New Emanuel Presentation

Howard Wiseman sends in:

I attach a link to Dr. Kerry Emanuel’s recent powerpoint presentation at the Southern New England Weather Conference which focuses on TC activity in the Atlantic. The download is a bit slow, and the presentation is pretty familiar ground. Emanuel’s conclusions are presented with breathless inevitability (surprise). No comments on the 2006 Atlantic TC season that wasn’t, although there might have been at the live presentation which I did not attend.

Kerry Emanuel on Global Warming and Hurricanes

And now, for your viewing pleasure, the link to Joe D’Aleo (“‘?Dr. Dewpoint”‘?). Some nice graphs and graphics here that are a pretty good counterpoint to the Warmanistas.

Joe D’Aleo: Climate Change

Last but certainly not least, RC’s take on the EPICA study and DO events and ocean circulation flips. Buried in there is an admission by Eric Steig of the possibility of a solar cycle driving a 1470 year long climate regularity (but not provable since the Holecene) and expressions of doubt and curiosity. Perhaps the penny has dropped on bad science??


  1. David Smith
    Posted Nov 8, 2006 at 10:04 PM | Permalink

    Howard, the Emanuel link works well, but I haven’t been able to reach the D’Aleo link. Perhaps a bad link. Thanks

  2. Howard Wiseman
    Posted Nov 8, 2006 at 10:21 PM | Permalink

    Here’s another try of D’Aleo

  3. Howard Wiseman
    Posted Nov 8, 2006 at 10:25 PM | Permalink

    David, Once more, with feeling. IE7 is messing with the script. Try the cut and paste approach.‘Aleo%20-%20Climate%20Change.ppt

  4. Howard Wiseman
    Posted Nov 8, 2006 at 10:43 PM | Permalink

    I cannot get through either. I am afraid I might have sent these folks more traffic than they bargained for.

    BTW,David, thanks for the referral to Steve M. My Very Own CA Topic. WOW!

    The conference is attended by amateur weather geeks like me, and is strongly supported by NWS Boston Office and its outstanding Mets. It is probably a not hard core math and physics, but a cut above nevertheless. I strongly recommend the outstanding presentation by David Vallee of NWS on the Coastal Bomb of 12/9/2005. The link is at the site, but I won’t post it right now, since it doesn’t look like they are equiped for CA traffic levels. I remember the storm well as I busted a fender that day ($1,000). I posted here on these warm core winter rogue storms when Dr. Curry was here. Using the thesis that hurricanes serve a primary climate function in transporting heat from the tropics to the temperate areas, I had suggested that as the northern latitudes warm and temperature disequilibrium with the tropics reduces, TC’s may grow less intense rather than stronger. JC rejected this notion, referring to the absence of TC’s during winter at the height of thermal differentials. I replied by referring to these hybrid warm core winter storms. And then she went away.

  5. Jaye
    Posted Nov 8, 2006 at 10:49 PM | Permalink

    The address that you are pasting is actually this:

    The link you want is here:


  6. Jaye
    Posted Nov 8, 2006 at 10:50 PM | Permalink

    Or here without the quotes…'Aleo%20-%20Climate%20Change.ppt

  7. Jaye
    Posted Nov 8, 2006 at 10:51 PM | Permalink

    Arghh…something on the server side is screwing up the link.

  8. Posted Nov 8, 2006 at 11:01 PM | Permalink

    Go to the home page at,
    click on “2006 Presentations” and scroll to the bottom.

  9. bender
    Posted Nov 9, 2006 at 1:33 AM | Permalink

    Aside from the already-discussed problems of smoothing and pinning and inflated correlations, (not to mention the leap-of-faith attribution of temperature rise to CO2), there is one serious problem with Emanuel’s presentation. His slide 49 predicts an increasing trend in landfalling hurricanes, especially cat 3 & 4. But there is no trend whatsoever in HURDAT landfalling hurricanes. The published increasing trend (Emanuel 2005, Mann & Emanuel 2006) (slides 8,9,13,14) is for basin-wide TCs, not landfalling TCs.

    This may simply because we haven’t observed the system long enough to observe the predicted increasing frequency of landfall (ergodicity, you Bayesians?). Or maybe the model is incomplete. Alternatively, maybe the hypothesis is false.

    Would be nice to ses a recognition and discussion of these issues. But maybe that would be admitting too much uncertainty?

  10. TAC
    Posted Nov 9, 2006 at 5:42 AM | Permalink

    Last but certainly not least, RC’s take on the EPICA study and DO events and ocean circulation flips. Buried in there is an admission by Eric Steig of the possibility of a solar cycle driving a 1470 year long climate regularity (but not provable since the Holecene) and expressions of doubt and curiosity.

    Along these lines, I note that Harry Lins (, who has been somewhat skeptical about the significance of reported climatological trends, is convening a session at the European Geosciences Union (EGU, the European equivalent of AGU) General Assembly in Vienna in April 2007 on the topic:

    HS41 Statistical concepts in understanding and modelling hydro-climatic change
    (co-listed in NP, CL and AS)
    Convener: Lins, H.
    Co-Convener: Bunde, A.; Dolman, H.; Koutsoyiannis, D.; Pegram, G.

    Assessment of hydro-climatic variability and change depends critically on the application of statistical concepts and models that reflect the enormous complexity of the hydro-climatic system and the ineffectiveness of purely deterministic descriptions. Concepts from probability, statistics, and stochastic processes provide tools for data analysis that quantify the climate and its evolution, and enhance understanding of the dynamics and causative mechanisms of hydro-climatic processes. They also enable the modelling of processes, the projection of their future evolution (either as stand alone tools or in combination with deterministic approaches), and the assessment of the uncertainty and limitations of projections. Traditionally, standard statistical models based on independence or simple (e.g., Markovian) dependence have been widely utilized in typical statistical tasks such as estimation and prediction of hydro-climatic parameters and processes. More recently, richer stochastic structures have been identified in the climatic system, which reflect the complex nonlinear character of this system and may have implications in assessing and predicting hydro-climatic change.

    This session is designed to explore the statistical characterization and modelling of the hydro-climatic system. In particular, contributions are sought that shed light on the following issues:

    – Characterization of climatic variability with emphasis on the study of teleconnections in hydro-climatic signals;

    – Recent advances with respect to trend detection, attribution, significance testing, and the statistics of extremes;

    – Assessment of uncertainty in hydro-climatic projections;

    – Coupling of deterministic and stochastic approaches, and probabilistic descriptions of causative mechanisms;

    – Spatio-temporal statistical approaches with an emphasis on statistical downscaling of deterministic climatic projections; and

    – Multifractal characterization of hydro-climatic processes.

  11. Spence_UK
    Posted Nov 9, 2006 at 7:30 AM | Permalink

    Re #10

    Thanks for pointing that one out – the issue of self-similarity and complex multi-scale variability in climate dynamics is a bit of a hobbyhorse of mine. To me, these questions are right on the button and exactly what climate scientists should be asking.

    I picked up on Demetris Koutsoyiannis’ work through a post on CA by Steve, and learnt a great deal through his published work and through some of David Stockwell’s posts on the subject as well. I’d recommend it to anyone interested in climate science.

    The beauty of science is even if it goes astray for a short while, the truth will out and the train will get back on the tracks eventually.

  12. beng
    Posted Nov 9, 2006 at 8:09 AM | Permalink

    And now, for your viewing pleasure, the link to Joe D’Aleo (“Dr. Dewpoint”).

    That’s funny — “Dr. Dewpoint” used to be featured regularly on He was a well-thought-out “skeptic”, and I wondered how long intellicast would tolerate it. Not very long.

  13. TAC
    Posted Nov 9, 2006 at 9:01 AM | Permalink

    #11 Note that you are welcome to submit an abstract and give a talk. I, for one, am thinking about doing so. This session presents an opportunity for some critical discussion about the concept of statistical significance with respect to climate trends. It could be interesting. Given what we know — and, more important, what we don’t know — about the spectrum of the noise in climatological data, even defining the concept of trend significance presents a challenge.

