter.dk/publications/scientific-report-series/ISAC_Final_Report.pdf/view

  341. Posted Apr 1, 2008 at 10:06 AM | Permalink

    330,336 (kim,jae): No there is no cancellation. The Earth is closer to the Sun in January than in July and BOTH hemispheres receive 90 W/m2 more in January than in July [I can't believe I have to say this twice]. If you want to nit-pick divide by 4.

  342. Posted Apr 1, 2008 at 10:15 AM | Permalink

    327 (kim): as this is spurious it is hard to make sense of, but I would say that the longer a cycle is, the more it warms. Here I assume that at solar minimum we get a certain base amount and that within each cycle we get a little bit extra due to solar activity, so the longer the cycle the more extra we get. But this is just hand waving, there are people that say we get cooling, warming, nothing, whatever. We are past logic here.

  343. Posted Apr 1, 2008 at 10:15 AM | Permalink

    For those looking for proxies to prove solar influence and/or climate sensitivity, might I suggest reviewing wind pattern anomolies (0-30 degrees, 30-60, 60-90)against the solar cycle rather than temperature. IF the cosmic ray theory is correct, or even partly so, we may be able to see it in wind pattern changes as viewed against the solar cycle. As we have all noticed in our lives, when a cloud passes overhead on a sunny day, a breeze can kick up in response to the temperature fluctuation on a local basis. El Nino and La Nina are in response to wind pattern changes, as are many of the changes being seen such as lower Arctic summer ice.

    Tucker

  344. jae
    Posted Apr 1, 2008 at 10:24 AM | Permalink

    339, AlanB; 341, Leif:

    Since the southern hemisphere is tilted toward the Sun at about the same time that the Earth reaches the closest approach to the Sun, the southern hemisphere receives slightly more energy from the Sun than does the northern over the course of a year.

    90 Wm-2 seems like more than a “slight” difference. Whatr is the total summer/winter difference at, say, 45 deg. lat.?

  345. Posted Apr 1, 2008 at 10:36 AM | Permalink

    333 (Erl):

    What else is needed?

    Can I suggest a little critical sense. There is no window through which the solar wind heats the tropics. Quoting [true, but] irrelevant stuff from Wikipedia does not help.

    It seems that there is plenty of heating potential in the troposphere from emissions from the sun both in terms of energetic non ionising radiation and also ionising radiation. Both forms vary with sunspot activity. Depending upon density and ionic populations the solar wind comes into play but it works directly with irradiance over all of the solar cycle except close to solar minimum. At minimum it seems to be separately influential producing heating in the absence of increases in irradiance.

    I’m a little sad that my hundreds of posts in this blog seem to have been in vain and for naught, if you still cling to the above near gibberish. :-(

  346. Posted Apr 1, 2008 at 10:44 AM | Permalink

    344 (jae): The quote is highly misleading and is not what the issue is. The ‘slightly’ more in the course of a year is because a Johan Kepler discovered, the Earth moves a little bit faster when closer to the Sun. But this is a very small effect. I’m talking about the huge effect of 90 Watt/m2 more input during January than during July, to BOTH hemispheres. [This is now the third time I say this]. It is precisely that large difference that is the reason we are not in the grip of an ice age right now.

  347. yorick
    Posted Apr 1, 2008 at 11:10 AM | Permalink

    There is a table on this link which breaks down percent of total insolation by lattitude. The next step is to adjust it for the difference in distance from the sun from month to month.

    Lenght of summer is not an insignificant consideration at some points in the Milankovich cycle, which varies due to Kepler, as Leif says. I just don’t know how different it is now. It is has been proposed that longer summers have more of an effect on rotting glaciers than warmer peak temps alone. During times when obliquity when oblliquity was different than now. Like during the Holocene Optimum (Am I still allowed to use that name?)

  348. Posted Apr 1, 2008 at 11:18 AM | Permalink

    Re Anthony 303, setting lambda = 100 with annual data is essentially the same as 1600 with quarterly or 14400 with monthly data, since sqrt(lambda) is approximately the number of periods at which the HP filter sets in. The HP filter itself preserves cycles longer than approximately this period, and smooths out cycles shorter than approximately this period.

    As Basil points out (#304), often what is of interest are the residuals after this “trend” has been subtracted out. If the series has the “typical economic power spectrum” discussed by Granger, ie the approximately hyperbolic spectrum of an I(1) series, the HP residuals will have the frequencies below the tuning frequency eliminated, so that a rounded peak will typically appear a little bit above this frequency. Eg if lambda = 100 with annual data, the spectral peak will appear at frequencies corresponding to periods of about 8 years, ie a little less than sqrt(lambda) = 10 years. Similarly for lambda = 1600 with quarterly data or 14400 with monthly data.

    I misspoke in #295, incidentally — in order to generate 11 year cycles in the HP residuals of a highly persistent monthly series, it would be necessary to set lambda = (11/8)^2 * 14400 = 27,225. I forgot to square the factor of 11/8 in my earlier post.

    RE Lief #312, the HP filter itself is just a lowpass filter, and its residuals are thus just high-pass filtered. The appearance of bandpass filtering comes when it is being applied to a highly persistent series like GDP (as is often the case in economics), since if you zero out the left half of a right hyperbola (or whatever the exact functional form of I(1) noise is), you end up with a function with a pronounced peak just to the right of the cutoff, without actually doing a bandpass. But yes, the period of the “cycles” that come out is just an artifact of the the tuning parameter that went in.

    The fashionable use of devices like the HP filter in econometrics starts to make “climate science” look positively rigorous! :=(

  349. yorick
    Posted Apr 1, 2008 at 11:19 AM | Permalink

    Also the effect of an increase in Antarctic ice would be heightened by the fact that it is exposed to the Sun during time of peak insolation due to eccintricity of the Earth’s orbit. I have no idea either about the magnitude of such an effect, just note that the Arctic is dark during this time, so it skips the peak entirely.

  350. jae
    Posted Apr 1, 2008 at 11:31 AM | Permalink

    Leif: thanks for your patience :)

  351. kim
    Posted Apr 1, 2008 at 12:04 PM | Permalink

    341 (Leif) Still, the 90 watt difference intra-annual is repetitive every year and shouldn’t have any effect on the decadal and centennial climate variations. Maybe the intracycle tenths of a watt difference don’t have much to do with them either, but something must explain the correlation between sunspot sparsity and cold weather. The Minima are solar and climatic.
    =====================

  352. Posted Apr 1, 2008 at 12:55 PM | Permalink

    351 (kim): in the protracted replying concerning the 90 W/m2 and one or two “Januaries” I think we lost the motivation for this, namely the claim by Erl that every little 0.01 W/m2 wiggle has a direct and immediate heating effect.

  353. Erl Happ
    Posted Apr 1, 2008 at 5:22 PM | Permalink

    352 (Leif)
    The extra 90 watts per metre TSI due to orbital considerations is reflected in the increased UV found at the surface in the Southern Hemisphere in February by comparison with that experienced in the northern hemisphere in July. However, the extra irradiance is readily absorbed into the oceans promoting a smaller atmospheric temperature increase than it would in the northern hemisphere. So it is that Margaret River is cooler in midsummer than Bordeaux despite the considerably higher latitude of the latter.

    The greater proportion of land mass in the Northern Hemisphere returns OLR to the atmosphere faster in July (than occurs in the Southern Hemisphere in February) thereby warming it and promoting a systematic 3% reduction in GLOBAL (not local) cloud cover as revealed in ISCCP data as it appears in Goode et al. Journal of Atmospheric and Solar-Terrestrial Physics. 69 (2007). This reduction in cloud cover is influential in terms of heat gain in the northern hemisphere in summer. If we knew the change in cloud cover in the northern hemisphere hemisphere we could balance the fall in cloud cover against the lower irradiance and work out the net effect.

    But these are short duration episodes in terms of ENSO and the solar cycle. It is the effect of the sustaned increases in irradiance and other possible influences that are of interest.

    Re: “I’m a little sad that my hundreds of posts in this blog seem to have been in vain and for naught, if you still cling to the above near gibberish.”

    Mars, I have read somewhere, has an atmosphere one hundred times thinner than Earth. Without a magnetic field as strong as Earths it has suffered depletion. Nevertheless I see Earths atmosphere as thin when one looks at the forces at play. Episodically, it is at risk. The combined influences of ionising radiation and the solar wind are the forces at play. Is it too much to suggest that these influences are not apparent in the troposphere where 70-80% of the atmosphere is located.

  354. Chris Knight
    Posted Apr 1, 2008 at 5:42 PM | Permalink

    341 (Leif): Not exactly true: During January the (a: area within the)Antarctic (b:Circle)is mostly insolated, with the (ditto a:)Arctic (ditto b:) mostly in darkness, during July the opposite applies.

    The hemispheric insolation always favours the south because of this, at this epoch. Precess the Equinoxes by 20,000-odd years, and there’d be something to talk about. But this has nothing to do with anything we should be discussing here.

  355. Posted Apr 1, 2008 at 6:43 PM | Permalink

    354 (Chris): yes, in this context it is exactly true. Of course, you have to modulate that with the Earth being round and tilted, and rotating. You could also nit-pick that we don’t get anything at night, etc. As you also say, all this doesn’t matter, so let’s stop stomping on the gnats.

  356. Posted Apr 1, 2008 at 8:07 PM | Permalink

    Because of all the confusion [?] and nit-picking about the variation of TSI, the following graph shows the TSI as observed by SORCE since the start of the mission in 2003:

    The red curve is what is actually observed at the Earth. This is what hits the atmosphere. This is the only thing that counts. The blue curve is what the Sun puts out [what is observed at 1 A.U.] and is interesting for solar physicists, but is totally irrelevant for climate. So matching up tiny wiggles in the blue curve with anything terrestrial is like worrying about the mote in thy neighbor’s eye, but ignoring the beam in thine own.

    Because of rotation, seasonal tilt, etc, the red variation is distributed [unevenly] over the whole globe and does not in itself cause any long-term effects because of the annual repetition, but it is absurd nonsense to think that our turbulent and complicated and interacting atmospheric and oceanic systems at all times precisely cancel out the very large red variation to the extent that we can ascribe any significance to the blue wiggles in real time.

    If you want to claim otherwise, you must show how this precisely cancellation happens, not just say that it does.

  357. Posted Apr 1, 2008 at 8:21 PM | Permalink

    355 (Erl):

    Episodically, it is at risk. The combined influences of ionising radiation and the solar wind are the forces at play. Is it too much to suggest that these influences are not apparent in the troposphere where 70-80% of the atmosphere is located

    Yes it is too much. Very rarely [only 58 have been observed since the first one was discovered in 1942] do energetic particles from the sun reach the ground in so-called GLEs [Ground-Level-Events, lasting from minutes to a few hours] and then at higher latitudes [not the tropics]. We have been down this road before: the solar wind and the FUV do not heat the troposphere, no matter how much of the atmosphere is located there.

  358. Erl Happ
    Posted Apr 1, 2008 at 8:22 PM | Permalink

    356 (Leif), 337 (Andrew)
    We are not talking about the weather here or the seasons. If the suns irradiance spurts line up with temperature increases in the tropical troposphere are you going to say it is entirely fortuitous?

    If they line up it means that the signal is clear. If there is amplification you have got the answer that you say you sought. If there is no solar cycle signal as such but just an ENSO signal can we forget about the ‘whole of cycle signal’ and start to concentrate on what the data is telling us….that ENSO is driven by the sun and that if we want to know why the Earth warmed we must look at the frequency and power of warming as against cooling events.

    This is elementary stuff.

  359. Erl Happ
    Posted Apr 1, 2008 at 8:28 PM | Permalink

    355 (Leif)
    I have never suggested that the solar wind heats the tropsophere. I am relieved to know that there are such things as ground level events. Like to know more about that.

    Sure as hell the stratosphere gets heated. What is doing it?

  360. Posted Apr 1, 2008 at 9:09 PM | Permalink

    359 (Erl):

    I have never suggested that the solar wind heats the tropsophere

    yet in earlier posts you said:

    It seems that there is plenty of heating potential in the troposphere from emissions from the sun both in terms of energetic non ionising radiation and also ionising radiation. Both forms vary with sunspot activity. Depending upon density and ionic populations the solar wind comes into play but it works directly with irradiance over all of the solar cycle except close to solar minimum. At minimum it seems to be separately influential producing heating in the absence of increases in irradiance.

    ‘solar wind’ bracketed by ‘heating’ does look like such a suggestion. But it is good to hear that it doesn’t happen; now, we just need to work on the FUV [the ionizing solar radiation] not heating the troposphere either and we shall be making progress.

  361. Posted Apr 1, 2008 at 9:15 PM | Permalink

    358 (Erl):

    suns irradiance spurts line up with temperature increases in the tropical troposphere

    you have to line up the red curve in #356 with whatever you are comparing it to.

  362. Erl Happ
    Posted Apr 1, 2008 at 9:21 PM | Permalink

    359 (Leif) The effect of the solar wind on the troposphere/stratosphere is indirect. The wind imparts energy to high altitude neutrals and shifts neutrals along with the ions. The effect is to open a window for all forms of solar radiation to penetrate to lower altitudes. Just a hypothesis to account for co-incidences that I see.

    Can we have your interpretaion of what is heating the stratosphere and why it is precluded from entering the troposphere?

    Can we have your interpretaion of the disconnect of heating events from irradiance close to solar minimum?

  363. Posted Apr 1, 2008 at 9:32 PM | Permalink

    362 (Erl):

    Just a hypothesis to account for co-incidences that I see

    Since the solar wind does not what you think it does ["opens windows"] you can’t based any hypothesis on that, and in any case if these event are co-incidences you don’t need any hypotheses.

    Can we have your interpretaion of what is heating the stratosphere and why it is precluded from entering the troposphere?

    Any textbook can tell you that [try Wiki, even], but whatever is absorbed in the stratosphere and heating it, evidently does not make it further down.

    Can we have your interpretaion of the disconnect of heating events from irradiance close to solar minimum?

    there isn’t any connection in the first place, so nothing to interpret.

  364. jae
    Posted Apr 1, 2008 at 9:53 PM | Permalink

    356, Leif:

    Because of rotation, seasonal tilt, etc, the red variation is distributed [unevenly] over the whole globe and does not in itself cause any long-term effects because of the annual repetition, but it is absurd nonsense to think that our turbulent and complicated and interacting atmospheric and oceanic systems at all times precisely cancel out the very large red variation to the extent that we can ascribe any significance to the blue wiggles in real time.

    If you want to claim otherwise, you must show how this precisely cancellation happens, not just say that it does.

    You are a devil’s advocate and a FENCE SITTER, par exellance. Now, if I may, do you have ANY idea about what cuses the MAJOR cyclic climatic temperature swings that are recorded in history, like the MWP and LIA? Do you have an opinion on whether it is not a Solar phenomenon? You seem to be against any and all ideas that don’t go with the IPCC flow, but you don’t seem prone to offer anything to explain what the heck is going on. Maybe you simply claim you do not know, which is OK by me. But please tell us if you have an opinion/theory/hypothesis/explanation for the CWP, MWP and LIA? Maybe you’ve answered this before and I missed it. If so, sorry.

  365. Posted Apr 1, 2008 at 10:41 PM | Permalink

    364 (jae):

    You are a devil’s advocate and a FENCE SITTER, par exellance.

    There are times where sitting on the fence is better than having fallen into the ditch [on either side]. Translation: I do not think the ‘science is settled’].

    Now, if I may, do you have ANY idea about what causes the MAJOR cyclic climatic temperature swings that are recorded in history, like the MWP and LIA? Do you have an opinion on whether it is not a Solar phenomenon?

    If the only evidence we had was the MWP and the LIA, then I would say that we cannot conclude that the sun is responsible, but there is other evidence that these may be part of ~1500 year cycles [Dansgaard, Bond, others] going far back. These may very well be solar-based. I may note here that we do not know what causes these longer solar cycles [if they are there], and also not the 80 year cycle, the 200 year cycle, the 420 year cycle, the 2300 year cycle, the …, etc. So, I’m uncomfortable with ascribing something we don’t understand [climate cycles] to something we also don’t understand [solar cycles], and I’m certainly most uncomfortable with pretending to the public that we do understand these things.

    You seem to be against any and all ideas that don’t go with the IPCC flow

    I can’t see how you can say something like that. What do I care what the IPCC flow is? I’m perfectly capable of reading the IPCC drivel and make up my own mind. Although one cannot be an expert on everything there are references and science is ONE. Any physicist can follow the arguments to their end if need be. BS stinks the same in any field.

    but you don’t seem prone to offer anything to explain what the heck is going on.

    If I could, perhaps I would be on my way to collect the Nobel prize. This is a large and difficult problem and we may not know for another hundred years what is going on.

    Maybe you simply claim you do not know, which is OK by me. But please tell us if you have an opinion/theory/hypothesis/explanation for the CWP, MWP and LIA? Maybe you’ve answered this before and I missed it. If so, sorry.