  14. beng
    Posted Nov 9, 2006 at 9:06 AM | Permalink

    RE 4: Howard says:

    Using the thesis that hurricanes serve a primary climate function in transporting heat from the tropics to the temperate areas,

    Obviously, if TSs move into temperate latitudes, that’s true. However, much of the lifetime of TSs are spent in the tropics/subtropics. In that situation, the “heat-valve” aspect is rather localized (the low-pressure area and the surrounding, supporting high-pressure) and looses heat directly to space.

  15. David Smith
    Posted Nov 9, 2006 at 9:36 AM | Permalink

    I’m just starting to read these Powerpoint presentations. Two early questions I would have had for Emanuel are the following:

    ” Dr. Emanuel, your slide #9 shows a correlation (r-squared = 0.74) between Atlantic storm count and Atlantic sea surface temperature since 1970. Yet, your slide #6 shows no trend in global storm count since 1970, an era of rising global temperatures (slide #28) including tropical sea surface temperature in the MDR regions? Why is the Atlantic behaving differently from the global trend?”

    “Dr. Emanuel, there appear to be good storm count, SST and intensity data on the Northeastern Pacific since 1970. The data is handled by the same people (NHC) who handle the Atlantic data, so the data should be apples-to-apples. What are the trends in Northeast Pacific SST and storm count? Or, Northeast Pacific PDI and SST? Do they confirm what you suggest about Atlantic SST and storm count/intensity? if not, why not?”

    Perhaps there are good answers to these questions. Howard, do you recall if he touched on these?

    (By the way, in looking at Northeast Pacific (EPAC) SST, I noticed an apparent error in Webster et al’s Figure 1. The plot of EPAC SST temperature does not match the NCEP data nor the SST plot of Hoyos et al. I assume it’s a mislabel or misplot. These Webster errors are inconsequential, no big deal. Personally, though, before I’d submit a paper to Science, I’d double-check my work because I make mistakes and I’m surprised Webster did not do the same.)

  16. bender
    Posted Nov 9, 2006 at 9:44 AM | Permalink

    Re #10
    Looks interesting. Is the climate system an ergodic system? To the extent that it is open and chaotic, the answer must be no. But of course some components of the climate system are more closed, more tightly bounded, and better behaved than others. Maybe Emanuel’s Atlantic basin subsystem is not far from ergodic? Seems to me this is a central problem that can not be ignored in a session on statistics and uncertainty. Whether you’re a frequentist or a Bayesian.

  17. gb
    Posted Nov 9, 2006 at 11:38 AM | Permalink

    Re # 10.

    The topic is hydro-climate changes, not climate changes. One has to read careful what the actual topic is of the session before submitting an abstract.

  18. Howard Wiseman
    Posted Nov 9, 2006 at 1:12 PM | Permalink

    Dear Beng:

    I would concur that heat transport doesn’t occur when TC’s don’t gain latitude. However, TC’s ultimately circulate around the boundry of upper tropospheric high pressure areas and spend their last westerly days moving north and undergoing extratropical transition. At this point heat gets dumped into the baroclinic weather and usually energizes it significantly.

  19. Brooks Hurd
    Posted Nov 9, 2006 at 5:13 PM | Permalink

    Emanuel states in slide 11 that the pre-1950 data is unreliable for hurricaine intensity. I do not disagree with him on this, but why does he believe that any of the pre-1950 data on tropical cyclones is reliable?

    Is the pre-1950 Atlantic storm count (slides 8 & 9) reliable?

    Is the pre-1950 Atlantic SST (slide 27) reliable?

    If any of these data are unreliable, then Emanuel’s statements on slides 3 and 4 (largest numbers, lowest pressures, largest PDI, etc.) are based on unreliable data.

    I can make the same argument for his summary (slide 50).

    The statements on slides 3, 4, and 50 are only correct if the data on which they are based is both consistent and reliable.

  20. Willis Eschenbach
    Posted Nov 9, 2006 at 7:39 PM | Permalink

    Re #18, Howard, once again, I have to ask for a mechanism. In your interesting post, you say:

    I would concur that heat transport doesn’t occur when TC’s don’t gain latitude. However, TC’s ultimately circulate around the boundry of upper tropospheric high pressure areas and spend their last westerly days moving north and undergoing extratropical transition. At this point heat gets dumped into the baroclinic weather and usually energizes it significantly.

    I still don’t understand how this is supposed to occur. The TC picks up heat from the ocean, moves it vertically on a time scale of minutes to hours, and either dumps it into space as radiation or into the ocean as rain. Compared to the amount of heat moved by the TC, the amount it contains at any given moment is small. In addition, the heat it contains at any moment was picked up in the near vicinity. So where is the lateral movement taking place?

    As a relevant aside, an oft-neglected aspect of weather is that it is one of the few places in nature where cold is transferred as well as heat. We are all familiar with the transport of heat from a warmer object to a colder object. This takes place by the normal means of convection, conduction, and radiation, and is generally a one-way process.

    But the atmosphere, by way of hydrometeors (rain, snow, graupel, hail, etc.), transfer cold from a colder object to a warmer one. When a TC passes over the land, the land does not get warmer, it gets colder.

    So what is the mechanism for the lateral transport of heat from TCs? I’m not picking on you, I’ve asked this question before without getting an answer. Do you have any citations with numbers for this claimed movement of energy (as opposed to citations that just claim it exists)?


  21. David Smith
    Posted Nov 9, 2006 at 7:54 PM | Permalink

    Figure 9 in Emanuel’s Powerpoint shows SST since 1870 versus storm count since 1870. Figure 2 in Emanuel/Mann (13 June 06, EOS) shows the same two things.

    The odd thing is, Emanuel’s changed the SST curve. Powerpoint uses a “smoothed curve” based on “SST at key latitudes”. Emanuel/Mann uses “decadally-smoothed MDR SST”.

    And, the storm count curves are different. Powerpoint uses “storm count smoothed with a weighted filter” while Emanuel/Mann uses “decadally smoothed count”.

    All I can figure is that massive smoothing is more statistically-appropriate for residuals calculations (Emanuel/Mann) while the other, weighted smoothing is more appropriate for calculation r-squared (Powerpoint).

    I don’t have a clue why he redefined the SST region.

  22. David Smith
    Posted Nov 9, 2006 at 9:01 PM | Permalink

    Re #19

    Further on Brooks’ points, Emanuel uses 1870-1950 intensity data, which Emanuel himself describes as “unreliable” data, to calculate PDIs and display them in slide 12!

    Interestingly, there are two PDI/SST charts. Slide 12 shows “storm max PDI” over 1870-2003 in a comparison with “SST at key latitudes”. The slide gives a visual impression of good correlation.

    Then, in slide 14, Emanuel switches to a different definition of PDI and a different SST region (the old Emanuel’s MDR box) to show a correlation over 1970-2000.

    Why start slide 14 at 1970 instead of 1950, or 1870? Why not use the same SST definition for both charts?
    I bet an internet dollar/euro/yen that the visual impressions of the charts would be different.

  23. bender
    Posted Nov 9, 2006 at 9:40 PM | Permalink

    Why start slide 14 at 1970 instead of 1950, or 1870? Why not use the same SST definition for both charts?

    Surely not to improve the optics in order to promote the alarmism coefficient, A?

    I have to say that from a superficial survey of what is presented there, his predicted increase of 50% in cat 3 and 100% increase in cat 4 hurricanes, stretched over that long a time scale, is not as alarming as I would have expected from this group. That trend would be absolutely dwarfed by the natural variability – which we have seen from 2005 vs 2006, is very spiky. This, I think is one of RPJr’s main arguments.

    Do you think insurance companies will increase rates commensurate with this modest increase in risk, or will they double or triple them over a much shorter time period in order to make a few extra bucks? Wonder if any consumer advocacy groups are watching this closely enough.

  24. David Smith
    Posted Nov 9, 2006 at 9:46 PM | Permalink

    Slide 16 of Emanuel is interesting.

    It shows PDI, storm count, storm intensity and storm duration for the Atlantic from 1950-2003. (Note: the wiggles of this third PDI curve do not match perfectly the other two PDI curves, for whatever reason, but I’m ignoring that here.)

    Slide 16 shows, for 1950-2003:

    * no trend in average Atlantic storm intensity
    * no trend in average Atlantic storm duration
    * recent upward trend in average Atlantic storm count

    To the extent that PDI is higher today, it’s because of the higher storm count, not due to higher intensity or duration.

    (I am assuming that I am correctly interpreting his unlabeled scale and chart.)

    Those are remarkable findings.

    The one thing showing an uptrend is Atlantic cyclone count, but global data shows no trend in global tropical cyclone count. That observed no-trend is consistent with the generally-accepted belief, by skeptics and warmers alike, that GW probably won’t appreciably increase the count of tropical cyclones.

    So, if the problem is not intensity, and it is not duration and (per global history) it is not storm count, then what is the problem? Am I reading his chart properly?

  25. Gerhard H. Wrodnigg
    Posted Nov 9, 2006 at 10:05 PM | Permalink

    Re # 10:

    This conference looks very interesting! There are many hot topics in the “Climate: Past, Present, Future – Programme”

    Since this conference is just round the corner to me, I will be there.

  26. Posted Nov 10, 2006 at 7:43 AM | Permalink

    #20 Willis…

    Latitudinal Heat Transfer in Tropical Cyclones

    Tropical Cyclones most often form deep in the warm tropics (there are exceptions) near the low pressure InterTropical Convergence Zone (ITCZ), a natural boundary where surface air drawn from the subtropical ridge (approx 30N & S of the equator) of high pressure in both hemispheres meet, having become warm and moisture laden along the way. The warm moist air rises with convective effects carrying it aloft to great heights where it radiate heats directly into space. The ITCZ tends to spend much of it’s time in the hemisphere experiencing Summer and Autumn, migrating twice a year across the equator to follow the seasons. It is more complex than this simple explanation, however it will do for our purposes here.

    Along the poleward side of the ITCZ most often in the hemisphere experiencing it’s Summer to Autumn season, and especially in disturbed areas where pulses of higher pressure air from the subtropical ridge interact with it, lower pressure vortexes of increased uplift and convection aided by the coriolis effect of the rotating earth may form that can establish themselves and begin to dominate the local environment, distorting the ITCZ by drawing the low pressure ‘monsoon trough’ running along it’s centre into the vortex, enabling the developing low to take full advantage of the flows of warm moist air being drawn into the trough from opposite directions in both hemispheres – we call these systems tropical lows or tropical depressions or monsoon depressions, depending on where and how big they are – a typical low may be many hundreds of kilometres in diameter, but sometimes they can be 1500 or more kilometres in diameter.

    In many cases, these systems move west, being steered that way by prevailing winds from the subtropical ridge while also drawing warm moist air from a huge area in both hemispheres into a large quasi-circular storm as increasingly strong convective bands wrap around the central core which begins to tighten up whilst increasing amounts of heat and moisture is being pumped aloft – such a system has become a fully fledged tropical cyclone, also called a tropical storm, and as it continues to intensify it may develop a central eye, becoming a severe tropical cyclone, also called a typhoon or hurricane.

    As you can see from the above, we have a huge area of warm moist air rotating around a central vortex that is pumping a large quantity of heat into the space whilst also drawing all the warm moist air it can from a huge area surrounding it from both hemispheres into it’s large circulation.

    As the system matures it gains so much mass in the form of moisture that the coriolis effect aiding it’s rotation also starts to push the system more polewards, so that it veers further from it’s initial more westward track, and sooner or later it will find a weakness in the subtropical ridge that has been trying to keep it steering westward, often caused by a temperate cold front or short wave trough eroding the ridge from the poleward side, which then allows the cyclone to turn poleward through the weakness, carrying it’s load of warm moist tropical air polewards with it.

    As you can see, a cyclone will vent heat into space, but it will also transfer a large heat load in the form of a huge circulation of warm moist air out of the tropics into the mid and temperate latitudes as it makes it’s final fatal dive polewards.

  27. David Smith
    Posted Nov 10, 2006 at 8:23 AM | Permalink

    Tropical Storm Rosa formed this week in the Pacific near Mexico. Rosa reached minimal storm strength (35 knot winds) for all of 12 hours, before degrading into a depression. A plot of Rosa is here, with the yellow line being its entire life as a tropical storm.

    I counted about two similar weak systems per year in the Atlantic, since 2000.

    It is doubtful that these would have been detected and reported pre-satellite era.

    This is yet another problem with comparing current storm count against historical numbers.

  28. Posted Nov 10, 2006 at 9:13 AM | Permalink

    BTW Willis, I do not know where you can find numbers on heat transferred latitudinally this way in cyclones, however the conceptual view I presented above can give you a picture to work with. You need to be aware that there is a lot of heat stored in the moisture in the lower atmosphere of any cyclone, and this is quite evident when on the poleward side within the closed circulation, where the heat and humidity streaming around the system from more tropical areas feels oppressive even when one is located in the sub-tropics.

    OFFTOPIC: Typhoon CHEBI has undergone a central pressure drop of 57 hPa (=MB) with winds increasing from 55 kts to 115 kts in just 6 hours as it approaches landfall in the Philippines – I think this sets a new record for rapid intensification!

  29. Brooks Hurd
    Posted Nov 10, 2006 at 12:10 PM | Permalink

    Re: 27; David

    I agree with you on the detection of weak storms pre satellite. I would take it a step further and point out that non-landfalling storms in the 1870 – 1930 time frame would only have been reported by survivors. Very strong non-landfalling TCs would have tended to remove those who might have reported their existence.

    The Existentialist question could be restated for Kerry Emanuel as:
    “If a TC develops and does not make landfall, did it exist if all its observers were lost at sea?”

  30. David Smith
    Posted Nov 10, 2006 at 3:56 PM | Permalink

    Re #29 Good one!

    Minor point: On a different note, I looked at Emanuel’s slide #23, wondering why his presentation included a chart of the temperature of the lower stratosphere. The common view of hurricanes is that they extend as high as the tropopause, and not to the stratosphere (though they can “bump” and distort the tropopause locally, raising it a bit). Also, the great majority of air outflow is below the tropopause. It seems like the better measure of a storm’s potential “heat sink” is the upper troposphere temperature, not the lower stratosphere.

    I guess that if Emanuel had used, say the 200mb upper-troposphere temperature, then he would have gotten a no-trend line (temperature hasn’t cooled in recent decades), which would hurt his attempts in slide #24. But, by choosing the lower stratospheric temperature, he gets to help his slide #24 and, perhaps as a small bonus, can put more onus on humans because the dropping stratospheric temperatures are credited to human activity (CFC, greenhouse, etc).

    More interesting point: In looking at atmospheric temperature profiles, I ran several for the tropical upper atmosphere. Here is the temperature plot for the lower stratosphere (100mb surface). This is radiosonde (weather-balloon) data. Two points:
    1. There is a remarkable spike in the temperature circa 1976. I continue to marvel at whatever it was that happened to the climate in the mid-1970s. Its fingerprints are all over the place, sometimes in the oddest corners of the atmosphere. (If I were a gradute student in climatology, I’d be digging everywhere to understand what happened in the mid-1970s.)
    2. As mentioned, CFCs and GHGs are generally credited for the drop in lower-stratospheric temperatures over the last twenty-five years. This chart makes me wonder if, as an alternative, recent temperatures are simply recovering from the 1976 temperature spike.

    Out of curiousity I looked at slide #23 to see if, by chance, Emanuel explained what happened in the mid-1970s. What I found, oddly, is that the slide uses MSU (satellite-derived) temperatures from 1980 forward but uses “predicted temperature” for 1950 to 1980. Why use “predicted” values into the past, when radiosonde readings exist and seem to correlate with MSU readings? And what assumptions are used to make these “predictions” about the past?