    I have said again and again [ad nauseam] that since the Sun varies less than we thought it did just a few years ago, the climate must either be much more sensitive to solar ‘forcings’ than currently assumed or have nothing to do with solar variability. I came on this blog to get an answer to that. I was reading that people understood the solar part and that they therefore with confidence could ‘subtract’ it and make assertions about other influences [e.g. human-caused]. I think I have shown that we do not have a handle on the solar equation and that we therefore must be less categorical in our claims when we foist these on the public.

    And I rail against any simple-minded and all-encompassing ‘armchar’-quality easy solutions because they detract from the real and difficult job ahead.

    Clear enough?

  366. Erl Happ
    Posted Apr 1, 2008 at 11:16 PM | Permalink

    363 (Leif)

    Re heating events at solar minimum.

    there isn’t any connection in the first place, so nothing to interpret.

    The interest arises precisely because of the lack of connection with irradiance when it is so well connected elsewhere.

    One question you appear to have missed is:

    If the suns irradiance spurts line up with temperature increases in the tropical troposphere (as the diagram in #308 reveals (outside solar minimum)) are you going to say it is entirely fortuitous?

    I well remember your suggestion that B should first be connected with A rather than C or D as proxies for A and that the relationship might be conditional on another factor F or G and should be approached on the basis of classification. That is what I am doing here.

    I well remember your instance of the solar wind and the BZ component (F or G) influencing the impact on the ionosphere.

    All I am doing is pointing out how the albedo effect of the clouds can provide the amplification necessary to produce a large reaction to a small stimulus. Along with that there is the suggestion that pre-conditions are important.

    These factors alone make the problem complicated and very difficult to model. Lets not pretend that it is more complicated than it needs to be.

  367. Posted Apr 1, 2008 at 11:23 PM | Permalink

    366 (Erl): when ‘lining up’, use the red curve not the blue. And you ask that the problem not be made more complicated than it is. I ask that the problem not be made less complicated than it is.

  368. Posted Apr 1, 2008 at 11:47 PM | Permalink

    From the The Lowell Observatory blog commenting on the SORCE 2008 Santa Fe Meeting:

    “We turned to the stars in the latter part of the afternoon. Tom Ayres presented a very nice review of results on Sun-like stars, leading into my talk on the activity and brightness behavior of solar analog stars. A key question here is: what happens to stars when they enter Maunder minima? A faint Maunder Minimum Sun could mean the Sun is more variable on century timescales than we think it is, but the stellar data just don’t support that hypothesis. In particular, observations of a star that likely transitioned from a flat activity to a cycling state in 2002 don’t show any significant total change — yet — in its brightness. That’s an unpublished result, so I’ll say more about it once we get it into a journal. In any case, since ice core records show the solar cycle didn’t completely shut down during the Maunder Minimum, and since recent observations (both by us and by others) of flat activity stars also provide evidence that such stars have not subsided into an anomalously low magnetic state, we do not think that the Maunder minimum Sun was necessarily fainter than it is at the modern solar minimum. This emphatically not to say that the Sun cannot exhibit a long term trend in its overall brightness — in fact, on a sufficiently long timescale, we know that it will. But the present data indicate that grand minima per se are not important drivers of total solar or stellar brightnesses.

  369. Erl Happ
    Posted Apr 2, 2008 at 12:29 AM | Permalink

    368 (Leif)

    the present data indicate that grand minima per se are not important drivers of total solar or stellar brightnesses.

    All the more reason to examine the effects of the minute variations in brightness (and anything else that varies) that occurs within and between cycles and get down to the nitty gritty of looking at the temperature response that occurs in the atmosphere and the oceans. Vital too to ask where that response is apparent.

    Re the choice of curves. If anything I looked at the green curve (optical) that I assumed best related to TSI. But, does it simply relate to visible light? Please explain significance of these choices.

  370. Erl Happ
    Posted Apr 2, 2008 at 12:40 AM | Permalink

    368 (Leif)

    since ice core records show the solar cycle didn’t completely shut down during the Maunder Minimum, and since recent observations (both by us and by others) of flat activity stars also provide evidence that such stars have not subsided into an anomalously low magnetic state,

    The ‘magnetic state’ refers to the solar wind influence on Earth process I presume. How does it manifest in the ice cores? What is the inference as to the climate? If there are no sunspots I imagine changes in irradiance are slight. This means that we must countenance a solar wind mechanism. What is that then? I dont see the planetary wave argument as offering a great deal of oomph.

  371. EW
    Posted Apr 2, 2008 at 4:08 AM | Permalink

    332 (Basil) I looked at the Brunetti paper about the alpine climate in your post. It’s interesting that in the 19th century they observed clear correlation between lots of sunspots and lots of cold and storms, and wrote papers about that, even arguing about who came up with it first.

    Looking at the Fig. 4, the pretty inverse relationship changed at the end of 19th century, giving rise to these no less pretty parallel Fries-Lassen type curves for sunspots and temperature, until the end of 20th century. Hmmm…maybe after some 15-30 years we can start write new papers about the sunspots bringing storms ;-)

  372. Posted Apr 2, 2008 at 4:40 AM | Permalink

    369 (Erl): the red curve of #356.
    370 (Erl): the magnetic state is the solar magnetic field. What Jeff Hall means is that the field during the Maunder minimum was just as it is today at a regular solar minimum.
    371 (EW): this is typical of an immature field where people correlate left and right, and correlations come and go. Just shows that we have not made any progress in sun-weather-climate relations. The simplest explanation is that there isn’t any.

  373. Posted Apr 2, 2008 at 4:51 AM | Permalink

    370 (Erl): I forgot above the ice cores. The sun’s magnetic field configuration controls the modulation of cosmic rays and hence the production of 10Be deposited in said ice cores.
    The solar magnetic field also controls the production of slow solar wind [c.f. results from the Hinode spacecraft released today here], which in turn creates interaction regions when it collides with faster solar wind. These regions help to screen us from cosmic rays. All Hall is saying is that all that seems to have operated as usual at current solar minima during the extended Maunder Minimum.

  374. EW
    Posted Apr 2, 2008 at 4:52 AM | Permalink

    372 (Leif): It’s interesting that the correlations and anticorrelations were quite stable and predictable for almost one century. It wasn’t as each solar cycle the relationship was random, different or none at all. And the switching of phases was quite abrupt. I really wonder, if we are into the inverted relationship again.

  375. Posted Apr 2, 2008 at 5:08 AM | Permalink

    374 (EW): The climate does have internal [?] decadal cycles. If unrelated to sunspots, the two curves will match for a while, then drift out of match and ‘anti-match’, then drift back into match, etc. Add some noise, and the phase shifts may seem abrupt. People looking for correlations will stick with one phase as long as possible, unwilling to give it up, until forced to do so by an obvious anti-phase. Thus it is people that change their mind abruptly rather than Mother Nature.

  376. Posted Apr 2, 2008 at 6:13 AM | Permalink

    369 (Erl):

    Re the choice of curves. If anything I looked at the green curve (optical) that I assumed best related to TSI. But, does it simply relate to visible light? Please explain significance of these choices.

    The graph itself clearly states that the red curve [#308] is ‘measurements’ and that all the others are various model calculations. So the red curve is the only legit one.

    But that was not the curve I meant. I meant the red curve in comment #356.

  377. Pete
    Posted Apr 2, 2008 at 6:17 AM | Permalink

    294(Pete): Myself

    I wish to illustrate, by using my own graph in #294 as an example, how we must be very aware how we are analysing data. The point I am making is that someone will need to re-adjust the entire Global Data record if you have any chance of achieving a more robust Solar Correlation.

    This example is perfect. It is a recent event, well monitored, and the most studied affect of just one natural influence on the Climate so far made. It is also an excellent example as it should highlight to anyone the pitfalls of just using raw Global Data.

    Question: How can 1 Volcano (and it was only a VEI of 4) across a 3-4 year period affect just one section of an analysis. Here is the answer.

    In #294 there was a rather nice 50 year analysis of a nice perbutation in the Global temperature lagging the Solar Cycle. It doesn’t explain the overall trend (and never did try to show this). Let’s just concentrate on that little 0.1C-0.2C fluctuation.

    Something went a bit wrong at Solar Cylce 22 though. The Temperature ‘bump’ appeared to occur closer to Solar Max with less of a lag than the rest of the series. More worrying the Solar Minimum of Cycle 22/23 appeared to come after the global temperature had already dipped. It was shown in #294 as the only ‘leading’ line.

    How can I explain this oddity? Below I’ve plotted the original data again (in Magenta) taken from the Raw Hadley Monthly Series. After researching the concensus of the effect of the 1991 Pinatubo eruption, I used the information I gathered – which was around a 0.4C drop from 1992 and then reducing slowly until the effect vanished by 1995. I spliced in my rough adjusted data and this is shown in Yellow (Monthly) and Light Blue (Filtered).

    After re-filtering, the original Magentia dip which happened at the wrong place, is now a Hump shifted forwards by around 2 Years from Solar Max. Excellent. My trough is also moved forward by 2 years, re-asserting a Lag now between Solar Min in 1996 with the dip in the temperature Min.

    Just one Volcano with a VEI of just 4, with the effect lasting just 3-4 years caused this error. Now think about applying this not only to Volcanoes since 1750, but Ocean oscillations as well, and you can see the task that is needed to reform the global temeprature to take into account all these events.

    Many of you will know this. But in 1883 there was a very famous eruption with a VEI of 5. Krakato. Don’t let the VEI number confuse you. It is like the Earthquake scale, So a VEI 5 has about 10,000 the Mass of the Material ejected than a VEI of 4.

    But, I bet most of you do not know that in the early 1770′s there were two eruptions of VEI 4 scale. How do we know how this affected the global temperature for 2-3 years after? The point Tamino concentrated on in his rebutal was that the Solar Peak in that Cycle was in 1770, but the temperature appeared to Peak in 1777? He specifically highlighted this faulty correlation. Unfortunately very little is known about these two VEI 4 eruptions in the 1770′s. Just like in 100 years time the VEI 4 of Pinatuba will not really be remembered for anything other than a few years of slighly lower temperatures. Not like the VEI 5′s and VEI 6′s, such as Tambora and Krakatoa to name but 2.

    So, be wary!

  378. Basil
    Posted Apr 2, 2008 at 6:26 AM | Permalink

    #375 (Leif)

    If unrelated to sunspots, the two curves will match for a while, then drift out of match and ‘anti-match’, then drift back into match, etc. Add some noise, and the phase shifts may seem abrupt.

    Couldn’t this be true even if related to sunspots, or some other solar driven process? I.e., two curves will match for a while, and then drift out of match because of non-solar influences. I’m thinking here of course of both short and long term ocean oscillations which have their own dynamics, and may at times be in cycle with solar influences, and at other times at odds with them. The moving of the curves in and out of match does not necessarily demonstrate that climate is unrelated to sunspots; it may just reinforce what you said in #365:

    And I rail against any simple-minded and all-encompassing ‘armchar’-quality easy solutions because they detract from the real and difficult job ahead.

    It might be that what Anthony and I have done came across as an “all-encompassing ‘armchair’-quality easy solution” but that wasn’t intended. There’s no question that there is a “real and difficult job ahead” in terms of understanding if and how solar processes influence terrestrial climate. I chose the term “solar imprint” in describing the results we found precisely because of its less than precise implication. With maybe one significant exception, what we’ve found simply replicates what has been found in numerous studies suggesting Hale cycle periodicities in various climate indexes or proxies. Which brings me to my question, if you’ve been kind enough to read this far.

    What is your “take” on all the evidence claiming to find Hale cycle periodicities in climate proxies or indexes? Are you entirely skeptical? Are you perhaps just agnostic about them? Do you think they could be indicative of a “solar connection” but that until a connection can be established to a specific physical mechanism, that the most we can say is “Well, that’s interesting, but I don’t know what it proves?”

    Thanks for all the time you put in here. Reading your exchanges with others is very educational.

    Basil

  379. Pete
    Posted Apr 2, 2008 at 6:30 AM | Permalink

    377(Pete) – Myself.

    Sorry – Volcanoes in the “1770s” should read as “1870s”

  380. kim
    Posted Apr 2, 2008 at 7:30 AM | Permalink

    365 (Leif) Despite the isotope data, don’t the sparse, large and southern sunspots during the Maunder Minimum have some meaning? Surely there is a clue, there. And please, call me Un Shirley.
    =============================================================

  381. kim
    Posted Apr 2, 2008 at 7:36 AM | Permalink

    377 (Pete) The Argo buoys are demonstrating that the GCM’s do not model the oceanic oscillations well, and of course, we have poor historical data about them. I suspect they are the noise masking a more obvious sun-climate connection. However, I’m unsurely.
    ===========================

  382. Erl Happ
    Posted Apr 2, 2008 at 8:53 AM | Permalink

    376 (Leif)
    Re the red curve in #356 and the statement in 375 that “The climate does have internal [?] decadal cycles”.

    Strange as it may seem my interpretation of the anti-correlation of radiation (red curve) with terrestrial temperature is in #353. The source of the anti-correlation lies in the difference in the distribution of land and sea in the two hemispheres. As a result, extra insolation in January is devoted to warming the oceans of the Southern Hemisphere. The seasonal reduction in cloud cover in the northern hemisphere in June-July opens up the window to admit more solar insolation. The reduction in cloud cover in the N. Hemisphere promotes such a rise in air temperature that any latitude in the north tends to be warmer than the comparable latitude in the southern hemisphere in the comparable summer season. So, the marked reduction in radiation intensity in July (due to orbital effects) is negated and overwhelmed by the reduction in cloud cover.

    The window, and it’s opening and closing is the key to terrestrial temperature response to the sun. What is to be established is how the sun does the opening and closing. The simplest mechanism is via direct solar heating of the atmosphere. Other proposed mechanisms like Cosmic Rays offering condensation nuclei may contribute. Both respond in the same way to solar forcing so as to be additive in effect.

    A reduction in cloud cover offers dramatic possibilities for amplification with enduring consequences for terrestrial temperature. The ocean can soak up lots of irradiance without influencing surface temperature greatly. As it does so, outgoing long wave radiation falls in the area where cloud disappears as is clearly shown in Svalgaard 3/510. The Earth stores heat and warms during ENSO events. The last three solar cycles have been ENSO positive.

    ENSO is not temperature neutral. It is not an ‘internal oscillation of the climate system’ or climate noise to be filtered out to get the true signal. It is the signal. It represents temperature change in that part of the globe where irradiance produces the greatest warming effect, where ultraviolet is strongest and humans must have brown skins to survive without going ‘Troppo”.

    ENSO is plainly driven by the sun.

    Leif You say that “The climate does have internal [?] decadal cycles.”

    I appreciate the significance of the question mark.

    I suggest that the decadal bit comes from the very frequent La Nina at solar minimum. Solar maximum has a much less definite central tendency and this may be due to high atmospheric moisture levels as a result of the big El Nino that occurs in the sunspot/irradiance upswing.

    334 (Kim) I agree.

  383. Posted Apr 2, 2008 at 9:11 AM | Permalink

    308 (Erl): Let’s look closer at the ‘lining up’ problem. Here is a part of the 308 plot:

    The circled ‘event’ seems to be a good heating event [the biggest besides the 1998 el Nino]. Now, first we note that the red curve in the plot is the one to compare with. This curve is a 121-day running mean of TSI. The temperature curve is [I think] monthly anomalies, i.e. with the seasonal [whatever that is in the Tropics] variation taken out. Erl lines up the TSI local max in the last half of 2004 with the temp max in early 2005. That is ok as one might assert that some small delay is ‘natural’.

    Let us now plot the TSI observations:

    The light-blue line is the 121-day running mean. The peak chosen by Erl is marked with the blue arrow. [We are looking at the top panel now]. So far, so good. The dark-blue curve is the unsmoothed daily values of TSI, and one can indeed discern a small rise of 0.3 W/m2 of the background corresponding to the smoothed peak. The trouble begins when we realize that the dark-blue curve is not what the Earth is exposed to, but is TSI reduced to 1 AU. The actual TSI that impinges on the Earth is given by the red wavy curve. The lower panel shows the three curves to the same scale and we now see the absurdity of lining up the little wiggles [where the red arrow points] that are completely swamped by the orbital variation. If anything one might argue that the temp curve matches the overall red curve rather well in the sense that there is 90 W/m2 extra in January 2005, matching the similar peak in the temps. I would actually expect some kind of match-up like that, as it seems to me that the atmosphere cannot possibly be so finely constructed as to always perfectly canceling out the 90 W/m2 variation. As to why the temp peaks line up with the little TSI wiggles, well, sometimes they don’t, e.g. 1998, 2007. So, in short, this whole thing is completely spurious, regardless of how much we would like it to be true, because ‘obviously’ the sun is the driver of climate and weather.