    By using “predicted” values for 1950-1980, and not radiosonde readings, the mid-1970s lower-troposphere spike is purged from the record. Perhaps the spike is, to coin a word, inconvenient.

    Maybe there is good reason to discount high-altitude radiosonde temperature data. If anyone knows why the radiosonde temperature record for the lower stratosphere in the 1970s should be purged from the record, please post. But, Emanuel has no problem using 1880s storm count and intensity data elsewhere in the presentation, so why are his standards higher now?

  31. BKC
    Posted Nov 10, 2006 at 4:32 PM | Permalink


    2. As mentioned, CFCs and GHGs are generally credited for the drop in lower-stratospheric temperatures over the last twenty-five years. This chart makes me wonder if, as an alternative, recent temperatures are simply recovering from the 1976 temperature spike.

    Very interesting. Isn’t the cooling stratosphere one of the planks for GHG forced AGW? If the data is (are?) accurate, it certainly doesn’t look like the cooling is caused by consistently increasing CO2.

  32. bender
    Posted Nov 10, 2006 at 8:47 PM | Permalink

    If the data is (are?) accurate

    “Data” is plural. “Datum” is singular. The data are accurate.

  33. Posted Nov 11, 2006 at 3:15 AM | Permalink

    A new article in Science on Hurricane data gives the following comment:

    New analyses show that most storm records have been skewed, producing the impression that tropical cyclones have been getting stronger globally. But reanalyzed records from the Atlantic Ocean going back to 1983 still show a sharp increase in hurricane intensity as tropical Atlantic waters warmed.

    From the full article:

    That pattern [of the Atlantic] supports a record that meteorologist and hurricane specialist Kerry Emanuel of the Massachusetts Institute of Technology published in 2005 (Science, 16 September 2005, p. 1807). Agreement was also very good in the eastern North Pacific, where energy release declined 60%. These two trends are well-supported, said Kossin.

    Then again, there was “not so good news everywhere else,” Kossin noted. Where the standard records from the northern Indian Ocean, the southern Indian Ocean, the western North Pacific, and the South Pacific showed rising trends of intensity, the reanalysis showed modest declines or no trend at all. And 85% of the world’s tropical cyclones occur in these ocean basins. Outside the Atlantic, Kossin concluded, storms show no signs of intensifying as the underlying waters warm, at least in the past 23 years.

  34. bender
    Posted Nov 11, 2006 at 4:45 AM | Permalink

    Re #33 “Global Warming May Be Homing In on Atlantic Hurricanes”

    And by what mechanism does this, er, “homing in” occur? [First the statistics, next the linguistics. Keep it up.]

  35. Nicholas
    Posted Nov 11, 2006 at 5:13 AM | Permalink

    Uh, am I reading #33 right – Kossin thinks it’s “bad news” that hurricanes are NOT increasing in most basins?

    I guess it may be bad news for his pet theory but it’s good news for the people who live in coastal areas.

  36. David Smith
    Posted Nov 11, 2006 at 7:41 AM | Permalink

    Re #33 I look forward to reading that article.

    A “60% decrease in energy released” in the Eastern Pacific due to SST?? That seems to mean that sea surface temperatures have declined in the East Pacific. But, here is the sea surface temperature plot for the Eastern Pacific (EPAC) tropical region. Temperatures have risen, not dropped.

  37. David Smith
    Posted Nov 11, 2006 at 8:31 AM | Permalink

    Re #33

    About the only way Science could find an East Pacific SST decline is if they used Figure 1 from Webster Curry (Science, 2005). The problem is, that plot is a mistake.

    Hoyos Webster Curry apparently recognized their error, because the 2006 Hoyos Curry Webster Figure 1A (Science, 2006) corrects that error for the East Pacific.

    I wonder if the Science writer used the plot that had the apparent error.

    (If anyone has a different interpretation of the plots or sees that I’m misreading or misunderstanding something, please post.)

    Here are the papers’ addresses:

    Click to access Webster2006c.pdf

  38. David Smith
    Posted Nov 11, 2006 at 9:07 AM | Permalink

    If there is a SST/hurricane connection in the Atlantic, my hypothesis is that it has more to do with geography than thermodynamics.

    I’ve been looking at NCEP data and see what I believe is a northward-shift in the Intertropical Convergence Zone (ITCZ) since 1995. The ITCZ is a line of thunderstorms and moist air, which are helpful in hurricane formation. When it shifts northward, it does three things. One, the coriolis effect is enhanced, which helps seedlings “spin”. Two, the ITCZ interacts more-strongly with the easterly waves and helps them develop. Three, fewer early-season seedlings “crash” into northern South America (land weakens or destroys them) because they are farther north and can travel through the waters of the Caribbean.

    My review of NCEP data indicates that the mean summer location of the ITCZ has shifted from 7N to about 10N over the last ten years. This is based on meridonal wind and pressure shifts. There may be a better way to locate the ITCZ, but that’s all I know about.

    The location of the ITCZ generally corresponds with the location of the highest SST, so as the North Atlantic has warmed relative to the South Atlantic, the ITCZ has moved northward, due to SST. I suspect there is also another reason, related to pressures north of the tropics, but that’s another topic.

  39. Howard Wiseman
    Posted Nov 11, 2006 at 6:21 PM | Permalink

    Dear Willis, Re #18. I agree that thermal transfers from the TC heat engine occur mostly at tropical latitudes. This probably continues at diminishing levels until the storm loses its warm core characteristics. The well vented TC with a true eye feature and anticyclonic circulation aloft will deliver its heat to tropospheric levels. Once extratropical transition occurs, the kinetic and baroclinic energy of the storm is delivered into the synoptic scale climate. I think this is what Carl Smith is saying in #20 as well. Your point regarding the source of my knowledge (mostly hearsay) is well taken and I intend to take a look at the literature on extratropical transition to see if anyone has tried to quantify these energy transfers.

  40. Willis Eschenbach
    Posted Nov 11, 2006 at 8:50 PM | Permalink

    Re 39, Howard, thank you for your comments. You say:

    Once extratropical transition occurs, the kinetic and baroclinic energy of the storm is delivered into the synoptic scale climate.

    Assuming this is true, we have to ask about the timescale of this transfer. If it is just picking energy up from the ocean and delivering it on the scale of hours to the “synoptic scale climate”, there’s very little net poleward transfer.

    I look forward to your further researches of the literature.


  41. David Smith
    Posted Nov 15, 2006 at 10:42 PM | Permalink

    Emanuel’s presentation includes a slide on lower stratospheric temperature. The chart shows a temperature decline, which is often presented as evidence of global (tropospheric) warming. Sounds interesting, so I decided to do some reading.

    First, some background. Stratospheric cooling has two main causes (see here for a nice summary). One cause is ozone depletion: ozone absorbs sunlight and converts it to IR, warming the vicinity. Less ozone = less warming. The reduction is ozone is generally attributed to human activities.
    The other cause is the increase in greenhouse gases. The reasoning is apparently rather complex, as the simplified explanations written for laymen vary quite a bit. Even Realclimate’s explanation, written by Working Climate Scientists, bounces around a bit. For my purposes here, I won’t get into the physics but rather will look for evidence of cooling.

    Now, some evidence. Here is the RSS satellite chart which shows “lower stratosphere” temperature over the last 25 years. One can see two volcanoes (warm spikes in the early 1980s and early 1990s) and what appears to be a downward slope. Oddly, there is little downward slant since the early 1990s. The cooling trend is neither dramatic nor steady.

    To attribute this weak cooling, particularly when it coincides with possible ozone-caused cooling,to AGW rather than a natural cycle, should require the proponent to show a robust understanding of the stratosphere. I am hoping that the readers here include such a person. Here are some questions which, in my opinion, would help show that robust understanding of the stratosphere.

    1. Why has the global lower stratospheric temperature, as measured by satellite and radiosondes, remained more-or-less steady for the last ten years, a period of increasing greenhouse gases?

    2. I looked at radiosonde data from NCEP. Here is the global temperature at 100 mb which is the lower part of the lower stratosphere. What happened in the mid-1970s to cause such a temperature spike, from which the stratosphere has now recovered? How much of the reported cooling of the lower troposphere is simply a recovery from theis (apparently) natural event in the 1970s?
    3. I looked at radiosonde day for global temperature at 70mb which is the “middle-upper stratosphere”. The shape approximates the most commonly-displayed satellite data, so I take it that 70 mb is representative of the lower stratosphere. I looked at the northern polar (65N to 90N) region ( (here) , to see the geographical pattern of cooling. What I see here, and at the Southern polar region, is a pattern of rising temperaure in the last 10 years . Why is that?

    4. What I see in the tropics at 70mb is a pattern of falling temperature in the last ten years. Why are the tropics different?

    5. When I look at the global 10mb temperature what I see is a decline, but I also see a lower stratospheric temperature back in the 1960s. Why is that? I thought the coldest temperatures are the recent temperatures.

    6. How do we untangle the greenhouse effects from the ozone effects?

    I have other questions but will stop here. If anyone can help provide answers, please do so. If I have flawed reasoning or understanding, please post.


  42. David Smith
    Posted Nov 16, 2006 at 10:06 AM | Permalink

    The October, 2006 NCDC (US) report on global climate, released today, is linked here .

    Looks like global warming continues to take a break.

  43. Steve Sadlov
    Posted Nov 16, 2006 at 11:45 AM | Permalink

    RE: #41 – one thing that happened between about 1967 and 1975 was the explosion in aircraft capable of flying above the tropopause. I am just old enough to remember the end of the older style. During much of the 1960s, other than certain transcontinental and transoceanic flights using 707s and DC-8s, they were still using Connies, DC-6/7s and the like. Especially for domestic, shorter flights. At best, they’d use a turboprop plane like an Electra or Convair (don’t recall the number). None of these planes were above the tropopause. Starting around ’67 we saw the introduction of 727s, DC-9s, and 737s, which replaced the older prop planes. And they were smoke belching straight turbojet propelled – we had yet to see widespread adoption of cleaner burning and more efficient turbofans. The turbofan planes were initially only wide bodies such as 747s, L1011s and DC-10s. Again, turbofans were only initially used for long haul, not for most domestic. Only after the mid 70s were 737s and DC-9s/MD-80s upgraded to have turbofans (727s were EOLed). Here is a short table to document what I am alluding to here:
    1910 – 1967: few planes above tropopause
    1967 – 1975: increasing number of dirty turbojet planes above tropopause
    1975 – 1983: slow adoption of cleaner turbofan planes above tropopause
    1983 – present: mostly cleaner turbofan planes above tropopause

  44. David Smith
    Posted Nov 28, 2006 at 9:11 PM | Permalink

    Emanuel focuses on anthropogenic factors (CO2, aerosols) as the drivers behind late-20’th century global warming. As a lukewarmer, I’m inclined to believe the warming has been due to a combination of anthropogenic and natural factors, and I don’t know which is bigger.

    The anthropogenic hypotheses are well-known, so I’ve been digging into possible natural oscillations.
    As usual, I have a mystery which is beyond my neural ability to solve. If anyone can explain it, please help.

    First, some background. The wintertime (Dec-Feb) sea surface temperature trend of the Indio-west Pacific warm pool is shown here . I have defined the region as 0 to 20N by 50E to 130E, based on my examination of maps in the literature.

    There is evidence (Quan et al, 2004) that the wintertime sea surface temperature in this region affects the strength and behavior of the associated Hadley cell. The strengthened wintertime Hadley cell then affects Northern Hemisphere westerlies in ways which keep frigid air from flowing to lower latitudes (= fewer cold blasts). There are additional effects on Northern Hemisphere weather, which I plan to write about this weekend. It is interesting stuff (at least for a stormhead like me) and explains a lot, in my opinion.

    Anyway, the mystery are the jumps in warm-pool sea surface temperature circa 1976 and circa 2000. I don’t have the ability to show a smoothed curve, so you’ll have to look at the unsmoothed data and imagine, say, 3-year averages.

    To my eyes, there are two jumps in the data, each followed by a (more or less) flat period (except for El Ninos and volcanoes).

    The years (1976 and 2000) are important. 1976 was when the globe switched from net cooling to net warming. Examination of satellite-derived tropospheric temperature shows a marked global temperature jump in 2000.

    What can cause such a SST jump, one that is sizeable enough to apparently affect Northern Hemisphere temperatures? If it is CO2/aerosols, I’d expect the temperature rise to be more or less relentless. If it is solar, I’d also expect smoother behavior. If it is weather pattern (windspeed, humidity, cloud cover, etc) I’d expect to see it in those factors, but those don’t seem to be apparent in the NCEP data.

    To my mind, that leaves ocean behavior (paths of currents, extent of upwelling, etc) as a suspect.

    Anyone have thoughts on how radiative forcing could cause such important jumpy behavior?

    Final note: the most interesting thing in Quan is the report that this key Hadley cell’s behavior changed in 1976, the same time that the globe switched from net cooling to net warming, especially in the Northern Hemisphere. Prior to 1976, the cell sent much of its outflow (heat) southward, into the Southern Hemisphere. After 1976, much of the outflow shifted into the Northern Hemisphere. Remarkable. Why the shift?

    Thanks for any help.

  45. Steve McIntyre
    Posted Nov 28, 2006 at 10:35 PM | Permalink

    David, you might be interested in looking at some of the Pacific proxy series – Fisher’s Mount Logan ice core series is hypothesized as indicating changes in zonal- and meridional- circulation. Newton et al 2006 plausibly interpret much proxy data as indicating N-S shifts in the ITCZ. Is it possible that there are micro-versions of such things involved here?

  46. Willis Eschenbach
    Posted Nov 29, 2006 at 3:13 AM | Permalink

    David, take a look on Google for “Pacific Decadal Oscillation”.


  47. David Smith
    Posted Dec 5, 2006 at 8:34 PM | Permalink

    Gray states that the warming over the last 35 years is due to natural (ocean-related) oscillations while Emanuel states that the warming is manmade (GHGs).

    For fun, I’m trying to understand what natural oscillations might cause decadal warming. Gray’s ocean hypotheses are hard for me to decipher, so I’m having to guess and look at all possibilities.

    Also, it’s a long time until hurricane season for a stormhead like me, so I’m looking for meteorological things to explore.

    One possibility is the Pacific Warm Pool and the related phenomena of El Nino / La Nina. The Pacific Warm Pool is a deep accumulation of warm water in the western Pacific. Pacific winds blow east-to-west at the equator, which pushes the warm surface water into the western Pacific. The warm water piles up in this area, forming the Warm Pool.

    Figure 1 , when “Assorted Plots” is clicked, shows a nice cross-section of the Pacific. Click on the chart to magnify it. Warm water is red while cool water is bluish. The western Pacific is on the left, and the accumulated (red) warm water can be seen there.

    The chart clearly shows the piled-up warm water in the western Pacific. It takes wind from the east to keep that warm water piled up and prevent it from spreading eastward.

    What happens if the easterly wind dies down? Well, that is “El Nino”. The wind no longer has the power to keep the warm water piled up in the western Pacific, so it spreads across the ocean surface towards the east. The anomalous warm water on the surface releases a lot of heat into the atmosphere, warming it and creating global temperature spikes like in 1998. (This also depletes some of the deep heat from the Warm Pool, but that’s another story.)

    There may also be a warming effect in addition to the direct release of ocean heat. This is due to the warmed air flowing poleward, which (due to the Coriolis effect) speeds up the upper winds in the temperate regions. The faster winds make it more difficult for cold polar air to push southward. My hypothesis is that the best removal of heat from Earth occurs when air mixes, bringing warm air to the Poles where there is good radiation while cool, dry air brings its lower humidity (water content) to the lower latitudes, aiding radiation and evaporation. El Nino slows this mixing, thus making heat removal from Earth less efficient.