  384. Andrew
    Posted Apr 2, 2008 at 9:32 AM | Permalink

    358 (Erl Happ): The problem is that their is a solar cycle signal, and I’ve linked at least twice to sources which demonstrated just that!

  385. Posted Apr 2, 2008 at 5:27 PM | Permalink

    378 (Basil):

    What is your “take” on all the evidence claiming to find Hale cycle periodicities in climate proxies or indexes? Are you entirely skeptical? Are you perhaps just agnostic about them? Do you think they could be indicative of a “solar connection” but that until a connection can be established to a specific physical mechanism, that the most we can say is “Well, that’s interesting, but I don’t know what it proves?”

    That is not how it works. Suppose you have a phenomenon and 1000 theories claiming to explain the thing. Being agnostic [without knowledge - or as Huxley who coined the term meant: not professing to a belief in something that cannot be proven] about them would mean that I would treat them all equally. But surely, many of them are ‘not even wrong’, but just junk. A few may deserve more attention. They have to ‘earn’ my attention by being physically/statistically sound, i.e. they have to be ‘convincing’ enough to me that I find them worthy of further study. And, so far, none of the Hale cycle correlations have met that standard. There is another reason as well, namely, the longer you make the period, the fewer of them we have observed, so the smaller their statistical significance. We are sure there is a rotational cycle [of 27 days] because we have observed thousands of them, but we have only about 4 examples of the 88-year so-called Gleissberg cycle, so we are not so sure that that cycle is real or coincidence.
    There are other criteria: If I have a physical theory [or understanding] from which I calculate that something should happen, and it does, I may only need one or a few cases to establish the point. But if all I have is a fit to past cases with no understanding of why those came about, then the burden of proof is a lot harder [even if turns out that my 'prediction' was 'correct']. We have a good case with the prediction of the size of the next sunspot cycle. At last count there were ~50 published [and mostly peer-reviewed] ‘predictions’ ranging from 40 to 250. One or more of these are going to be ‘correct’ in the sense that they land on or near the right number. Only a few [perhaps two or three] are based on physics, and these even differ by a factor of two [75 and 160]. How can they differ if they are based on sound physics? Because there are some ‘boundary-conditions’ we don’t know: is the solar dynamo shallow [75] or deep [160]. If one of these numbers [75 or 160] is close to what the Sun ultimately comes up with, we have learned something about which boundary condition is likely to be the better approximation to reality. And our next prediction for cycle 25 can then be put on a stronger footing. If that prediction is also correct, I think most will accept that we ‘know’ what is going on. If solar cycle 24 should come out significantly different from either prediction, they were both wrong and we have to go back to the drawing board and find what we were missing. Only true and honest prediction rather than postdiction will count here. And, returning to your question, we have no understanding of how a Hale cycle period can happen in climate, so we can hardly make a compelling case, until we have such understanding or [as with rotations] we have accumulated enough cases that statistics alone can decide. IMHO, we are not there yet, as simple as that.

  386. Richard Sharpe
    Posted Apr 2, 2008 at 5:37 PM | Permalink

    Leif (385 and all the others):

    What you say makes an extraordinary amount of sense, just don’t get a big head :-)

  387. Posted Apr 2, 2008 at 6:09 PM | Permalink

    386 (Richard): my tirade was boilerplate scientific method [not my invention]. I realize that there are a significant number of lurkers that may benefit from hard-won insight by practicing scientists. This is also sorely needed elsewhere. I just an hour ago saw this:

    MODELS LOOK GOOD WHEN PREDICTING CLIMATE CHANGE, April 02
    The accuracy of computer models that predict climate change over the coming decades has been the subject of debate among politicians, environmentalists and even scientists

    [emphasis added].

  388. Not sure
    Posted Apr 2, 2008 at 9:18 PM | Permalink

    264(Leif) Dr. Svalgaard, cool graph. Out of curiosity, why not include the cycle with a minimum around 1997?

    Also, at least one person who claims that cosmic rays influence the climate would say that proxies like CLIMAX and the Moscow neutron monitor are irrelevant because they measure the flux of cosmic rays whose energy is too low to influence the climate.

    http://www.realclimate.org/index.php/archives/2006/05/thank-you-for-emitting/

    http://www.sciencebits.com/CO2orSolar

    (Relevant bits are in the comments for both links.)

  389. Erl Happ
    Posted Apr 2, 2008 at 11:00 PM | Permalink

    384 (Andrew) Dont be shy. Lets have those references again.

  390. Erl Happ
    Posted Apr 2, 2008 at 11:28 PM | Permalink

    383 (Leif) We are looking at a smoothing function on a fluctuation that occurs over years, not days or months.

    The climate system itself dampens the response. The oceans are the flywheel. You would not expect a response to the gyrations in the blue curve. Why expect it to the red curve?

    One little problem that one has in lining up graphs is that people never tell you whether the observations lie on the dates or between them. Then the computer has a little resolution problem and won’t let you place a line on the spot that it needs to be on. Next thing is you change the scale to fit it to a page and the whole thing goes out of whack and you start again.

    But none of this is important because the pattern sticks out like DB’s anyway and the peaks match in a most excellent fashion.

    So far as the peaks that don’t match are concerned I have this quotation that is apt:

    “Is there any point to which you would wish to draw my attention?”
    “To the curious incident of the dog in the night-time.”
    “The dog did nothing in the night-time.”
    “That was the curious incident,” remarked Sherlock Holmes.

  391. Dennis Wingo
    Posted Apr 2, 2008 at 11:45 PM | Permalink

    Leif

    Would you update your cycle progression charts?

    Thanks

  392. Dennis Wingo
    Posted Apr 2, 2008 at 11:57 PM | Permalink

    394 (Pete).

    Sorry – Volcanoes in the “1770s” should read as “1870s”

    Does this correlation extend to the two somewhat earlier volcanos of known global climatic effects near the end of the LIA? (Laki in Iceland in 1783 and Tambora in 1814-15). We know for a fact that in England and North America that there were significant climate effects from these Volcanos.

  393. stargazer
    Posted Apr 3, 2008 at 1:17 AM | Permalink

    http://news.bbc.co.uk/1/hi/sci/tech/7327393.stm

    from link

    “Rutherford-Appleton Laboratory. He showed that over the last 20 years, solar activity has been rising, which should have led to a drop in global temperatures if the theory was correct.”

    Huh…this is backwards. ‘rising’ activity would give a warmer earth. and low activity a cooling earth. did I miss somthing ?

  394. EW
    Posted Apr 3, 2008 at 2:11 AM | Permalink

    393(stargazer)

    Mike Lockwood from the UK’s Rutherford-Appleton Laboratory. He showed that over the last 20 years, solar activity has been rising,

    And that being the same Mike Lockwood that said in his ‘sceptics debunking” article there is no increase in solar activity that might explain the recent temperature rise?

  395. Pete
    Posted Apr 3, 2008 at 2:51 AM | Permalink

    393(stargazer):

    This report is based on This Paper

    May be Leif could comment? Seems to me they found a strong anti-correlation for Cycle 22 (The Volcano Cycle by-the-way), but not so much for Cycle 23. They then just lumped them together as a 20 Year period and said it all anti-correlated. Odd really.

  396. Pete
    Posted Apr 3, 2008 at 2:59 AM | Permalink

    392(Dennis Wingo): Does this correlation extend to the two somewhat earlier volcanos of known global climatic effects near the end of the LIA?

    Dennis, the two Volcanoes you mention are well known. There were far more VEI 4 eruptions than I thought when I checked the records. Over a Dozen. Now, not all Volcanoes affect Climate, depends on their locations and composition.

    My point was, that I showed how 4 years of accounting for just 1 Volcano can affect the trend of the filtered data to such an extent it actually altered where the Peaks and Troughs were.

    I would say this is significant when nit-picking over the Peaks of actual Solar Cycles and Warming Peaks. How can one do it without considering these internal short term Climate Changing events.

  397. Posted Apr 3, 2008 at 4:05 AM | Permalink

    390 (Erl):

    We are looking at a smoothing function on a fluctuation that occurs over years, not days or months.

    if so, why line up wiggles on a shorter time scale? Anyway, I do recall you saying that it is not the slow year-scale fluctuations that matter, but the short, sharp impulses. I always thought that only the longer-term fluctuations would be important, and I’m glad that you have come around to see that as well.

    You would not expect a response to the gyrations in the blue curve. Why expect it to the red curve?

    If by ‘you‘ you mean ‘me‘, the answer would be: “because the red variation is a hundred times bigger”.

    the peaks match in a most excellent fashion.

    “`Contrariwise,’ continued Tweedledee, `if it was so, it might be; and if it were so, it would be; but as it isn’t, it ain’t. That’s logic.’”

    Dear Erl, there is a time when it is prudent to back off from an untenable position. For you, that time has come.

  398. Posted Apr 3, 2008 at 4:37 AM | Permalink

    388 (Not…): In #11 I showed the run of cosmic ray flux for the past several cycles. I noted the alternating peaked and rounded maxima [which are understood in terms of drifts and polar fields]. Every other flux maximum [at solar minimum] can thus be compared. The 1997 [and the 1976] minimum belongs to the other set [with rounded maxima] and can thus not be directly compared to 1965, 1987, and 2008, because the shape is different.

    I’ll let the cosmic ray proponents have their slugfest in peace as to what is important and what is not. Once they have settled that, they can come and tell me, and I’ll look at what they have.

    391 (Dennis): which charts?

    393,394,395 (EW,Pete): I would tend to agree with the conclusion of the paper cited: “We have examined this hypothesis to look for evidence to corroborate it. None has been found and so our conclusions are to doubt it.”
    About Mike Lockwood: he has been saying “rising” ['doubling in the last 100 years'], but has now backed off ['falling since 1985 or so']. Also, his utterings have been filtered through the media, and they get things backwards at times. Mike’s problem is the same as Erl’s: not knowing when it is time to slink away from an untenable position.

  399. AlanB
    Posted Apr 3, 2008 at 5:22 AM | Permalink

    394 (EW) As usual Leif is quite correct

    his utterings have been filtered through the media, and they get things backwards at times.

    and the BBC post now reads:

    Mike Lockwood from the UK’s Rutherford-Appleton Laboratory …

    showed that over the last 20 years, solar activity has been slowly declining

  400. Andrew
    Posted Apr 3, 2008 at 5:39 AM | Permalink

    389 (Erl): Here you go:

    http://www.spacecenter.dk/publications/scientific-report-series/ISAC_Final_Report.pdf/view

    Looking again I see that the first ref was somewhat the same as this ref, but here’s another:

    http://arxiv.org/ftp/physics/papers/0411/0411002.pdf

  401. EW
    Posted Apr 3, 2008 at 5:58 AM | Permalink

    #399 (AlanB) Well, what an example of an abrupt solar change! ;-) But in fact it is not funny – more like an example of an Orwellian memory hole.

  402. Posted Apr 3, 2008 at 6:12 AM | Permalink

    399 (AlanB):

    Mike Lockwood from the UK’s Rutherford-Appleton Laboratory …
    showed that over the last 20 years, solar activity has been slowly declining…

    This is, of course, not Mike Lockwood’s discovery [although the quote suggests so], but rather his grudging admission that solar activity is not on an ever upwards path as his earlier work claimed.

  403. stargazer
    Posted Apr 3, 2008 at 6:15 AM | Permalink

    399(AlanB)

    Wow yes it indeed it does….

    “Mike Lockwood from the UK’s Rutherford-Appleton Laboratory.

    He showed that over the last 20 years, solar activity has been slowly declining, which should have led to a drop in global temperatures if the theory was correct.”

    So we have had a drop in global temp over the past 7-10 years according to the above this makes the theory correct… (just saying… grin)

  404. Andrew
    Posted Apr 3, 2008 at 6:24 AM | Permalink

    403 (stargazer)

    which should have led to a drop in global temperatures if the theory was correct.

    By this same token, obviously the rotation of the Earth doesn’t cuase the day/night temperature difference, since temperatures increase for a full two hours after the sun reaches its highest position in the sky!

    Its like “this is what happens when you let a biologist out of the lab” but more “this is what happens when a solar scientist dabbles in climatology”

  405. Posted Apr 3, 2008 at 6:29 AM | Permalink

    404 (Andrew):

    this is what happens when a solar scientist dabbles in climatology

    Hey! that would apply to me too :-)
    BTW, Mike is no solar physicist, but an atmospheric scientist, albeit the high atmosphere [ionosphere and beyond].

  406. Basil
    Posted Apr 3, 2008 at 7:48 AM | Permalink

    #396 Pete

    I would say this is significant when nit-picking over the Peaks of actual Solar Cycles and Warming Peaks. How can one do it without considering these internal short term Climate Changing events.

    Presumably, with a filter like Hodrick-Prescott, those events would be part of the cyclic component that is filtered out. When I can get to it, I plan to look more carefully at the cyclic component, and see if or how well it correlates to known shorter climate cycles. I think that will have to be a part of any eventual acceptance of HP as a useful tool for analysis of time serious climate data.

  407. Basil
    Posted Apr 3, 2008 at 7:59 AM | Permalink

    #385 Leif

    Only true and honest prediction rather than postdiction will count here. And, returning to your question, we have no understanding of how a Hale cycle period can happen in climate, so we can hardly make a compelling case, until we have such understanding or [as with rotations] we have accumulated enough cases that statistics alone can decide. IMHO, we are not there yet, as simple as that.

    Thank you.

  408. Andrew
    Posted Apr 3, 2008 at 9:19 AM | Permalink

    405 (Leif): Sorry, no offense was intended. It kind of seems odd, though, that an atmospheric scientist wouldn’t know about the basic effect I just described. And I’m not sure what difference the upper atmosphere makes here.

    400 (Me): I think I’m going to write to Professor Douglass and ask what effect using the ACRIM composite would have on his results (purely out of curiousity).

  409. Stephen Richards
    Posted Apr 3, 2008 at 9:34 AM | Permalink

    Mike Lockwood and a collegue produced probably the worst piece of so called scientific reasoning I have ever seen in a recent paper (last several months) about the sun activity and its effect on the climate.

    I cannot now bring myself to read any more of his junk. He should stick to his day job and hope he does that better than his climate rantings.

  410. Erl Happ
    Posted Apr 3, 2008 at 9:37 AM | Permalink

    384, 400 (Andrew) Thanks for that. I have read the Danish report but some time ago. Where in particular should I look for the solar cycle signal? It’s a big paper.

    I see the second paper finds a solar cycle response of about 0.10 K. This is not big biccies.

    I think the guys who did a long period statistical analysis found a bit more but it was also very small.

    I suggest that the La Nina at solar minimum will give this sort of response all on its own regardless of what is happening at solar maximum. Remember that ENSO temperature fluctuations swamp the measured solar cycle fluctuation in terms of magnitude.

  411. jae
    Posted Apr 3, 2008 at 9:58 AM | Permalink

    400(Andrew): Fig. 5 in your first reference is pretty impressive:

    Ocean Temperatures
    One example of a positive correlation is the apparent solar response of Sea Surface
    Temperatures (Reid, 1987; Reid, 1991; Reid, 2000). Sea Surface Temperatures (SSTs)
    have been obtained from ocean going ships since the middle of the 19th
    century. During
    the first part of the 20th century the observed SSTs increased, and then flattened out
    during the years 1940 and 1970, before continuing with the overall increasing trend.
    Figure 5 indicates that this long-term variability in SSTs is in phase with the 80-90 year
    envelope that modulates the approximately 11-year sunspot cycle

    It seems to me that the proper metric for looking at these relationships is SST. Air temperatures are too chaotic, quickly changing, and unreliable. I think we will eventually find that the sunspot cycle has a big influence on world climate. The “magnifying ” effect is probably cloudiness (cosmic rays? humidity?).

  412. Erl Happ
    Posted Apr 3, 2008 at 10:09 AM | Permalink

    383 (Leif)
    Please find the temperature data for the northern and the southern hemisphere and the globe as a whole, plot it against the red line of irradiance and see if I am correct or not in my statement that the temperature of the globe reaches a maximum in mid year when irradiance is least. Then perhaps you can judge whether my explanation in terms of cloud response to atmospheric heating from warm land masses in the northern hemisphere is reasonable or not. If it seems unreasonable check the cloud data in the Palle paper that I referenced to verify the point.

    Then perhaps, realising the importance of the cloud cover dynamic, you can re-assess your position.

  413. Posted Apr 3, 2008 at 10:13 AM | Permalink

    408 (Andrew): ACRIM. If the result depends on which composite you use, I for one would not accept it.

    409 (Stephen): I was not impressed either, although I know and respect the co-author [Claus Froehlich].