    What happens if the east winds increase, piling up the warm water in the west even more? That is known as “La Nina”. The warm water covers less surface area and, importantly, cool water upwells in the eastern Pacific. The surface of the equatorial Pacific thus cools, reducing the heat flowing to the atmosphere. It also reduces the westerly winds in the temperate latitudes, which helps the mixing of cool and warm air, helping net heat removal from Earth (my opinion).

    So, an era of greater El Ninos and fewer La Ninas should tend to warm the earth’s surface. How have El Nino and La Nina behaved? Figure 2 shows the ENSO index, which is an indicator of El Nino and La Nina. As the chart shows, there appears to be a shift circa 1976 from predominately cool la Ninas to warm El Ninos.

    Conceivably, this may account for part of the surface warming since warming began in the mid-1970s.

    Now, an alternate explanation is that excess heat due to GHGs has been accumulating in the Warm Pool, triggering more-frequent releases in the latter half of the 20’th century. Maybe so, but the evidence is not conclusive on that. Also, the rather sudden change in frquency in the mid-1970s looks more like a natural mode shift rather than a gradual shift due to gradual accumulation of GHGs.

    The UN IPCC opinion is that the Warm Pool / ENSO connection to climate is complex (indeed) and that GHG forcing could manifest itself via changes in modes or frequencies of natural oscillations like ENSO.

    I presume that Dr. Gray’s hypothesis is that a predominantly El Nino mode, like the one we’re in, temporarily warms the atmosphere by releasing heat stored in the deep Pacific warm pool. The earth’s surface warms while the ocean’s Warm Pool heat is depleted. After some decades, the oscillation shifts to a La Nina mode, reducing the release of heat from the Warm Pool.

    It’d be interesting to see if there is some statistical way to correlate the ENSO index to 1980-2005 mid-tropospheric satellite-derived temperatures and then use the correlation to estimate ENSOs (La Nina’s) effects on global temperatures before 1976.

    My belief is that the shift to predominantly El Ninos may be a contributor to recent global warming but it is not a sole or dominant contributor.

    Next: a look at shifts in the Indo-Pacific Hadley Cell

  48. Steve Sadlov
    Posted Dec 6, 2006 at 10:13 AM | Permalink

    RE: #47 – There has been some debate out and about whether or not we switched to PDO cold phase in the late 90s / early 00s. In any case, based on my completely non scientific pneumonic of general coldness, length of winter (especially lateness of spring onset) and a few other things I look at, for the mid Pacific Coast of the US, things are more La Ninaesque now than they were 1976 – late 90s (or so).

  49. Steve Sadlov
    Posted Dec 6, 2006 at 10:18 AM | Permalink

    RE: #48 – A major breakpoint, by my reckoning, was how we went from a classic warm and very rainy El Nino pattern during early 1998 (endless days of rain, temperatures locked in bewtween 50 and 65 Deg F) to a completely opposite pattern, of a remarkable interior outbreak and widespread snow flurries down to sea level, on December 20, 1998, followed by a freeze lasting into January 1999, eclipsed only by the 25 year freeze of late December 1991, here where I live and in nearby environs.

  50. Steve Sadlov
    Posted Dec 6, 2006 at 10:19 AM | Permalink

    RE: #49 – Sorry, the 25 year freeze was late Dec 1990, not late Dec 1991.

  51. jae
    Posted Dec 6, 2006 at 10:48 AM | Permalink

    David: If you look at Figure 17 in Jan Veizer’s paper on solar connections, you will note a close relationship between El Nino’s and increases in total solar irradiance and decreases in cosmic ray flux. Just chance? I doubt it

  52. Steve Sadlov
    Posted Dec 6, 2006 at 11:27 AM | Permalink

    RE: #51 – Causitive might be overall increase in convection events in East Pacific, due to increased cloud formation (possibly aided by uptick in cosmic radiation) impacting the trades and hence El Nino? Who knows ….

  53. David Smith
    Posted Dec 6, 2006 at 7:04 PM | Permalink

    Re #48 The more I read, the more I’m of the opinion that we haven’t identified all the major climate oscillations. Something happened in 1976 that was much more than a PDO shift. I call it a Climatological Earthquake, which affected many facets of the atmosphere.

    Once I master the procedure for posting graphs, I’ll post examples of the mid-1970s changes.

    One example I’ll describe in detail soon is the 1976 shift in the Indo-Pacific Hadley cell. Before 1976 it sent much of its outflow (heat) into the Southern Hemisphere but, after 1976, the heat flowed northward. To me, this is important stuff, but it’s buried in an obscure technical article.

    I wish that studies of natural oscillations had the same “sex appeal” as AGW studies, but they don’t, so they get little attention.

  54. David Smith
    Posted Dec 6, 2006 at 7:07 PM | Permalink

    Re #51, #52 I have wondered if changes in solar radiation affect the stratospheric circulation. Our understanding of how the stratosphere works, and how it affects the tropospheric decadal weather patterns (= climate), is poor.

  55. Ken Fritsch
    Posted Dec 6, 2006 at 7:31 PM | Permalink

    I wish that studies of natural oscillations had the same “sex appeal” as AGW studies, but they don’t, so they get little attention.

    I know that this is the proper thread on which to post your climate oscillation tutorials, but older and lesser used threads can have added posts that I’ll miss if I do not stay up to date. Therefore, I wanted to once again let you know that I enjoy and appreciate the material you have posted on climate oscillations and your personal evolving theories on the subject. It may or may not be “sexy”, but I find it good reading and another reason for my tip jar contribution being a real bargain.

  56. Posted Dec 6, 2006 at 8:41 PM | Permalink

    #54 David:

    Re #51, #52 I have wondered if changes in solar radiation affect the stratospheric circulation. Our understanding of how the stratosphere works, and how it affects the tropospheric decadal weather patterns (= climate), is poor.

    Have you read the paper:

    Sunspots, the QBO, and the Stratosphere in the North Polar Region – 20 Years later” (1MB PDF file)

    We have shown in earlier studies the size of the changes in the lower stratosphere which can be attributed to the 11-year sunspot cycle (SSC). We showed further that in order to detect the solar signal it is necessary to group the data according to the phase of the Quasi-Biennial Oscillation (QBO). Although this is valid throughout the year it was always obvious that the effect of the SSC and the QBO on the stratosphere was largest during the northern winters (January/February).
    Here we extend our first study (Labitzke 1987) by using additional data. Instead of 30 years of data, we now have 65 years. Results for the entire data set fully confim the early findings and suggest a significant effect of the SSC on the strenght of the stratospheric polar vortex and the mean meridional circulation.

  57. David Smith
    Posted Dec 6, 2006 at 9:31 PM | Permalink

    Ken, thanks, it’s a fun exercise. I’m simply glad that Steve M is kind enough to let laymen like us probe other aspects of the global warming puzzle on CA.

    When I look at a chart like temperature of tropical upper troposphere , where I see a giant spike in temperature in 1976, it raises my curiosity. What in the world happened then? Why do so few climate scientists seem intrigued by it?

    Carl, I have some familiarity with the QBO but have not read that paper, which I’ll do. Thanks for the link.

  58. jae
    Posted Dec 7, 2006 at 11:44 AM | Permalink

    57: There was a huge increase in solar irradiance, with consequent large decrease in cosmic rays in 1976. See Jan Veiser’s paper, Figure 14.

  59. jae
    Posted Dec 7, 2006 at 11:47 AM | Permalink

    56: Carl: That study is a great example of using out-of-sample data to support a hyypothesis. I don’t understand everything I read, but it’s pretty clear that the sun has a great effect on the oscillations.

  60. Steve Sadlov
    Posted Dec 7, 2006 at 11:48 AM | Permalink

    RE: #53 – I would reckon there to be substantial harmonics of the currently known oscillations, as well as resonances and “rogue wave” type events where interactions between them result in things like mega El Ninos and mega La Ninas. And also, to your point, what about oscillations encompassing multiple ocean basins, maybe even all of them.

  61. David Smith
    Posted Dec 16, 2006 at 8:48 AM | Permalink

    People are still scrambling to explain the fizzled 2006 Atlantic hurricane season.

    The NOAA climate summary recently blamed El Nino. A problem with that is that wind shear, the disruptive offspring of El Nino, did not arrive until late in the season. The fizzle was underway before El Nino played a major role.

    The other is Saharan dust. Yesterday Jeff masters wrote an easy-to-read article on Sahel dust and its hypothesized effect on hurricane formation. Basically, the dust absorbs sunlight and warms the middle atmosphere, which makes it stable and thus hard for rainclouds to form. There are problems matching the hypothesis with reality, which Masters touches in his article.

    Bottom line is that, despite the public appearance of “we know what happened”, I think that many experts are still scratching their heads in private.

    Interestingly, pre-1995 there was some correlation between Sahel rainfall in the fall and hurricanes in the following year. More autumn rainfell meant less dust, and less dust meant more hurricanes. That correlation broke down in 1995, when the Atlantic SST pattern changed. Gray continued to use it, though, which likely hurt his prognostications.

    Master’s article is here .

  62. David Smith
    Posted Dec 19, 2006 at 7:32 PM | Permalink

    Well, five months after it happened, the National Hurricane center has decided that was, probably, a 10’th tropical storm in 2006. The news release is here .

    This additional storm has no significance. But, someone tell me how one can compare 2006 storm count with, say 1885, 1925 or even 1965, times when detection capability was quite limited and nobody was going to spend time debating whether a mid-Atlantic thunderstorm cluster had spin and warm-enough temperatures.

  63. David Smith
    Posted Dec 19, 2006 at 7:36 PM | Permalink

    Re #62 This newly-classified tropical storm existed for (drumroll) a grand total of 18 hours.

  64. Steve Sadlov
    Posted Dec 20, 2006 at 10:37 AM | Permalink

    RE: #63 – And in the pre aircraft / electronic systems era, one must wonder how many slightly spinning thunderstorm clusters were never witnessed by any human. In the 1800s, while there was sea traffic, the population in the Americas was very sparse as compared with today, so one can imagine, boats at any given time across the Atlantic would have been few and far between. Any powered boat would have taken a northern route rather than a southern route especially going to the most concentrated population area between Boston and Newport News. There must have been many TSs that never were noticed prior to the early 1900s.

  65. Steve Sadlov
    Posted Dec 22, 2006 at 2:20 PM | Permalink

    Well, well, well …. what do we have here:

    Curry and Pielke Jr. debating – very telling

  66. jae
    Posted Dec 22, 2006 at 3:12 PM | Permalink

    65: An interesting discussion there. Sounds kind of familiar. Some people have a gard time admitting to mistakes, even unimportant ones.

  67. David Smith
    Posted Dec 26, 2006 at 10:16 AM | Permalink

    Re #65 The Pielke-Curry exchange shows that atmospheric science is still in its infancy, in more ways than one.

  68. bender
    Posted Dec 26, 2006 at 11:49 AM | Permalink

    Re #67 Heh heh. I didn’t want to say it myself, but you took the words right out of my mouth – as you often do.
    May scientific truth – and the quest for it – prevail in 2007.

  69. Michael Jankowski
    Posted Dec 26, 2006 at 11:59 AM | Permalink


    Good stuff! I had read much of the early exchange but forgotten about it over the holidays.

  70. Steve McIntyre
    Posted Dec 26, 2006 at 8:40 PM | Permalink

    IT has been possible to make reasonable projections of final 2006 hurricane statistics for some time – I can say this because I’ve done so. It’s not a particular specialty of mine and it didn’t take me long to do this from scratch. Nevertheless Holland and Webster failed to included 2006 data in their results. When the hurricane statistics went the other way in 2005, Emanuel included 2005 statistics in Dec 2005 in his reply to Landsea. This time last year, Webster and AGU reported a marvelous otherwise-unreported European prediction for the 2005 hurricane season. But no one’s seen the corresponding “prediction” for the 2006 season.

    In business, you’d be expected to reporting 2006 results in Dec 2006 – not 2005 results. But hey, this is climate science.

  71. David Smith
    Posted Dec 26, 2006 at 9:04 PM | Permalink

    Steve M, bender or Willis, I wonder if you have any comments on Greg Holland’s conclusion in slide 4 of this presentation .

    This was, apparently, a presentation to the US Senate.

  72. David Smith
    Posted Dec 26, 2006 at 9:21 PM | Permalink

    Re #71 On Holland’s slide #4 I find his use of pre-1900 tropical storm count in an analysis with modern storm count to be poor technique.

    How many times in the 1890s did they use their 1890 satellite images to reanalyze a squally area and conclude that, yes indeed, that squally area was a tropical storm for 18 hours? This was done last week for 2006.

    In 1900 the weather people struggled to even keep track of hurricanes in settled regions (see the Galveston hurricane of 1900, which killed 8,000), much less detect and track storms far from shorelines. How many storms at sea were never detected?

    His comment on SST “explaining 60%” of variation in storm count is the one I hope that someone statistically adept will comment about. Others, including Webster I think, have seen little connection in global storm count and SST. I wonder if Holland is proposing that the Atlantic is special.

    His slide 12, second sentence, is a leap of logic in my opinion.

  73. bender
    Posted Dec 26, 2006 at 10:10 PM | Permalink

    Decadally averaged (9-y MA) SST might “explain” 60% of the variation in decadal hurricane frequency; but the question is what is the significance level on that r^2? Holland’s Neff (effective sample size) after correction for autocorrelation is probably far lower than he thinks. He’s got 156 points in his time-series. But I bet Neff is as low as 30 when you use a 9-year running average. Willis is a whiz at digitizing the data and he’s got the formula for making that calculation. Good eye, David.

  74. bender
    Posted Dec 26, 2006 at 10:22 PM | Permalink

    As for his slide 12:

    The strong relationship between increases in storm and hurricane numbers and increases in SSTs leads to the inescapable conclusion that the majority of current hurricane activity is a direct result of greenhouse warming.

    I would invite Dr. Holland to come to CA to clarify and try to defend this statement. I will not say that it is entirely wrong; but it isn’t quite right either. It is ambiguous and it is highly presumptuous – to the point of being utterly misleading. Dr. Curry, too, would be welcome to try to defend it. I’ll debate her on it.

  75. Steve Bloom
    Posted Dec 26, 2006 at 10:55 PM | Permalink

    Re #70: There are rumors that the Euro modelers will be going public this spring.

    Re #71: The presentation you linked to was from the October AMS policy seminar (series archives here) and wasn’t sponsored by the Senate, although I would assume there was attendance by science committee staffers. The graphic you’re wondering about was taken from the new Holland/Webster paper, which is still available on Holland’s site.

  76. bender
    Posted Dec 27, 2006 at 4:12 AM | Permalink

    Re #70: There are rumors that the Euro modelers will be going public this spring.

    And there are rumors they won’t. Shows you what rumors are worth. I’ll believe it when I see it.

  77. Posted Dec 27, 2006 at 4:18 AM | Permalink

    There are rumors that James Hansen predicted a “super El Nino”. But then again maybe he didn’t.

  78. bender
    Posted Dec 27, 2006 at 5:11 AM | Permalink

    Re #75

    The graphic you’re wondering about was taken from the new Holland/Webster paper, which is still available on Holland’s site.

    Here is the “paper”. It is not actually a published “paper”, but a submitted manuscript. And it has flaws, such as the one I pointed to in #73. On p 13:

    SST anomalies explain over 60% of the tropical cyclone variance since 1905

    No reviewer should let statements on statistical association pass without a statement of confidence (p value) and full disclosure of the methods of analysis. And no time-series correlations whould ever be reported without having done a sensitivity analysis on the time-points (to avoid cherry-picking dates that favor the hypothesis). Why 1905? What happens if you pick 1906, or 1910, and so on?

    Figure 4 is comical. The hypothesis here is that a step function fits the hurricane proportion data better than a smooth oscillatory function. Where is the hypothesis test? How were the parameters of the step function chosen? Where are the methods? Malcolm Hughes would give this figure a failing C grade if it were the product of a graduate student.