  414. Andrew
    Posted Apr 3, 2008 at 10:25 AM | Permalink

    413 (Leif): I don’t think it will make a huge difference. I’m just curious becuase with PMOD, solar is a net negative forcing and ACRIM net positive. What Douglass and Knox found was that the Tropospheric temperature is well predicted by solar, Nino, Volcanoes, and a linear trend. However, they found that the sensitivity value is very high for TSI, but low for volcanoes (which suggests to me that there are amplifiers at work). The result might change the linear trend value if reached with ACRIM, but I think you’d get the same basic thing: a solar cycle too strong to be explained by TSI without a large amount of positive feedback, which volcanoes contradict.

  415. Andrew
    Posted Apr 3, 2008 at 10:34 AM | Permalink

    410 (Erl): They discuss it in many places…the figures on page 54 to start.

    411 (jae) the problem is that it is only extended to 1980 (I think) and sunspot numbers at the end are about the same (se the figure) then as in 1980, I think. And sunspots have gone down since then. But, I agree that it is impressive. There must be some amplifier, in my opinion CRF->LACC.

  416. jae
    Posted Apr 3, 2008 at 11:02 AM | Permalink

    415, Andrew: Funny they didn’t update the figure. It looks like there might be a 3-5 year lag in SST, so the maximum effects of cycle 23 would not show until around 2006. And the SST should now be going down. By gosh, it is going down. :)

  417. Posted Apr 3, 2008 at 11:16 AM | Permalink

    412 (Erl): I have no position on how the hemispheres distribute the heat. There must be a zillion people on this blog who can confirm your ideas. All I rant against is the lining up of the tiny less than 1 W/m2 wiggles in TSI with ‘heating’ events in the tropics as long as they ride on an annual [not seasonal] variation a hundred times a strong.

  418. Posted Apr 3, 2008 at 11:25 AM | Permalink

    414 (Andrew): I think that ACRIM overshoots, PMOD undershoots, and deWitte’s (see #204) is just right. That would translate to a net solar forcing near zero if ACRIM > 0 and PMOD < 0, which is where it belongs [ :-) ].

  419. Andrew
    Posted Apr 3, 2008 at 12:47 PM | Permalink

    418 (Leif):Fair enough, but as the analysis stands using PMOD, it may produce an artificially high trend, if solar forcing is not net negative.

  420. oggy
    Posted Apr 3, 2008 at 3:12 PM | Permalink

    In all the above posts I don’t recollect any comments about the effects of solar flares and their links to El Nino events. Can anyone through any light or links on this?

  421. Andrew
    Posted Apr 3, 2008 at 3:15 PM | Permalink

    420 (oggy): I believe Erl has been going on about this, check his posts.

  422. Posted Apr 3, 2008 at 4:20 PM | Permalink

    412 (Erl): in email you comment on the 90 W/m2 wave:

    But the effect of the distribution of land and sea on atmospheric temperatures in the Northern Hemisphere and the consequent reduction in cloud cover swamps the stimulus entirely.

    If this effect can completely swamp a 90 W stimulus then it also swamps a 1 W wiggle on top of the 90 W wave, no? If not, how does the system know to swamp an amount of 45*cos(day-of-year/365*2pi) on a given day and not a milliwatt more [or less if there is a small dip in TSI due to a sunspot]?

  423. Posted Apr 3, 2008 at 4:33 PM | Permalink

    From here

    Research published today, Thursday 3 April, in the Institute of Physics’ Environmental Research Letters shows how a team from Lancaster and Durham Universities sought a means to prove the correlation between the ionizing cosmic rays and the production of low cloud cover.

    Previous research had shown a possible hint of such a correlation, using the results of the International Satellite Cloud Climatology Project, and this had been used to propose that global warming was all down to cosmic rays.

    The new research shows that change in cloud cover over the Earth does not correlate to changes in cosmic ray intensity. Neither does it show increases and decreases during the sporadic bursts and decreases in the cosmic ray intensity which occur regularly.

    One such very large burst caused the magnetic storm which blacked out the power in Quebec in 1989.

    Professors Sloan from Lancaster University and Wolfendale from Durham University write, “No evidence could be found of changes in the low cloud cover from known changes in the cosmic ray ionization rate.”

  424. Erl Happ
    Posted Apr 3, 2008 at 4:39 PM | Permalink

    417 (Leif)

    I have no position on how the hemispheres distribute the heat. There must be a zillion people on this blog who can confirm your ideas. All I rant against is the lining up of the tiny less than 1 W/m2 wiggles in TSI with ‘heating’ events in the tropics as long as they ride on an annual [not seasonal] variation a hundred times a strong.

    Less ranting and more ‘following where the data leads you’ will enable the question to be resolved.
    Let’s not overestimate the numbers that follow this thread, or the willingness to get involved in what can be a bruising experience.

    It’s not a question of ‘how the hemispheres distribute the heat’ but how the hemispheres respond to the irradiance change due to the orbital variation. One hemisphere amplifies the change in irradiance and the other accepts the energy into the oceans with little variation in atmospheric temperature.

  425. Posted Apr 3, 2008 at 4:56 PM | Permalink

    417 (Erl):

    Let’s not overestimate the numbers that follow this thread, or the willingness to get involved in what can be a bruising experience.

    Since you are so willing to undergo such an experience then we may perhaps appeal to the ‘numbers that follow this thread’ to signal if they understand why the precise orbital 90 W/m2 are swamped, but the irregular tiny 0.3 W/m2 [121-day means] are not at the same time and by the same process, and how the systems know to leave those tiny variations alone so that they cause heating events. I do not understand how this can happen, but am, as always, willing to be educated [which is always joyful and never bruising].

  426. jae
    Posted Apr 3, 2008 at 5:03 PM | Permalink

    423, Leif: Haven’t looked much yet, but a quick scan indicates they only look at solar cosmic rays, not GCRs?

  427. Posted Apr 3, 2008 at 5:18 PM | Permalink

    423 (jae): I have not read the paper yet, if you have, lead me to it, but since nobody [in the field] uses the term ‘solar cosmic rays’ anymore [they are now called 'solar energetic particles {SEPs}], I doubt the authors said or meant solar cosmic rays. The 1989 event was a solar explosion that did cause a significant decrease in GCRs, as can be directly seen in my plots in #11 and #264. So, I tentatively concluded they mean GCRs.

  428. kim
    Posted Apr 3, 2008 at 5:20 PM | Permalink

    New Idea: If solar cycles alternate pointed and rounded maxima, and the Pacific Decadal Oscillation alternates cooling and warming phases on a 30 some year cycle, then each flip of the oscillation has two solar cycles with two of one type of maxima and one of the other. I’ll leave it to my betters to pooh-pooh this, or suggest a mechanism for cooling or warming.
    ==============================================

  429. Andrew
    Posted Apr 3, 2008 at 5:23 PM | Permalink

    423 (Leif): see Lubos’s comments:

    http://motls.blogspot.com/2008/03/three-preprints-on-cosmoclimatology.html

    I’m not sure they’ve actually presented anything new, and I’m not terribly impressed by it. Incidentally, the statement “this had been used to propose that global warming was all down to cosmic rays.” is untrue, and a rather typical attempt to marginalize opposition to AGW.

  430. Andrew
    Posted Apr 3, 2008 at 5:31 PM | Permalink

    423 (Leif): I’m not sure if this refers to the same paper (it appears not) but this may not be the first time they have argued against the idea:

    The Cloud cover data over the US (Udelhofen & Cess, 2001) or the cloud data following the Chernobyl accident (Sloan & Wolfendale 2007) does not exhibit variations expected from the CRF/cloud-cover link. These expectations rest on the assumption that the CRF climate link should operate relatively uniformly over the globe. However, the lower troposphere over land is filled with naturally occurring CCNs, such as dust particles. Thus, one would expect the link to operate primarily in the clean marine environments.

  431. Andrew
    Posted Apr 3, 2008 at 5:42 PM | Permalink

    423 (Leif): I think that using their method, I could disprove a connection of anything to climate. In some places temperature will be anti-correlated with CO2, but it doesn’t mean it isn’t global important. I hope that better explains why I’m not impressed with their results.

  432. Posted Apr 3, 2008 at 5:42 PM | Permalink

    423 (jae): They mean GCRs. The “all down to cosmic rays” is just the journalist and not the paper. I found the paper [thanks to 429 Andrew] and it looks basically OK to me [not great, but then very few papers are great]. They confirm the ‘signal’ during SC22, but don’t see it in SC23 and the details go against the hypothesis. But I don’t see this paper having any impact on the die-hard GCR freaks. So, maybe I should not have mentioned it, after all, but it is a contribution.

  433. Not sure
    Posted Apr 3, 2008 at 5:53 PM | Permalink

    (423)Leif

    I just skimmed it, and it looks like they’ve found (again) that there is no correlation between climate and cosmic ray flux as detected by Climax and various neutron detectors. Nir Shaviv has said again and again that such findings are irrelevant because those instruments detect cosmic rays whose energy is too low to affect cloud formation in the troposphere.

    The full paper is here

  434. Posted Apr 3, 2008 at 5:53 PM | Permalink

    428 (kim): I don’t think you will get any takers, either way. Now, tell me if you understand Erl’s swamping effect?

  435. Posted Apr 3, 2008 at 6:05 PM | Permalink

    433 (Not…): And Svensmark & Co. have again and again showed that the correlation is with the neutron monitor data from Climax and other stations, the energy range that Shaviv says again and again is irrelevant. If the proponents can’t get their story straight, I’m not buying any of it. You can’t have it both ways.

  436. kim
    Posted Apr 3, 2008 at 6:12 PM | Permalink

    434 (JLeif) Oh, heck, I was proud of that one. The best I can do with the swamping is that the 90 watts is repetitive every year, and the 1 watt is not and is (magically) amplified.
    ===========================================

  437. kim
    Posted Apr 3, 2008 at 6:15 PM | Permalink

    436 (kim) Also, I wonder if the correlation noted between cosmic rays and clouds in one cycle, but not in the next, also alternates repetitively.
    ======================

  438. Not sure
    Posted Apr 3, 2008 at 6:18 PM | Permalink

    (435)Leif I don’t see how it follows that Shaviv must be wrong because “Svensmark & Co.” are wrong.

  439. Andrew
    Posted Apr 3, 2008 at 6:24 PM | Permalink

    435 (Leif):The problem is that the energies Nir is looking at do not have data over the whole period for which we have cloud data, so no comparison can be done. Rather unfortunate, I think, as it might answer some important questions. I’ll try not to be offended by the “freaks” remark. ;)

  440. Posted Apr 3, 2008 at 6:31 PM | Permalink

    438 (Not…): but it follows that they cannot both be right. I have seen statements such as: “there is evidence that cosmic rays drives the climate [Svensmark, xxxx; Shaviv, zzzz]” where they supposedly support each other. My personal opinion is that [once again] if the proponents cannot agree, there is no need for me to take any of them seriously. This may be a bit harsh, but such is the house that science builds: one result builds on and confirms earlier results.

  441. Posted Apr 3, 2008 at 6:36 PM | Permalink

    439 (Andrew): it takes one to know one :-) Well, I thought that if a remark is outrageous enough it is not offensive. The way I read Nir is that Svensmark’s rays are irrelevant, then Svensmark ‘result’ is spurious. If not spurious, then the rays are not irrelevant. Am I wrong?

  442. Andrew
    Posted Apr 3, 2008 at 7:02 PM | Permalink

    441 (Leif): Well, what I was getting at is that I see no reason why cosmic rays of one set of energies would be correlated and not another. So, if they found that there is a solar cycle in LACC, why would it not work equally well with a different energy set? As I said, the ion chamber data doesn’t cover the whole period of cloud data. The Danish team is fully aware of that data (see figure 2), so I’m speculating as to the reason they don’t compare it to clouds. I believe that Nir is saying we can establish that clouds appear to be correlated with cosmic rays with any solar index, but that this is what is relevant for climate.

  443. Posted Apr 3, 2008 at 7:23 PM | Permalink

    442 (Andrew): I’m missing something here [first a confession: I don't read their papers carefully because on their face they don't inspire confidence]. I thought it was the Danish team that claimed there was a connection with clouds… using neutron monitor data. And I thought you and Shaviv were saying that that data and the correlation were irrelevant. Who cares what the ion-chamber results show, that is not real climate data [yet]?

  444. jae
    Posted Apr 3, 2008 at 8:25 PM | Permalink

    Leif, 432:

    But I don’t see this paper having any impact on the die-hard GCR freaks.

    LOL, you are starting to sound like Judith Curry. Freaks?

  445. jae
    Posted Apr 3, 2008 at 8:28 PM | Permalink

    433, Not Sure: That’s what I was trying to get at.

  446. jae
    Posted Apr 3, 2008 at 8:30 PM | Permalink

    435, Leif:

    And Svensmark & Co. have again and again showed that the correlation is with the neutron monitor data from Climax and other stations, the energy range that Shaviv says again and again is irrelevant. If the proponents can’t get their story straight, I’m not buying any of it. You can’t have it both ways.

    I don’t understand what you are saying here. Could you flesh it out a little?

  447. jae
    Posted Apr 3, 2008 at 8:34 PM | Permalink

    440, Leif:

    My personal opinion is that [once again] if the proponents cannot agree, there is no need for me to take any of them seriously. This may be a bit harsh, but such is the house that science builds: one result builds on and confirms earlier results.

    Are we to believe that only you know the true facts about solar science? Sorry to be so blunt, but you are coming across this way.

  448. Mike Davis
    Posted Apr 3, 2008 at 8:56 PM | Permalink

    447(jae):
    I would agree with Leif on this. If everyone wants to promote their own personal theory and spend their time showing why everyone elses theory is wrong no advances will be achieved.
    I think the the answers to the solar climate connection lie in the future with more and better observations.
    I also think the the answers to the climate change question will require more and better observations.
    Nature will prove the theorys right or wrong in time. It just seems to have a way of doing that.
    Mankind needs to learn to have patience as we can not know the future (when we think we know what will happen nature throws a curve ball).

  449. Dennis Wingo
    Posted Apr 3, 2008 at 9:47 PM | Permalink

    398 which charts?

    Most Recent Charts (your first link).

    Thanks

  450. Posted Apr 4, 2008 at 2:09 AM | Permalink

    445,446 (jae): Svensmark says that there is a strong correlation between CGRs of energy X and clouds. Shaviv says that energy X is too low to affect cloud formation in the troposphere. Hence, Shaviv must discount the correlation as spurious, and has his own set of [different] observational evidence. I, the fence sitter, am not ready to fall into either ditch. I’ll stay on the fence. And I said that that was my own personal opinion, and did not imply that I know better than either of the proponents; rather that I know less, and therefore cannot throw my hat in with either.
    444 (jae): my attempt at levity at least did ferret out a LOL from you. Mission accomplished :-)

    449 (Dennis): update in progress. problem is that the sun changes so slowly that it is hard to see any change. The Ulysses people have not released any new data, so that section must wait a bit. Which of the Figures are of special interest to you?

  451. Allan MacRae
    Posted Apr 4, 2008 at 4:30 AM | Permalink

    450 etc (Leif and all)

    Nir Shaviv comments on the above topic at

    http://www.sciencebits.com/RealClimateSlurs

    See also Figure 2 in Veizer (2005 – GAC), which is derived from Svensmark and Marsh (2003 and 2005) – the GCR:cloud correlation looks pretty good – I’ll email you the paper if you want.

    Best, Allan

  452. Posted Apr 4, 2008 at 6:12 AM | Permalink

    451 (Alan): So Shaviv does not dispute the GCR-climate correlation. Nowhere did I see that he said that the energy was too low as maintained by “NotSure” in #433. So there is no problem. Should teach me not to react to something without double checking. But, me point is still valid, if proponents of something disagree, I count them both out.

  453. jae
    Posted Apr 4, 2008 at 9:00 AM | Permalink

    Leif: FWIW, I’m on the fence, too. I keep hoping that someone will clear this all up and produce a well-supported theory. Sorry about my impatience. :)

  454. Andrew
    Posted Apr 4, 2008 at 9:20 AM | Permalink

    452 (Leif): I’ve just been saying that, for comparing to clouds, it doesn’t matter which data set you use, so they just used the data that covered the whole period. You could compare it to TSI even, becuase we only have cloud data over the last two solar cycles. However, some have used the Nuetron monitor data to argue that the correlation means that clouds haven’t changed since 1950. Here, and at no time before, does the set of energies you use become meaningful, becuase the long term changes will only be captured by those energies which actually matter. I know its a bit confusing, but the idea is that you can use the Nuetron monitor data to show a correlation, but it is not physically meaningful. However, the Nuetron monitor correlation should be extendable to any data with a solar cycle, and the ion data have a solar cycle. If the ion chamber data are the relevant energies, then this correlation is what is physically meaningful.

    In short, I see no contradiction, clouds are equally correlated with both. The reason for using Nuetron monitors is that ion chambers stopped counting in 1994 or so for some reason, but we want to show that clouds are still correlated with CRF.

  455. Posted Apr 4, 2008 at 10:04 AM | Permalink

    454 (Andrew): I didn’t initially see any contradictions, but then out of the blue came those “yeah, but Shaviv says that the neutron data are irrelevant”. Of course, since all cosmic ray series follow the sunspot cycle, any index that correlates with that can be used for the correlations.
    Now, while we are on the ion chamber data: the ion data goes back to the 1930s. The problem with the ion data is that they are ‘relative’ counters only and do not have a stable calibration, which neutron monitors do. So, can we put this subject to rest: it is OK to use neutron data for GCR correlations with clouds? What I see coming is that when [if?] the correlations begin to fail, that people will say again “yeah, but the neutron data is no good”.

    BTW, McCracken [2007; see links on my website, especially page 15 of this one] claims that there is a significant [as large as the solar cycle variation itself] step-like change in ion chamber counts around 1944. If this is real, and the ions are the ‘physically relevant’ that would seem to have implications for a climate link.

    To be credible, a correlation or mechanism should not be a moving target. And people should begin to converge on something that can be built upon. Tempered, perhaps, by Napoleon’s adage: “truth is a bunch of lies all agreed upon”.

  456. Andrew
    Posted Apr 4, 2008 at 10:27 AM | Permalink

    455 (Leif): Neutron monitor data is okay to show cloud correlations. You bring up an interesting point that I was not aware of with the ion chamber data, which may be problematic for the apparent long term trend (which is essentially what the bickering over ion->neutron is all about) I don’t know enough about the issue, but for me it is not crucial that there be any long term trend. However, since there may/may not be one, it is to early to say, as RC does “no trend since 1950″.

  457. Posted Apr 4, 2008 at 11:12 AM | Permalink

    456 (Andrew): the ‘jump’ in 1944 in the ion count is an error in my opinion [see here]. The significance is that we do not even know accurately what the trend [if any] is, making it hard to compare with climate trends.

  458. Dennis Wingo
    Posted Apr 4, 2008 at 11:33 AM | Permalink

    (450) Solar cycle progression.

  459. Erl Happ
    Posted Apr 4, 2008 at 12:47 PM | Permalink

    356 (Leif)

    NCDC’s Revised Monthly Averages 1961-1990 (°C) (Surface temperatures)
    Source: http://www.junkscience.com/MSU_Temps/Warming_Look.html

    Jan 12.0, Feb 12.1, Mar12.7, Apr 13.7, May 14.8, Jun 15.5, Jul 15.8, Aug 15.6, Sep 15.0, Oct 14.0, Nov 12.9, Dec 12.2.

    Despite the 9% lower irradiance in mid year due to the orbital variation, the Earth is warmer at that time. This is undoubtedly due to a regular reduction in cloud cover (about 3% documented by satellite observation) that is due to the heat radiating from the land masses of the Earth that are predominantly located in the northern hemisphere.

    Could there be any better illustration of the effect of changing cloud cover on surface temperature?

    Is it not conceivable that tiny variations in irradiance from the sun (over periods that can be greater than a year) that clearly align with variations at the surface (and in the troposphere) owe the amplified response at the surface (and a more amplified response in the troposphere) to the same factor that produces a warmer Earth in July than in January.

    Is it not reasonable to suggest that the greater variation that is seen in the troposphere is due to interception of some elements of that irradiance by the atmosphere itself.

    Is it not plain that a warming atmosphere will (initially at least) be less cloudy.

    NASA has documented an increase in humidity in the troposphere and in the stratosphere over the period of the satellite record. This affects stratospheric aerosol and ozone content with important consequences for the rate of transmission of solar radiation into and through the troposphere. This contributes to the warming trend at the surface.

    A dominance of warming events over recent solar cycles is likely initiating the surface warming and producing the atmospheric humidification that has occurred since 1976. These warming events have their origin in small changes in irradiance (and other factors that appear to produce atmospheric warming when changes in irradiance are plainly not responsible). It is very likely that these unknowns are also related to aspects of solar activity.

    The Earth is a radiating body with very efficient mechanisms for heat dissipation. ‘Internal oscillations of the climate system’ and the concept of ‘chaos in the climate system’ or ‘noise in the signal’ are all BS. These concepts have their origin in the chaos and confusion in the minds of their proponents.

  460. Erl Happ
    Posted Apr 4, 2008 at 1:20 PM | Permalink

    346 (Leif)

    I’m talking about the huge effect of 90 Watt/m2 more input during January than during July, to BOTH hemispheres. [This is now the third time I say this]. It is precisely that large difference that is the reason we are not in the grip of an ice age right now.

    I wonder about this ice age statement: If there was more irradiance in July than in January a lot of the heat that is immediately returned as Outgoing Long Wave Radiation from heating land masses would be speedily lost. Because the currently enhanced radiation benefiting the Southern hemisphere that is probably well absorbed into the Southern Hemisphere ocean (by virtue of the very large cloud free areas in the tropics that manifests in neutral to El Nino times) it is geo-effective and keeps the Earth warmer.

    An ice age would be promoted if the situation were reversed. It only takes a little bit of cooling for the massive land area of the Northern Hemisphere to acquire more highly reflective snow in winter. It will be interesting to observe the consequences for ‘length of frost free growing season’ of the current La Nina. La Nina brings a strong collapse in water temperatures in the southern hemisphere. There are strong knock on effects in terms of the chill experienced in the Northern Hemisphere winter as demonstrated in 2007-8.

    I see water temperatures falling strongly in the southern hemisphere currently and suspect that there will be a strengthening of the trend for Equatorial Pacific chilling by August 2008. I suspect that this La Nina will continue until we see a rise in solar activity.

  461. Andrew
    Posted Apr 4, 2008 at 1:39 PM | Permalink

    459 (Erl):

    ‘Internal oscillations of the climate system’ and the concept of ‘chaos in the climate system’ or ‘noise in the signal’ are all BS. These concepts have their origin in the chaos and confusion in the minds of their proponents.

    Are you prepared to defend this idea in the face of incontrovertible evidence to the contrary?

    457 (Leif): Thanks, I’ll look this over further later. Many of these issues are beyond me. I frequently, as a layman, just try to interpret the data, not deeply analyze the relative merits and demerits of them, and I’m certainly not very familiar with the involved issues. I hope this issues gets resolved so that I don’t have to learn much more solar physics etc. than I really want to…

  462. Posted Apr 4, 2008 at 1:43 PM | Permalink

    459 (Erl):

    Is it not conceivable that tiny variations in irradiance from the sun (over periods that can be greater than a year) that clearly align with variations at the surface (and in the troposphere) owe the amplified response at the surface (and a more amplified response in the troposphere) to the same factor that produces a warmer Earth in July than in January.

    Yes, it is very conceivable, and one can calculate how much. If we assume that the 4 degree July-January difference is caused by the different hemispheric response to 90 W/m2, then I calculate that the 0.2 W/m2 bump in TSI circled in #383 results in a heating of 0.009 degrees by the same process; you are quite right, after all.

  463. Posted Apr 4, 2008 at 1:51 PM | Permalink

    460 (Erl): you said:

    more irradiance in July than in January [...] keeps the Earth warmer. [...] An ice age would be promoted if the situation were reversed.

    Situation reversed? Meaning more irradiance in January? Well, that is what we have now and we are between ice ages. I think you got your argument backwards.

  464. Andrew
    Posted Apr 4, 2008 at 2:27 PM | Permalink

    I recall seeing (I don’t know where) a graph of Aurora observations, which appeared to show more Auroras recently, and also a high during the MWP, but not as high as the present. Does anyone know anything about this?

  465. Erl Happ
    Posted Apr 4, 2008 at 2:41 PM | Permalink

    400 (Andrew)
    From page 54 of the Danish report

    in Figure 30, we show the correlation between global tropospheric temperature anomalies and EUV radiation and galactic cosmic rays. In the top panels are shown the raw data giving correlations of 0.33 with EUV and –0.31 with GCR.

    I’ll back the EUV as the best indicator but as seen in #308 the relationship with solar irradiance looks excellent. The variable amplifier that mucks up the correlation with EUV is undoubtedly cloud cover and the mechanism is likely a very simple one, namely direct heating of the atmosphere.

    GCR’s vary with the solar cycle. Temperatures on Earth vary very weakly with the solar cycle and very strongly with ENSO and solar irradiance.

    I go with the KISS principle.

  466. Andrew
    Posted Apr 4, 2008 at 2:53 PM | Permalink

    465 (Erl): If you read the report carefully, correlation with CRF are usually the only ones that retain significance when the eleven year cycle is removed, implying that they are the strongest solar signal, to me at least. However, if you choose to believe that direct heating by solar cuases ENSO events to warm the Earth, go ahead. I just don’t believe that the small energy input is sufficient.

  467. Erl Happ
    Posted Apr 4, 2008 at 2:54 PM | Permalink

    483 (Leif)
    Sorry if my prose is obscure

    I said ‘If there was more irradiance in July (hypothetical) than in January (current state of play)’.

    Previous post I said ‘Despite the 9% lower irradiance in mid year due to the orbital variation, the Earth is warmer at that time.’

  468. Erl Happ
    Posted Apr 4, 2008 at 3:02 PM | Permalink

    466 Andrew
    I acknowledge your point. That is what the report says. However, I am mindful of the old adage that there are lies, damned lies and statistics.

    ENSO and temperature change are synonymous. When you take out ENSO what are you left with?

  469. Andrew
    Posted Apr 4, 2008 at 3:17 PM | Permalink

    468 (Erl): in the papers I qouted above, the ENSO signals were removed, and there were lots of other signals left, including a solar cycle signal. But of course, you are entitled to your view of statistics (though, I must say, that is a strange view to have on website about statistics) used to get these signals. BTW the multivariate ENSO index is based on six main observed variables over the tropical Pacific: sea-level pressure (P), zonal (U) and meridional (V) components of the surface wind, sea surface temperature (S), surface air temperature (A), and total cloudiness fraction of the sky (C). (Wolter, 1987, Wolter and Timlin, 1993).

  470. Posted Apr 4, 2008 at 3:19 PM | Permalink

    467 (Erl): I said: “It is precisely that large difference that is the reason we are not in the grip of an ice age right now.”

    and you continued:
    “I wonder about this ice age statement: If there was more irradiance in July than in January”

    as if you somehow disagreed with my statement, but evidently you did not. Still leaves the effect of the tiny TSI wiggles at the 0.01 degree or below.

  471. Posted Apr 4, 2008 at 3:32 PM | Permalink

    469 (Andrew): Erl is falsification-proof (no need for statistics), to wit: this quote from #459: “These warming events have their origin in small changes in irradiance (and other factors that appear to produce atmospheric warming when changes in irradiance are plainly not responsible)”, so no matter what, he is covered. When TSI wiggles don’t do it, there are other factors conveniently appearing as needed to make up for that. How can one possibly counter that? so you can see why no statistics is needed. It also means that no amount of discourse is going to make any difference, so perhaps, Erl, it is time voluntarily to help conserve bandwidth (especially the human variety).

  472. Mike Davis
    Posted Apr 4, 2008 at 6:39 PM | Permalink

    463(leif): Sorry to correct you but Earth has been experiencing an Ice Age for at least 2 million years and we are currently experiencing an “Interglacial Period”. Which started with the Halocine optimum and sometime in the near future (geologic time wise) we will be sliding back into a longer period of glaciation. Statistically we are lucky to be alive during this period and experiencing the warmth that we currently have. As far as I know the “Ice Age will not end until the glaciers are gone. That includes Antartica and Greenland Ice Sheets. That would take a lot of extra warmth and quite a number of years according to what I have read.
    Of course that is just my opinion based on things I have read as I have always found history and geology interesting subjects.

  473. maksimovich
    Posted Apr 4, 2008 at 8:06 PM | Permalink

    re 469(Andrew):

    in the papers I qouted above, the ENSO signals were removed, and there were lots of other signals left, including a solar cycle signal.

    Read what Ruzmaiken says above in 153 especially the last paragraph.

    Here is another paper of interest by AR.

    Abstract. ENS0 (El Nino and the Southern Oscillation) is considered a stochastic driver that excites the atmospheric anomaly states, such as the Pacific North American pattern. This can make the 11- year solar activity forcing feasible to climate through stochastic resonance — a phenomenon that amplifies a weak input to a nonlinear bistable system by the assistance of noise.

    In far from equilibrium systems such as the atmosphere with complex dynamics there are a number of constraints in over simplification.

    1) Prediction and determinism are incompatible: we cannot predict long-term behaviour of complex systems, even if know their precise mathematical description.
    2) Reducing does not simplify, interaction is important and interaction means inseparability.
    3) Simple linear causality does not apply to Chaos and Complexity.
    4) Complex dynamics give birth to forces of self-organisation(auto catalysis).

    The latter being of the upmost importance.Self-organization behaviour can be exhibited by far-from-equilibrium chemical systems as it was shown by the Nobel-prize winner (1977) in chemistry Ilya Prigogine. According to the results of his studies, the inorganic chemical systems can exist in highly non-equilibrium conditions impregnated with a potential for emergence of self-organizing chemical structures. The more complex the aggregation of these structures, the stronger the tendency for macro-molecules to organize themselves.

    Haigh (1999)also makes similar observations.

    Abstract

    Computer simulations of the impact on climate of solar variability generally fall into four categories. First, there are lower atmosphere GCM experiments, in which enhanced solar activity is represented by changes in spectrally integrated solar constant. Secondly, there are GCM studies of the dynamical response of the middle atmosphere to changes in solar ultraviolet, mainly concentrating on the northern hemisphere winter, and how these impact the troposphere. These studies have been instructive in providing an understanding of some of the mechanisms involved but, because of the very different nature of the assumptions made, give rather different suggestions as to potential patterns of change. In particular predicted zonal mean temperature changes in the lower stratosphere are usually of opposite sign in these two types of experiment. None of these GCM studies include interactive photochemistry and the third category of modelling work is concerned with the photochemical response of the middle atmosphere to enhanced solar ultraviolet. These generally employ 2D models to predict changes in ozone and other gaseous species. Recently it has been realised that the responses (to a variety of external forcings) of the lower and middle atmospheres are linked through both radiative and dynamical mechanisms and should not be viewed in isolation from each other. Thus the fourth type of modelling study, which is still in its infancy, attempts to represent solar variability by realistic changes in both irradiance and ozone concentrations. In this paper these various modelling studies are reviewed and some new results presented which confirm previous theoretical suggestions that, in the northern hemisphere winter, the atmosphere may respond to solar changes in a similar way as to the injection of volcanic aerosol. The implications of the results of the model studies for the detection of solar-induced climate change are discussed.

  474. Posted Apr 4, 2008 at 8:17 PM | Permalink

    472 (Mike): You are, of course, correct. I should, perhaps, have said ‘ice episode’. But my meaning should have been clear, regardless.

    473 (maks…): As is evident, there is a yearning for the simple stuff, more simplicity than Nature can live up to. Real life is messy, complicated, involved, chaotic, non-linear, …

  475. maksimovich
    Posted Apr 4, 2008 at 8:28 PM | Permalink

    Re 474(Leif)

    I think Alvin Toffler summed it up as impatience.The TV age says the detective must solve the complex problem,or the hero defeat the enemy and win the beautiful woman in 60 minutes including ad breaks.

    Science must take a little longer.

  476. Erl Happ
    Posted Apr 4, 2008 at 10:09 PM | Permalink

    475 ((Maksimovich ) Impatience: Here is an excellent summary of the social, political and environmental consequences of rampant environmentalism from a person who experienced at first hand the consequences of a tyranny based on the wisdom of a messianic minority. http://www.nzcpr.com/guest88.htm

    Leif, are you suggesting that I should be snipped? Or that I should shrivel.

    470 (Leif) Sorry, I did misinterpret your statement. An irradiance maximum in January serves to incorporate more enduring warmth into the Earth system than a maximum in July would. However, I am yet to hear you acknowledge that it is due to absorption into the oceans of the southern hemisphere yielding little advance in atmospheric temperature in January.

    Puzzles me that you can not concede that the irradiance minimum in July is turned into a temperature maximum by the Earth system itself, due to a reduction in cloud cover in mid year. Or have you actually said that? Is there another reason that you have in mind? Are you suggesting that some of the other temperature peaks in the troposphere are also related to the irradiance maximum in January as you suggested in the following words in #383 , or is it just this one.

    If anything one might argue that the temp curve matches the overall red curve rather well in the sense that there is 90 W/m2 extra in January 2005, matching the similar peak in the temps. I would actually expect some kind of match-up like that, as it seems to me that the atmosphere cannot possibly be so finely constructed as to always perfectly canceling out the 90 W/m2 variation.

    But, it in fact does just that.

    Puzzles me that you can not acknowledge that changing cloud cover is an amplifier of changes in irradiance to produce amplified changes in temperature in the troposphere. I guess what we have learned from this discussion is that this amplification is always likely to be greater in the northern than the southern hemisphere.