    In discussing Fig 4 I was glad to see the reference to Lorenz on p. 9:

    This sharp transition from one climatic state to another is a feature of non-linear time series (Lorenz 1963)

    But this statement, as written, is false. Non-linear time-series need not exhibit state-switching behavior. It is a peculiar feature of the highly non-linear Lorenz “butterfly” attractor.

    In short, this paper has a ways to go before being acceptable for publication. What other mistakes/oversights does it contain?

  79. lucia
    Posted Dec 27, 2006 at 8:12 AM | Permalink

    Does anyone know where the underlying data for those plots may be found?

  80. David Smith
    Posted Dec 27, 2006 at 8:33 AM | Permalink

    I’ve only scanned the Holland/Webster paper and don’t have time to read it in detail until later this week, but something did catch my eye.

    Figure 3b is a time plot of hurricanes (the blue line) as a proportion of all storms. From 1900 forward the proportion looks trendless to me, but the pre-1900 proportion is noticeably higher. Why is that?

    Well, in the text (Section 3.1) they seem to dismiss this with the words, “The higher proportion of hurricanes in the early record (Fig 3b) was probably due to a tendency to miss weaker systems in earlier years.” I certainly agree with that.

    So, there is an acknowledgement that weaker earlier storms were likely missed. Nevertheless, they seem to go ahead and use that problemed historical data in their key Figure 1!

    Suppose that we adjust the number of pre-1900 storms upwards, so as to make the hurricane proportion the same as post-1900. That adjustment would add about 20% to the pre-1900 storm count (eyeball estimate).

    Now, use that +20% to adjust Figure 1, so that it is apples-to-apples. that would bring the average annual storm count to about 9.5, which looks very close to the 10 average of the period 1935-1995. Basically, no step change. Worse that that for their argument is that SST stepped upwards around 1920. So, SST rose while the storm count stayed constant.

    Perhaps there is an explanation, which I’ll find when I read the entire report.

  81. Judith Curry
    Posted Dec 27, 2006 at 8:49 AM | Permalink

    A few comments on issues raised re the Holland paper and ppt presentation. I appreciate the more substantive discussion of Holland’s paper that is appearing at climateaudit (rather than the sophistry over irrelevant issues over at prometheus).

    1. The North Atlantic is the only basin where the number of tropical cyclones is increasing. Globally, the number of tropical cyclones remains the same (implying a slight decrease outside the North Atlantic). Bill Gray has been saying for decades that the Atlantic tropical cyclones show different behavior from those in other basins, so yes the Atlantic is “special” when it comes to tropical cyclones. The number of NATL tropical cyclones (when filtered for el nino and other short term variability) does show a correlation with SST. Webster has done the calculation for 5 year running means, and the correlation between NATL SST and TC counts from 1910-2005 gives a correlation of 0.78 that is significant using the Mann Kendall statistic (this is a curry/webster paper that is in press, to be published around June 07). If Steve M, Bender, or Willis can provide any further insights into the statistical significance of this relationship, I would certainly be interested in pondering it and would certainly forward to Holland/Webster. However, when data is limited (in terms of amount and quality), an absence of statistical significance doesn’t mean we should ignore the relationship between SST and the number of tropical cyclones. Scientists use such information to point them in the direction of trying to figure out causal mechanisms between the two variables. The highest resolution climate model simulation run by Oouchi et al. found an increase in the number of NATL TCs in a doubled CO2 scenario (although the rate of increase with SST was only 20% of that inferred from the historical NATL data (note Oouchi et al. found that the number of TCs outside the NATL to decrease slightly). So apparently there is something in the coupled atmosphere/ocean system that causes the number of NATL TCs to increase with SST. An important issue to understand, and several groups are trying to understand this relationship in the context of physical mechanisms, statistical analysis, and model simulations. While i agree that robust statistical analysis is important (and that our field is often too sloppy in this area), none of our arguments stand or fall solely based upon statistical analyses of imperfect data.

    2. Re the European TC forecasts, a paper has been submitted (Palmer, Vitart et al.) that is under review (last time I heard). They do plan to go operational for the 2007 season

    3. Re Holland’s statement in the ppt presentation:
    The strong relationship between increases in storm and hurricane numbers and increases in SSTs leads to the inescapable conclusion that the majority of current hurricane activity is a direct result of greenhouse warming.
    This is a short cut version of the argument presented in our BAMS article, which rested on the causal chain:
    1. observation of increased hurricane activity (global intensity and NATL number)
    2. statistical link of global intensity and NATL number with SST (see Hoyos et al. for global intensity link with SST)
    3. attribution (through climate model simulations) of increased SST (at least since about 1970) to greenhouse warming

    The issue of data quality is rather overwhelming outside the NATL, so recent attention has focused on the NATL where the data quality is more credible. The “inescapable” word in the ppt presentation is too strong, but .ppt presentations don’t tell the entire story of what was actually said (I wasn’t there so i don’t know what was actually said).

    4. Re 2006 data. The Holland manuscript was submitted last summer and accepted for publication before the end of the TC season. The TC data is still under revision (note a 10th storm was recently added), but arguably good enough to be included at this point. However, the 2006 SST data (at least in a form that is consistent with the rest of the climatological data set) is not yet available.

  82. Paul Linsay
    Posted Dec 27, 2006 at 9:00 AM | Permalink


    But this statement, as written, is false. Non-linear time-series need not exhibit state-switching behavior. It is a peculiar feature of the highly non-linear Lorenz “butterfly” attractor.

    Actually it’s quite common in non-linear systems. The atmosphere displays it too, for example the PDO shift in 1976, and, I’d contend, the global temperature as measured by the satellites, flat before the 1998 El Nino, then a step up afterwards. We probably need more time to see if it’s really a step up, but it looks like it. Needless to say, I don’t buy into the linear trendline.

  83. Judith Curry
    Posted Dec 27, 2006 at 10:11 AM | Permalink

    I am going through the holland paper now in some detail. Re statistical significance testing etc., in section 3 the paper states:

    Spectral analysis of the tropical cyclone database (not shown) reveals distinct peaks at 2-3, 5, 9 year and multi-decadal time periods. As longer period variations and trends are the focus of this paper, we require the removal of short-term variations, such as associated with the El Nino-Southern Oscillation (ENSO) and quasi-biennial oscillation (QBO), which are both of order 2-3 years in period. We tested filters and averaging periods between 3 and 15 years. The averaging showed considerable sensitivity to smoothing period out to 5 years but relatively stable at longer averaging periods. All data are therefore smoothed by a simple 9-y running mean.
    Standard statistical tests are used throughout. Unless otherwise indicated all variance and correlation coefficients are valid at the 95% level using the t-statistic with correction for serial correlation (Davis 1976).

  84. Dave Dardinger
    Posted Dec 27, 2006 at 10:27 AM | Permalink

    Re: #81 Dr. Curry,

    However, when data is limited (in terms of amount and quality), an absence of statistical significance doesn’t mean we should ignore the relationship between SST and the number of tropical cyclones. Scientists use such information to point them in the direction of trying to figure out causal mechanisms between the two variables.

    Maybe you’re using “absence of statistical significance” differently that I’ve seen it used elsewhere, but I thought the entire purpose of demanding statistical significance was to allow people to ignore relationships which may have occurred purely by chance. I suppose one could use near significance as a guide for further research, but surely not as evidence for a theory. And even then, I’d think there are plenty of other pointers which could be persued than looking to ones which lack statistical significance.

    I think I see a bit of “bias” creeping in, as researchers comb through data searching for something which will butress their existing ideas. But this means that other pointers, which would/could point in the opposite direction are ipso facto being ignored. This is why the sceptics are so worried about claims of “concensus.” It’s the old story of “to a person with a hammer, everything looks like a nail.”

  85. Steve McIntyre
    Posted Dec 27, 2006 at 10:34 AM | Permalink

    I’ve put up a new thread here for this paper Let’s use a new thread.

%d