    If you need statistics the Danes did find that temperatures in the troposphere correlate with EUV (#465).

  477. Andrew
    Posted Apr 4, 2008 at 10:12 PM | Permalink

    472 (Mike): Are you familiar with anthropic reasoning? We live in an interglacial becuase we would be here to notice if we didn’t ;)

    473 (maksimovich): I’ll look into it but (an this will get on Erl’s nerves) I still see ENSO as internal and unrelated to solar activity (though I am still curious about you know who’s ENSO predictions, which were quite good)

  478. Erl Happ
    Posted Apr 4, 2008 at 10:36 PM | Permalink

    For those with a head for statistics and a curiosity as to the importance of ENSO in climate change go to http://rankexploits.com/musings/2008/accounting-for-enso-cochrane-orcutt/ where you will find this:

    Accounting for the effect of ENSO by introducing the MEI index into the regression for temperature, the best estimate for trend of m= -0.5 C/century.

    Yes, that’s minus 0.5°C per century.

    Fortunately, some aspects of scientific endeavour are speedy and some people are outrageous enough to question the conventional wisdom.

  479. Posted Apr 5, 2008 at 3:59 AM | Permalink

    476 (Erl): No, I’m not suggesting that you be snipped, I’m suggesting that you are too repetitive without coming to the core of the matter, which is that if 90 W/m2 induces temperature changes [by whatever mechanism] of 4 degrees, then 0.2 W/m2 changes the temperature by 0.01 degrees. And, that whatever might amplify the latter, in equal measure amplifies the former.

  480. Andrew
    Posted Apr 5, 2008 at 4:13 AM | Permalink

    479 (Leif) I’m not sure whether the time scales are comparable in this way (although what I mean by that offer Erl no help). The response to brief burst may be smaller. On the other hand, since the rapid 90 W/m2 swings never allow the planet to reach equilibrium, I would advise against applying this to long term solar trends (not that I see you doing that here). By the IPCC’s estimate of climate sentivity of about 3 C for doubling CO2, I calculate that 90 W/m2 radiative foring translates to an eventual rise in temperature in excess of 70 degrees.

    These papers are interesting:

    http://arxiv.org/pdf/0803.2765

    http://arxiv.org/pdf/0803.2766

  481. Posted Apr 5, 2008 at 4:44 AM | Permalink

    480 (Andrew): I don’t see what CO2 and IPCC have to do with the response to a sustained 50 W/m2 forcing. The correct response is about 5 degrees. Here is a little table to put all of this into perspective:
    <pre>
    50 W/m2 -> 2.5 degree
    5 W/m2 -> 0.25 d ; this was what the Maunder Minimum was supposed to give, c.f Hoyt and Schatten and early Lean.
    1 W/m2 -> 0.05 d ; solar cycle variation 0.2 W/m2 -> 0.01 d ; Erl’s little wiggles
    </pre>

    All of these are for sustained forcing. For shorter time scales, the response [as you point out] is smaller, to non-existent [if short enough].

    If I read Erl correctly [and that is hard because although voluminous the information content is low] there is a kind of Maxwell demon [which I'll call Erl's demon], that sits at the top of the atmosphere and looks at each photon: “ah, this one is one of the 90 W/m2 crowd, leave him alone, but that one, lo and behold, belongs to one of the wiggles, be sure to amplify him”.

  482. Posted Apr 5, 2008 at 5:00 AM | Permalink

    481 (me): the “to a sustained 50 W/m2 forcing” should have been 90. of course. And my attempt to use the ‘code’ tag failed. May I suggest that the blog software doesn’t work with this? or does someone know how to? [Steve?]

    I’ll try another way:
       50 W/m2 -> 2.5 degree
        5 W/m2 -> 0.25 d ; this was what the Maunder Minimum was supposed
                                   to give, c.f Hoyt and Schatten and early Lean.
        1 W/m2 -> 0.05 d ; solar cycle variation
      0.2 W/m2 -> 0.01 d ; Erl’s little wiggles

  483. Posted Apr 5, 2008 at 5:33 AM | Permalink


     50 W/m2 -> 2.5 degree
      5 W/m2 -> 0.25 d ; this was what the Maunder Minimum was supposed
                         to give, c.f Hoyt and Schatten and early Lean.
      1 W/m2 -> 0.05 d ; solar cycle variation
    0.2 W/m2 -> 0.01 d ; Erl's little wiggles

    success…

    <table>
    <tr>
    <td>1
    <td>2
    <td>3
    </tr>
    </table>

  484. Posted Apr 5, 2008 at 5:43 AM | Permalink

    Steve, can you also snip the table stuff at the bottom of 483, thanks.

  485. cba
    Posted Apr 5, 2008 at 7:40 AM | Permalink

    481 (Leif):

    Leif, where is the 4 deg. / forcing / deltaT coming from? Your post on early Lean, H&S etc. suggests .05 deg / w/m^2 but I must’ve missed this latest goodie. Also, is there a measured T observation that is indicating 4 Deg between Jan and Jul? My own little OLR – albedo – ACRIM chart suggests there’s maybe 15 w/m^2 difference between jul and jan – except since 2002 or 2004 when it dropped to a difference of half that much roughly. In jan there is a net energy absorption and in jul a net energy emission over 33 yrs. Also, the albedo is just an averaged number based on the jan and jul measured albedo only over 3 or 5 years since there apparently were no sat. measurements prior.

    The little wiggle at the end of the chart is somewhat interesting as the jan energy balance goes up (- means energy absorption, + means more energy is radiated) significantly but not to break even and the jul balance drops, but not down to break even. Both are the largest changes on each respective graph line and both seem to happen at the same time- around 2002 or 2004. The rest is just bouncing around several W/m^2 – which suggests that there could be several W/m^2 variations that last for years which are just rather random in appearance. Unfortunately, this two sample months in a year doesn’t guarantee there is that much variation actually going on over the entire year though rather than the shifting from one month to the next.

  486. Posted Apr 5, 2008 at 9:18 AM | Permalink

    485 (cba): when you plot TSI as a function of time it should be the real observed TSI at Earth if you are discussing the effect of TSI on the Earth. That TSI has a 90 W/m2 annual modulation, see comment #383. If you want to see what the Sun is doing you plot TSI reduced to a constant distance [1 AU]. Almost every plot of TSI out there is of the latter variety. It is just amazing how many people just don’t [or won't] get this. You tell them a zillion times and it has absolutely no effect :-)

    There are two ways to get to the numbers I gave. The equilibrium temperature of the Earth be be estimated by equating incoming and outgoing energy. The incoming is Fin = S/4 (1-a) where S is the TSI and a is the albedo. For an average black body the long-wave emitted flux is Fout = s T^4, where T is the temperature and s is 5.7*10^(-8) W/m2 per degree Kelvin^4, the Stefan-Boltzmann constant. Equating Fin and Fout, you get: T = ((1-a)S/(4s))^(1/4). With a=0.30, you find that T varies 0.25K for a change of S of 5 W/m2. [Hint: dT/T versus dS/S]. The same dT you would get is ‘a’ varied by 0.0025. Then my little table follows, and the 90 W/m2 then means a 4.5K dT. All this must be [should have been] old hat to 99% of everybody here.

    The other way is to forestall any notion of “yeah, but the climate systems don’t vary according to your simple-minded physics”. To do this I used the table given by Erl in #459, that showed a 3.8K [rounded up to 4K] difference between January and July of global temperature. Now, according to Erl, the clouds somehow shifts the 90 /w/m2 hump in T from January to July to give this 4K response. Setting dT/T = Q dS/S, where Q represents this unknown process, we find 4/290 = Q 90/1365, hence Q = 0.2, and dT = Q * dS/S * T = 0.2 * 0.2/1365 * 290 = 0.0085K for 0.2 W/m2 dS. So, if some process shifts the hump from Jan to Jul, that SAME process [with same Q] would have a response of less than 0.01K to a variation of 0.2 W/m2 on the same time scale as the annual wave.

    The last half of your posting seems irrelevant in view of the above explanation.

  487. Posted Apr 5, 2008 at 9:22 AM | Permalink

    486 (me): “The same dT you would get if ‘a’ varied by 0.0025″, for the nit-pickers.

  488. Erl Happ
    Posted Apr 5, 2008 at 9:35 AM | Permalink

    485 (cba) Re 4°C between January and July

    See #458. Actually it’s 3.8°C increase in temperature for a 6.4% drop in irradiance(90 Watts per metre compared to 1410 as shown in #383). How’s that for amplification?

  489. Posted Apr 5, 2008 at 9:57 AM | Permalink

    488 (Erl): if 90 W/m2 gives a dT of 3.8K, then a 0.2 W/m2 which is 450 times smaller would give a dT of 3.8/450 = 0.008K, no?

  490. Erl Happ
    Posted Apr 5, 2008 at 10:32 AM | Permalink

    489 (Leif) Emphatically no.

    For a start let’s acknowledge that radiation falls between January and July and temperature goes up. Why?

  491. Posted Apr 5, 2008 at 11:02 AM | Permalink

    490 (Erl): Let’s try another tack: the global temperature would go up even if the Earth’s orbit was circular, because the Northern Hemisphere heats up [for obvious reasons, even if there were no clouds at all] more during its summer than the Southern. The higher TSI during Northern winter helps [by adding 2.5K] to ensure that the Northern Hemisphere doesn’t cool off enough to restart a glaciation. At any time, a 0.2 W/m2 bump in TSI will cause 0.008K heating overall, completely insignificant, unless Erl’s Demon does its magic trick and singles those photons out for special, highly non-linear, un-calculatory, treatment.

  492. kim
    Posted Apr 5, 2008 at 11:10 AM | Permalink

    It’s not a daemon, it’s a genie in a bottle, the bottle being the ocean and the genie being the vaporous mists arising from it. And yes, it is a powerful genie. A small variable on what is functionally a constant(the 90 watts) can have strong additive effects.
    ===================================

  493. Posted Apr 5, 2008 at 11:38 AM | Permalink

    492 (kim): The 90W is not a constant, it is the total swing in 182 days, so the TSI changes continuously from day to day by on the average 90/182 = 0.5W per day. The 0.2W in question [the 121-day running mean] is a change or 0.2/60 = 0.003W per day. The genie is indeed powerful, but it is hardly science.

  494. kim
    Posted Apr 5, 2008 at 12:08 PM | Permalink

    Yes, but it is applied constantly, year after year after year. The variable amount, though small, is added, or subtracted over a period of years.
    =================================

  495. Posted Apr 5, 2008 at 12:20 PM | Permalink

    494 (kim): it varies from day to day, added 0.5W per day for 183 days, then subtracted 0.5W per day for the next six months, so is [almost] as variable as the 0.2W smoothed variation on #383. In any case, a sustained input [over months] is more likely to have any effect than a rapidly varying component that varies from day to day, pushing one way today [or this week] and the other way tomorrow [or next week].
    And you are ignoring the energetics of it. 0.2W is very tiny, and we know of no mechanism [show me one if you know] that will amplify that without amplifying the 90W as well.

  496. cba
    Posted Apr 5, 2008 at 12:33 PM | Permalink

    486 (Leif):

    Sorry, I didn’t catch that it was a theoretical calc. rather than a measured one. There appears to be more OLR (measured) going on in Jul than in Jan despite the 90 W/m^^2 due to the jan perihelion and current eccentricity. Unfortunately, I’m working from memory as my charts are at the office and I’m not about to go there this weekend as I’m exhausted after 3 – 12 hr+ days instead of the typical 2 days in the middle of the week. There is 10-15 W/m^2 greater net OLR – TOAInsolation*(1-Albedo)in jul than in jan. Jan is a net sink, jul is net emission according to measured values. This is using the TOA value and not the 1 AU value for TSI.

    As a quick aside, I think Roy Spence recently showed that our actual atmospheric effects relate to the theoretical SB calculations by a 50% proportionality constant. I.E., if SB says 1 K, expect 0.5 K. This suggests that there is a strong negative feedback going on rather than a strong positive feedback as promoted by the model makers.

    While mostly, I’m remembering the net energy balance chart, I think the OLR chart was similar in that there was more OLR in Jul than in Jan. This suggests there could be a net T difference between Jul and Jan that is not in accord with the added TOA insolation in Jan. Note too that the measured Albedo varies from Jan to Jul by 2-3 %, but much less from year to year.

    The Q process might be associated with ocean absorption in some way. Low delta T for the W/m^2 power absorbed perhaps being transported for 6 months in currents and change depths etc. before it is transferred out into the atmosphere on average.

    It seems to me that Erl’s TSI minute variation stuff is not about w/m^2 but rather varying chemistry/effects from the differences in the spectral make up having different effects or effects at different locations(altitudes) which causes significant temporary changes in chemistry etc. which in turn results in more potent factors entering into play. The idea that such could be possible seems sound but there seems to be detail problems etc. and I’ve got very limited time to try to wade through this stuff – which is mostly out of my area of knowledge and interest.

  497. kim
    Posted Apr 5, 2008 at 12:49 PM | Permalink

    496 (cba) If there weren’t ‘detail problems’ this would have been already figured out.

    495 (Leif) We don’t know the mechanism. I understand your point about the wiggles in a big stream. Nonetheless, if it is the sun causing climate variation, those wiggles must do it somehow. It is theoretically possible for the variability of the sun to cause the changes. The mechanism must be both powerful and complicated.

    How about my idea of two solar cycles of one type and one of the other in each PDO flip? That could give 30 years of warming, then 30 years of cooling, as we have noted over the last century or so. Particularly so if cosmic rays correlate with cloudiness in alternating solar cycles. And those big, sparse, southern hemisphere spots during the Maunder Minimum have meaning, and the meaning implies solar control of climate. My ignorance is vast, but I don’t think these questions have been satisfactorily answered. Erl Happ seems on the trail of the mythical, magical beast.
    =================================================

  498. kim
    Posted Apr 5, 2008 at 12:50 PM | Permalink

    497 (kim) For more precision substitute ‘around 33 years’ where I write ’30 years.
    =========================================

  499. Posted Apr 5, 2008 at 1:02 PM | Permalink

    496 (cba): there is a measured 3.8 degree difference with July being warming, and OLR duly reflects that.

    Erl’s minute variations sometimes line up with dT and sometimes not. The lining up is not statistically significant, but Erl claims that statistics is worse that ‘damned lies’ anyway so he is unfazed. His specific mechanism [as far as can be discerned] seems to invoke direct heating of the troposphere by UV that somehow managed to survive the passage through the stratosphere. Now, if one wants to assert that the ‘active’ agent is the short-wave UV and beyond [which is much more likely, because of its larger variation], then to compare wiggles in TSI with anything does not make much sense because we have excellent UV, FUV, EUV, and Xray time series to compare with. Erl’s problem [as I see it] is that his ‘effects’ have become hostage to his theory about what causes them, and that does not help advance the field.

  500. Posted Apr 5, 2008 at 1:15 PM | Permalink

    497 (kim):

    if it is the sun causing climate variation, those wiggles must do it somehow.

    but here you are assuming that the sun is causing climate variation in the first place; there is the distinct possibility that it is not [and very probably not in the short run, comparable to the time scale of the wiggles].

    The cosmic ray difference between cycles is not in the amount but in the exact shape of the modulation, and is thus a minor second order effect.

    There are thousands of papers claiming sun-climate effects. It seems to me that the beast has its claws deep in many willing believers rather than they chasing the beast. A hallmark of such a situation is that no amount of rational discourse is going to make any difference.

  501. Posted Apr 5, 2008 at 1:25 PM | Permalink

    500 (me): about the slightly different shape of the GCR modulation: If, after another hundred years we have still not figured out what drives climate change, it should turn out that the climate variables [clouds, temps, phaetons, ...] also have the same characteristic shape changes, then we will have good evidence that the GCRs are important, but right now we are still on shaky ground.

  502. kim
    Posted Apr 5, 2008 at 1:38 PM | Permalink

    The beast we embrace is the only one strong enough to cause the changes, despite its diffidence.
    ===============================================

  503. pochas
    Posted Apr 5, 2008 at 2:18 PM | Permalink

    501 (Leif):

    Now, if one wants to assert that the ‘active’ agent is the short-wave UV and beyond…

    Well, I’ll repeat an observation I made a while ago, namely that the 9.6 micron secondary long-wave radiation from ozone can reach the surface directly. This may provide a way for UV to be downshifted to longwave and thus reach the surface. I haven’t estimated a forcing for this, but I sure wish somebody would. I can’t believe that with all of the radiation calculations that have been done this would have been missed, though. It was objected that temperatures in the thermosphere are somehow too low or “undefined”, but the signature is certainly present in the downwelling radiation spectrum.

  504. Mike Davis
    Posted Apr 5, 2008 at 2:39 PM | Permalink

    Leif:
    Maybe new question.
    If the sun is relativly constant what causes the cooling periods that the earth experiences?
    Maybe we are looking at this issue backwards. Instead of trying to figure out what is causing extra warmth. We should be trying to figure out what is blocking the warmth coming from the sun that has given us in the past the periods such as the H.O.,the R.W.P.,the M.W.P., and other warm periods during the Halocine Interglacial.
    From previous Interglacials we know that the sea level was at least 100 ft higher. Why not now?
    We know that the Swiss glaciers were smaller during this Interglacial. Why not now?
    I would propose that there is an optimum temperature for the planet and we have not reached that during this period because something has blocked the warmth of the sun from reaching the surface or failed to hold the warmth at the surface.

  505. Posted Apr 5, 2008 at 4:07 PM | Permalink

    I have two questions regarding uv and upper-atmospheric ozone:

    1. (This is the simple one) Where can I find a plot of the variation in solar uv over time?

    2. (Perhaps more difficult) Suppose that TOA uv increases by 1%. I presume that would increase the amount of ozone in the upper atmosphere. My question is, where on this plot of ozone partial pressure would the increase occur? Would it be evenly distributed (more or less) throughout the stratosphere (region B) or would it mainly affect the lower region (A)?

    Thanks in advance.

  506. Erl Happ
    Posted Apr 5, 2008 at 4:25 PM | Permalink

    Fantastic to get all that interest in the subject while I was sleeping. Cba has a grip on the thing and Kim is close. When cba says ‘Sorry, I didn’t catch that it was a theoretical calc. rather than a measured one.’ He is right on the money.

    The key to the difference in the response to solar radiation between January and July is the difference in the distribution of land and sea between the hemispheres.

    The two hemispheres react very differently to solar radiation. (it is not helpful to be talking about photons at the top of the atmosphere and genies etc).

    The different reaction to radiation (in terms of contemporaneous surface temperature change) lies in the different balance between absorption and emission between the hemispheres. One is a banker and the other a spender.

    The Southern hemisphere is mostly ocean and it absorbs radiation strongly. Despite the extra intensity of radiation in January the low ratio of emissions (via long wave radiation) produces a small temperature response for the hemisphere and the globe as a whole. This hemisphere is a banker. It stores energy.

    The Northern Hemisphere has much more land and it is on balance much more of an emitter, producing a strong rise in surface temperature in July despite the lower net irradiance at that time. This hemisphere is a profligate spender.

    The strong rise in surface temperature in the Northern Hemisphere (due to the increase in outgoing long wave radiation in July) heats the atmosphere, lowers relative humidity and evaporates the clouds. This reduces albedo. A greater proportion of the admittedly lower irradiance in July is actually effective at the surface. This produces an average 3.8°C increase in temperature over that in January. The response is to radiation at the surface, not to that at the top of the atmosphere. (genies go home, there is no magic necessary).

    It is erroneous to suggest that the sort of reaction that one should get to irradiation on a monthly scale will also apply to changes in radiation on the annual scale.

    Leif’s introduction of the perihelion/ aphelion argument is a red herring.

    Now, can we get back to a discussion of what happens in the atmosphere on annual scales when irradiance increases with the benefit of an understanding of the importance of changing humidity and cloud cover?

    Can we take into account what happens in the Southern Hemisphere when irradiance increases? The extra energy is banked. It is later spent in heating the Northern Hemisphere by virtue of ocean current transfer.

  507. maksimovich
    Posted Apr 5, 2008 at 4:25 PM | Permalink

    503(Pochas):

    Well, I’ll repeat an observation I made a while ago, namely that the 9.6 micron secondary long-wave radiation from ozone can reach the surface directly. This may provide a way for UV to be downshifted to longwave and thus reach the surface.

    The Radiative exchange between different layers of the atmosphere and the ground become dominant in “fixing” the temperature. The principal effect of decreasing O3 in the Stratosphere is to decrease the temperature because less radiation from the ground is absorbed by O3 in the 9.6 um band.

    Increased solar flux reaching the surface,is more then compensated by surface cooling through reduction of the greenhouse effect at the 9.3 um vibration-rotation band.

    Ramanathan et al 1976

  508. Barney Frank
    Posted Apr 5, 2008 at 5:07 PM | Permalink

    If, after another hundred years we have still not figured out what drives climate change

    If we haven’t figured out what drives climate change in a hundred years my descendants will be eating Ted Turner’s. :)

  509. steven mosher
    Posted Apr 5, 2008 at 5:15 PM | Permalink

    re 508. your descendants are going to aspire to eating brainless rat bastards?

    Hey. I’ve met hanoi Jane. slept in her friggin guest house. swam in her daddies pool.
    Her and ted should have procreated something, justs so we would have proof positive
    that they were human.

  510. Barney Frank
    Posted Apr 5, 2008 at 5:46 PM | Permalink

    509 (Steven)
    Ted says that’s all that’s going to be on the menu when the temps go up.
    Surely, it’s more humane to eat a brainless one than a sentient one.

  511. Posted Apr 5, 2008 at 5:48 PM | Permalink

    505 (DavidS):
    1) A place to start is here. But see also my criticism of the data here.

    2) Since the formation of Ozone depends on the number of energetic photons, if you increase these, the production would also increase in the upper atmosphere [where the big peak is]. This would eat up the photons and none would make it to lower layers (region A and below). The ozone in the troposphere comes from below, mostly from burning of biomass. To put the ozone content of the atmosphere in perspective: if we collect all the ozone from the TOA to the surface and make from it a layer with standard temperature and pressure, that layer would be 3 millimeter [~one eighth of an inch] thick.

    503 (pochas): 9.6 micron IR from O3. This was discussed back at #400 of Svalgaard #3. Guess the conclusion was that the energy involved was too minute. But if someone has a precise calculation of the forcing I would like to see it too.

  512. Posted Apr 5, 2008 at 5:58 PM | Permalink

    506 (Erl):

    Leif’s introduction of the perihelion/ aphelion argument is a red herring.

    Bad style, Erl… as red herrings imply intent. The issue was and still is how the atmosphere reacts to 0.2 W/m2 and not to 90 W/m2, both on a time scale of months. Remember that some of that ~90 W/m2 is the 0.2 W/m2. How does the atmosphere know to react differently on just those photons?

  513. Posted Apr 5, 2008 at 6:00 PM | Permalink

    506 (StevenMos): bad style and OT. Please refrain.

  514. Posted Apr 5, 2008 at 6:02 PM | Permalink

    509 (StevenMos) it was… not 506.

  515. Posted Apr 5, 2008 at 6:38 PM | Permalink

    Re #512 Thanks, Leif.

    One of the things I’m trying to understand is the behavior of the tropical upper troposphere, specifically the 15km-18km layer illustrated in this Gettelman et al figure . It’s a layer that probably plays an important role in tropical cloud behavior and overall heat removal.

    The intriguing thing (to me) is that the 15km-18km layer has a very low radiative-cooling rate. It’s low enough to where ozone heating plays a small but nontrivial role (see plot). In concept, a change in ozone concentration could cause a change in the layer’s temperature profile which could affect its interaction with rising air parcels (top-of-thunderstorm cirrus behavior). That, in concept, could affect tropical heat removal and thus global temperature.

    But, the variation in uv and ozone is too small to have a meaningful impact.

  516. Posted Apr 5, 2008 at 7:15 PM | Permalink

    506 (Erl):

    Now, can we get back to a discussion of what happens in the atmosphere on annual scales when irradiance increases

    All your plots and explanations (“spurts” of TSI) showed and were concerned with variation shorter than a year. If we plot yearly means ['annual scale' I presume] there are no spurts just a slow solar cycle variation. And of course hardly any ’0.2 W’ wiggles:

    Don’t be a moving target. Now, if you insist on annual scales, then, of course the 0.2W wiggles disappear and we don’t need to consider them anymore, which is actually just what I wanted to end up at.

  517. Posted Apr 5, 2008 at 7:20 PM | Permalink

    515 (DavidS):

    But, the variation in uv and ozone is too small to have a meaningful impact.

    I think that this is the fundamental truth if we are considering direct heating. What influence these things may have on other processes [circulation, planetary waves, etc] we don’t have a handle on.

  518. Erl Happ
    Posted Apr 5, 2008 at 10:44 PM | Permalink

    512, 516 (Leif)
    I want to have a close look at the variations comparing the surface with the troposphere and irradiance variations together with the perihelion/aphelion effect and come back to this question. Below is just ideas, not assertions, it’s just me thinking.

    I am curious as to why many of these peaks in temperature in the troposphere occur in December-January when global surface temperature is on the average 3.8°C cooler than in June-July. It occurs to me that it might have something to do with clearly associated peaks in solar irradiance, however small in amount, because logic suggests that with cooler global temperatures OLR should be low in January. That suggests that the peaks in temperature in the troposphere are not associated with surface trends.

    I am mindful of the statement by cba that : “Jan is a net sink, jul is net emission according to measured values. This is using the TOA value and not the 1 AU value for TSI. “

    Ocean absorption in December – February should have a big impact on the ENSO dynamic and ultimately northern hemisphere temperatures. The big reduction in cloud cover in the northern hemisphere summer will also strongly heat the oceans there, smaller in extent though they be. 3000 ARGO buoys say that the ocean is not warming after 2005. We had a little La Nina in 2006 and a big one in 2008. Ocean anomalies are currently negative over the bulk of the area.

    If one wants solar radiation to be geo-effective in raising temperature then December-February is the time to have it. As Leif said in relation to Ice Age avoidance this is the most favourable pattern. Where are we in relation to this cycle?

    A fuller understanding must face up to these questions: Is the temperature variation in the tropical troposphere at variation with surface temperatures? Which leads? Which follows? Is the pattern of variation in the tropics similar to that outside the tropics? Is global OLR behaving similarly to tropical OLR or are there lag effects?

    The conventional wisdom is that it is the surface that heats the atmosphere, rather than atmospheric heating reducing cloud cover and allowing more radiation to pass to the surface. That wisdom should be questioned. There is so little ozone in the stratosphere. How much energetic radiation can it absorb? David Smith raises the question in 505 as to whether energetic radiation affects cloud cover.

    For there to be a consistent pattern of increased irradiance (due to sunspot activity) in December-February one would have to ask whether the Earth is producing tidal effects on the sun due to its proximity at that time. The wisdom that says it is not, should be questioned.

    Mike Davis says ‘something has blocked the warmth of the sun from reaching the surface’ Changes in albedo are probably the key to surface temperature gain. Kim raises the question in 502. Any trend away from the status quo most probably requires an external agency. Kim makes this point in 497, 502. As I remarked earlier, the Earth has efficient processes for removal of heat. Without the sun it would be a one way trend to cooling.

  519. Posted Apr 5, 2008 at 10:48 PM | Permalink

    Hey Leif, I think that Hathaway is getting nervous. Check this link.

  520. Andrew
    Posted Apr 5, 2008 at 10:48 PM | Permalink

    516 (Leif): that graphic seems odd to me, as instead of converging, the different TSI’s seem to be diverging. Is there a reason for this, I mean are our measuring techniques getting less accurate, rathe than more? BTW, do you have data for TSI with monthly values (not the variations cuased by the annual cycle) and if so, where? I’d like to do some analysis relating to monthly values of temperature, but I don’t have enough data points with just the annual data on your site…

  521. Posted Apr 5, 2008 at 10:57 PM | Permalink

    (342) Leif

    as this is spurious it is hard to make sense of, but I would say that the longer a cycle is, the more it warms. Here I assume that at solar minimum we get a certain base amount and that within each cycle we get a little bit extra due to solar activity, so the longer the cycle the more extra we get. But this is just hand waving, there are people that say we get cooling, warming, nothing, whatever. We are past logic here.

    Leif

    Is not the “extra” TSI a function of the area under the curve divided by the time period of the curve? If weak cycles are longer then the total area under the curve divided by the time less?

    This is what intrigued me about a graph you did recently that showed the number of spotless days per cycle, with the weaker cycles having many more spotless days. It seems that the C-23/C-24 boundary area that we are in now has a lot of spotless days in comparison to recent strong cycles.

  522. stargazer
    Posted Apr 6, 2008 at 1:46 AM | Permalink

    (504)Mike Davis:

    I too have been wondering this for awhile..

    I would love for the folk here to consider putting figures to it.

  523. stargazer
    Posted Apr 6, 2008 at 2:07 AM | Permalink

    (504)Mike Davis:

    As Mike says, perhaps the sun (during an interglacial at least) is mostly ‘constant’ (this would suit Leif’s’ figures),

    and looking at graphs of the current interglacial ovet ~10,000 years ‘normal’= warm would seem to be the case…

    and (for some reason) the sun? ’causes’ the temp to drop from time to time during this period. So as Mike

    says perhaps this is where to look.(and this might better suit those looking for a solar conection)

  524. Posted Apr 6, 2008 at 3:46 AM | Permalink

    519 (Dennis):

    Hathaway:

    “The sun is at a critical point in terms of activity and observations,” Hathaway said. “We are starting to see new sunspots that could be part of a new cycle of activity.”

    The sun goes through roughly an 11-year cycle of minimum and maximum activity. It peaked in 2001 and early 2002, but now the activity could be starting to kick up again, Hathaway said.

    “It’s still within a couple of years away, but the activity is picking up we think, he said.

    Not only nervous, but slightly incoherent, I think. “We are starting to see new sunspots that could be part of a new cycle of activity.” The uptick so far has been old cycle spots. I hate it when people say “could be”. They either are or aren’t. And those weren’t. He cannot possibly be talking about that solitary tiny spot in January.

  525. kim
    Posted Apr 6, 2008 at 3:52 AM | Permalink

    I, too, was disappointed with that ‘could be’. Sure activity jumped from 70 to 90. From old cycle spots. Even I know better. Like I’ve said, I’m a better judge of rhetoric than science, and this from Hathaway is lamentable.
    ============================================

  526. kim
    Posted Apr 6, 2008 at 3:54 AM | Permalink

    524 (Leif) ‘Incoherent’ is generous. It is a lie, calculated to fool a journalist.
    =============================================

  527. Posted Apr 6, 2008 at 4:06 AM | Permalink

    520 (Andrew):
    TSIK data: Here are two of the prominent ones: ACRIM PMOD
    The series do seem to be diverging. The detectors are getting old and one of the main problems is how to correct for this ‘ageing’. And how to ‘stich’ newer data onto the old. The best ‘new’ data is SORCE. The ‘absolute’ level is unknown by about 5W/m2, so there is plenty of room to diverge.

  528. Posted Apr 6, 2008 at 4:09 AM | Permalink

    526 (kim): David Hathaway is a good man and doesn’t ‘lie’ or ‘calculate’. Please refrain from such accusations.

  529. kim
    Posted Apr 6, 2008 at 4:17 AM | Permalink

    528 (Leif) Alright, I believe you. But ‘could be’ is disingenuous, allowing plausible deniability. You know the man, I know the language, and how it is interpreted by a layman.
    ====================================

  530. Posted Apr 6, 2008 at 4:24 AM | Permalink

    518 (Erl):

    This is using the TOA value and not the 1 AU value for TSI

    And you [and Andrew and anybody else] should also be using the TOA value. Using the other [at 1 AU] value automatically disqualifies your analysis as junk.

    Your comment on tidal influences of the Earth upon the Sun, gets you perilously close to be snipped, so tread carefully here.

    521 (Dennis):

    Is not the “extra” TSI a function of the area under the curve divided by the time period of the curve?

    I would think, ‘minus’ the base value is the ‘extra’. But suit yourself; as I remarked: this is spurious and beyond logic.

    It seems that the C-23/C-24 boundary area that we are in now has a lot of spotless days in comparison to recent strong cycles.

    to me that is not a surprise if the next cycle is to be weak.

  531. kim
    Posted Apr 6, 2008 at 4:26 AM | Permalink

    497 (kim) Leif, please tell me what your belief is about the data that show clouds correlating with cosmic rays in alternate solar cycles. I believe that is a preliminary finding based on the analysis of only two cycles, but it is curious.
    ====================================================

  532. Posted Apr 6, 2008 at 4:40 AM | Permalink

    532 (kim): refresh my memory about the data. Where? Who? The GCR-climate thing is not on my front-burner.

  533. kim
    Posted Apr 6, 2008 at 4:50 AM | Permalink

    532 (Leif) Richard Black’s poorly written 4/3/08 BBC article mentions a Sloan and Wolfendale study published 3/15/08 finding a weak correlation between clouds and cosmic rays in one solar cycle and none in the next, using the same technique. It isn’t much evidence, and apparently Svensmark has denigrated it, already.
    =========================================

  534. cba
    Posted Apr 6, 2008 at 5:02 AM | Permalink

    524 (Leif):

    There were three rather visibile spots and the SSN got up to 63 and a M1.7 flare occurred and I think there was a CME observed this last week as well. All of this could be old cycle or some could be new cycle as the spots were all low latitude and I don’t recall hearing about the field reversal or not on them. Seems there was also a G1 and then a G2 magnetic storm. As ‘whoopi skidoo’ as these might be, they were significantly above the levels for quite some time even if it just might be a last ‘hurrah’. The low lat. suggests the stuff was last cycle but definitely doesn’t guarantee it.

  535. Posted Apr 6, 2008 at 5:08 AM | Permalink

    518 (Erl):

    There is so little ozone in the stratosphere. How much energetic radiation can it absorb?

    Get the physics straight! it is molecular Oxygen, not Ozone, that completely removes [absorbs] all radiation below 242 nm. Of the less energetic radiation above 242 nm, Ozone becomes the most efficient absorber. But the absorption is never starved for lack of Ozone. As we crank up the uv flux were get more Ozone automatically as there is always enough Oxygen around.

  536. Posted Apr 6, 2008 at 5:12 AM | Permalink

    534 (cba): all that activity was very definitely old cycle 23 stuff. NONE of it was SC24. The magnetic polarity was as for SC23, not 24, guaranteed.

  537. Dr. Gerhard Loebert
    Posted Apr 6, 2008 at 5:17 AM | Permalink

    [snip]

    Steve: I try to discourage this site being used for presentation of personal theories. Its primary purpose is analysis of peer reviewed articles being relied upon by IPCC.

  538. Posted Apr 6, 2008 at 5:30 AM | Permalink

    533 (kim): Ah, that one. They did not find a difference in response between the two cycles. They found no response in one of the cycles, suggesting that what was seen in the first cycle was spurious. This often happens, people see something, but it doesn’t hold up when more data comes in. Possibly as simple as that.

  539. lgl
    Posted Apr 6, 2008 at 5:46 AM | Permalink

    518, Erl

    Isn’t there far more water vapor in the atmosphere in july compared to january?
    Couldn’t it be the good old GH-effect peaking the temperature in july.

  540. Posted Apr 6, 2008 at 6:24 AM | Permalink

    537 (Loebert): seems to be 5 days too late.

  541. Erl Happ
    Posted Apr 6, 2008 at 6:54 AM | Permalink

    539 (lgl)
    Plenty of blue sky and warm land masses to do the heating in July via contact, convection and the release of the latent heat of condensation. These are first order things. Less cloud would mean a reduction in any greenhouse factor at that time.

  542. Phil.
    Posted Apr 6, 2008 at 8:20 AM | Permalink

    Re #521

    It seems that the C-23/C-24 boundary area that we are in now has a lot of spotless days in comparison to recent strong cycles.

    Significantly more than the last minimum apparently but I don’t think I’d call it ‘a lot’.

  543. kim
    Posted Apr 6, 2008 at 8:32 AM | Permalink

    538 (Leif) Thanks, Prof. Judith Curry has a link to the Sloan and Wolfendale paper in comment #492 of the Unthreaded #32 post.
    =====================================================

  544. Posted Apr 6, 2008 at 8:43 AM | Permalink

    537 (SteveM):

    I try to discourage this site being used for presentation of personal theories. Its primary purpose is analysis of peer reviewed articles being relied upon by IPCC.

    Erl’s theories seem to fall a bit outside of that, but so does some of my stuff. It is hard to draw a line sometimes, although in the case of Loebert it looked easy enough.

  545. Posted Apr 6, 2008 at 8:48 AM | Permalink

    542 (phil): I would defend ‘a lot’, because the numbers are 138 (for prev. cycle) vs. 170 for this one, and the previous cycles were even stronger with fewer spotless days. Note, that the next two or three months might make a real difference, if cycle 24 stays away.

  546. Posted Apr 6, 2008 at 9:09 AM | Permalink

    545 (me): I can’t even post correctly: 123 vs. 170, which is 38% more. That’s where the 138 came from [170/123=1.38]. April will likely have many spotless days [STEREO don't see anything so we may be in for a quiet stretch]. Unless Hathaway is right that cycle 24 is about to kick-off in a big way.

  547. kim
    Posted Apr 6, 2008 at 9:27 AM | Permalink

    546 (Leif) When I think of a Tesla ball and how little it takes to move all that electromagnetic force around, I compare it to the manifestation of the sun’s electromagnetic forces revealed by sunspots. Might not it take relatively weak forces to show large changes at the surface? Oops, barycentrism rears one of its hydra heads.
    ========================================

  548. Posted Apr 6, 2008 at 9:37 AM | Permalink

    547 (kim): the electromagnetic force is a million billion billion billion billion times stronger than the gravitational force. And can we not get away from the “might it not be”, “is it not thinkable”, “is it not conceivable”, etc, type of questions. I would like to raise the level of the discussion a bit.

  549. lgl
    Posted Apr 6, 2008 at 9:58 AM | Permalink

    541, Erl

    Well, since this is not a GHG-thread I’ll limit it to: http://isccp.giss.nasa.gov/cgi-bin/fetch_graphs.cgi
    found here: http://isccp.giss.nasa.gov/products/browseatmos.html
    shows a close to 16% annual variation in water vapor, which should give well above 3 C temperature variation if wv is responsible for more than 60% of the 33 C GH-warming. And the global low level cloud cover peaks mid year so increased solar doesn’t seem likely.

  550. Posted Apr 6, 2008 at 11:20 AM | Permalink

    549 (Igl): you should not a ‘cgi’ in a link as the response is dynamically generated.

  551. lgl
    Posted Apr 6, 2008 at 11:43 AM | Permalink

    549, 550

    Sorry, the cgi I mean is the plot created if you select Total Column Water Vapor and Global in the lower portion of the html

  552. EW
    Posted Apr 6, 2008 at 11:59 AM | Permalink

    More cycles and correlations ;-) I was searching for more about that detrended global T and length-of-the-day and other variables found this. The solar connections do not look very high, though. As it is a SORCE meeting paper – any comments, Leif?

    According to that link at FAO from 2001 about the relationship between temp and LOD we were due for a cooling:
    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.
    I wonder, what LOD forecasts these days…a paper with LOD graph (behind the money wall) up to Nov. 2006 suggested that somewhere around 2010-11 there will be warming again.

  553. Posted Apr 6, 2008 at 6:18 PM | Permalink

    552 (EW): although there are theoretical reasons for a LOD link to climate and to solar, I agree that the latter is weak [as almost any solar connection to climate]. I would not base my future purchases of heating oil on any of these forecasts.

  554. CAWolk
    Posted Apr 6, 2008 at 9:45 PM | Permalink

    This debate has become extremely confusing. I’d like to go back to some comments made much earlier.

    In Erl Happ’s initial posting, he wrote, “Since energy gain at the surface promotes evaporation (energy gain is almost completely resolved in increased evaporation rather than surface temperature increase in the tropical oceans) an increase in atmospheric transparency due to reduced cloud cover relies upon the change in irradiance being fast enough, and with sufficient persistence, to promote a net drying of the atmosphere against the countervailing force of rising humidity tending to restore cloud cover.”

    In response, and referring to the big, swinging red curve, Leif wrote, “Because of rotation, seasonal tilt, etc, the red variation is distributed [unevenly] over the whole globe and does not in itself cause any long-term effects because of the annual repetition, but it is absurd nonsense to think that our turbulent and complicated and interacting atmospheric and oceanic systems at all times precisely cancel out the very large red variation to the extent that we can ascribe any significance to the blue wiggles in real time.”

    Could the condition that the rise in TSI must be fast enough to dry the atmosphere not exist in the red curve after the climate’s reaction to the annual swings and only exist in the 1 AU curve?

    Secondly, it would be helpful to plot the first derivative of the TSI curve Erl Happ presented against temperature during the same period.

  555. Allan MacRae
    Posted Apr 6, 2008 at 10:11 PM | Permalink

    I haven’t been able to read this site regularly, but am disappointed to see suggestions that Hathaway is acting in bad faith. Haven’t we climate skeptics (aka “deniers”) been subjected to this sort of unfair treatment, and can we not rise above it? It is clear that there are two camps within NASA regarding SC 24 predictions – one high and one low – this suggests the subject is complex and, if I may borrow a phrase, the science is not settled.

    On a lighter note, NASA is well-covered since its SC24 predictions span the full range of possibilities.

    I’ve started a poll on CS and so far, all SC24 predictions are on the low end of the range. Mind you, many of the contributors are not experts, and admit to having used voodoo to reach their predictions. So what happens if they are right?

    Best, Allan

  556. Pete
    Posted Apr 7, 2008 at 2:57 AM | Permalink

    Leif,

    What are your thoughts on this latest statistical research by Jan Janssens, posted over the weekend, suggesting we are still in the downward slope of transition, and Solar Minimum can be expected March 2009 +-6

    Janssens

  557. Geoff Sherrington
    Posted Apr 7, 2008 at 5:34 AM | Permalink

    Leif,

    My son graduated in Surveying and is good at math. I’m not. I always thought that the reversals in the earth’s magnetic field were caused by earthly processes. Son writes

    The sun has just exited a polar reversal. The magnetic north of the sun is now south if the plane of the ecliptic is taken as reference to the magnetic north of the earth. China has produced maps which have a magnetic reading indicating a south that is 18 degrees short (162 degrees current north magnetic/approx elliptic). Speculation is that around 5000 years ago when these tables were engraved and stored, that there were local south’s showing that an Earth flip was drawing to a close.

    The Planets excluding Pluto and one moon all have magnetic north in similar directions as the current earth north. And rotate in the same directions excluding previous.

    Running through sedimentary layers over millions of years shows Iron particles aligned North South to South North but 87% are either NS or SN. (Australian Geology, MacMillian.1979. pp?) A high school science experiment shows that running a bar magnet around another bar magnet that is inverted causes warming, it is simply referred to as magnetic drag. See any article on Maglev trains.

    Is there mileage here? It’s beyond me.

  558. Posted Apr 7, 2008 at 6:58 AM | Permalink

    554 (CAWolk): cites Erl’s contradiction in terms: “relies upon the change in irradiance being fast enough, and with sufficient persistence” as being a meaningful condition. These ‘changes’, as I have shown, lead to undetectable direct heating of less than 0.01 degrees. CAW states rightly that Erl’s ideas are very confusing [or at least result in a very confusing 'debate']. Ordinarily in scientific discourse, peer review prevents such a debate from getting started in the first place. In a blog such as this we lower the entry bar a bit to allow a less strict review and to benefit from the interaction between posters and commentators. May I remind everybody about what Steve M said in comment #537:

    Steve: I try to discourage this site being used for presentation of personal theories. Its primary purpose is analysis of peer reviewed articles being relied upon by IPCC.

    We have broadened that a little bit as solar activity was only peripheral to the IPCC. But all within reason and for the purpose of increased clarity in this murky field. In this particular case that CAW cites, we are not achieving this purpose. This is, of course, not unique to this particular debate, and there are [standard] ways of dealing with such a situation, like ‘pruning the decision tree’, i.e. taking smaller steps and examining each to see if that direction is good, and if not, not go down that branch.

  559. Posted Apr 7, 2008 at 7:13 AM | Permalink

    555 (Alan): I agree. About voodoo predicting a low cycle 24, there is also good science doing that, so things are not that bleak.

    556 (pete): cycle 23 will peter out on its own schedule regardless of when cycle 24 starts, so we cannot really use past behavior of 23 to predict when 24 starts, so minimum is really impossible to ‘predict’ as such. Any ‘prediction’ has a built-in assumption of what cycle 24 will bring. Let me remind all that cycle 20 was long too, and yet cycle 21 was very strong.
    What does seem certain is that minimum was not in March 2008 or earlier. Moving minimum six months ahead would seem reasonable. More than that, e.g. a year ahead, is just speculation [which may turn out true].

  560. Posted Apr 7, 2008 at 7:24 AM | Permalink

    557 (Geoff): Son may be good at math, but is weak on facts and science. There is no mileage there. Sun’s magnetic field near the Earth is 100,000 times weaker than the school experiment’s bar magnet.

  561. Posted Apr 7, 2008 at 9:03 AM | Permalink

    [snip- personal hobbyhorse theory. please discuss elsewhere]

  562. Posted Apr 7, 2008 at 9:20 AM | Permalink

    3 (cba): back in#3 cba asked:

    What is the method used to extract the orbital position modulation of the TOA from the data, used either for SORCE or for ACRIM (which is the one I need since SORCE already has it)? Is it just a daily orbital position correction from 1AU? That’s nice for analyzing the sun but it’s a royal pain trying to deal with that when concerning actual TOA incident radiation.

    The orbital modulation does not have to be added to the actual TOA data. TOA includes the modulation and is what is actually observed and is what we should use when studying the Earth. For studying the sun, we simply divide by the square of the distance to the sun in AU. No pain at all. Of course, you need to know what the distance, R, is. Here is a formula to calculate R:
    First express time, t, in Julian millennia from the epoch J2000.0 as
    t = (JDE – 2451545.0)/365250, where JDE is the Julian Ephemeris Day [find out yourself to calculate that - try google], then
    R = R[0] + R[1]*t + R[2]*t^2 + …
    Each of the R[n]‘s is itself a series of cycles [the ancient Greeks would delight in this - we today still use a form of their epicycles] as follows:
    R[n] = sum{i=1 and up}(A[n,i] * cos(B[n,i]+C[n,i]*t))
    The coefficients B and C are in radians.
    Here are the more significant values:
    for R[0]:
    i=1 : A=1.00014 B=0 C=0
    i=2 : A=0.01671 B=3.09846 C=6283.07585
    i=3 : A=0.00014 B=3.05525 C=12566.1517
    i=4 : A=0.00003 B=5.19850 C=77713.7715
    for R[1]:
    i=1 : A=0.00103 B=1.10749 C=6283.07585
    for R[2]:
    i=1 : A=0.00004 B=5.78460 C=6283.07585
    Higher terms are not needed for TSI corrections [Modern theory goes to dozens of epicycles]

  563. Barney Frank
    Posted Apr 7, 2008 at 9:36 AM | Permalink

    561 (Jonh Ackerman)

    I didn’t get to read Dr. Loebert’s theory in 537 but surely it was not more worthy of a snip than 561.

  564. Posted Apr 7, 2008 at 9:54 AM | Permalink

    562 (me): OK, to calculate the Julian Day JD:
    Y = year
    M = month
    D = day
    if M=1 or 2: Y = Y-1, M=M+12
    A = int(Y/100)
    B = 2 – A + int(A/4)
    then
    JD = int(365.25*(Y+4716)) + int(30.6001*(M+1)) + D + B – 1524.5
    The difference between JD and JDE [ephemeris time, is of the order of 100 seconds and can be ignored].

  565. Posted Apr 7, 2008 at 9:57 AM | Permalink

    563 (Barney): Steve might snip 561 too. Luckily the link is dead, so not much harm.

  566. cba
    Posted Apr 7, 2008 at 10:35 AM | Permalink

    562 565 (Leif):
    Thanks for the details. There’s no problem with SORCE as it offers a TOA set. I think I was only able to find a 1 AU data set for ACRIM though. I can’t even remember offhand as I haven’t had time for a couple of weeks to even work on this project.

  567. Posted Apr 7, 2008 at 11:38 AM | Permalink

    566 (cba): and one should also apply a little common sense to my postings. In 562 I said to study the sun we “divide by” where I meant “multiply by”. Of course, to get TOA from TSI you divide.

  568. jae
    Posted Apr 7, 2008 at 12:02 PM | Permalink

    Leif, 553:

    552 (EW): although there are theoretical reasons for a LOD link to climate and to solar, I agree that the latter is weak [as almost any solar connection to climate]. I would not base my future purchases of heating oil on any of these forecasts.

    Well, this looks to be a better basis for predicting my purchases than tossing a coin. It seem right up to April, 2008. Do you know of a better way of predicting?

  569. Bill P
    Posted Apr 7, 2008 at 12:35 PM | Permalink

    Dr. Svalgaard:

    I’ve seen you refer to the sun’s weak effects on climate (and therefore biological proxies of climate), but I haven’t seen the question posed per se (haven’t reviewed all your thousands of comments, so please forgive a topic you’ve either dealt with or dismissed): have you seen evidences of solar cycles represented in any long-term dendrochronological record?

  570. Posted Apr 7, 2008 at 2:57 PM | Permalink

    568 (jae): here are some forecasting methods for temperature:
    1: forecasting using Earth weather/climate models
    2: forecasting solar activity [e.g. using my method), then using the [weak] solar-climate connection, forecasting the climate from the forecast solar activity
    3: using the LOD to forecast solar activity, then the solar-climate connection to forecast climate [or rather long-term weather {next year}]
    4: tossing a coin

    Take your pick. With 4 you’ll be correct half of the time. I’m not so sure about the others.

  571. Posted Apr 7, 2008 at 3:05 PM | Permalink

    569 (BillP): no, I don’t see any such [solid] evidence, but there are thousands of people that do. Try googling “tree rings solar cycle”.

Follow

Get every new post delivered to your Inbox.

Join 3,381 other followers

%d bloggers like this: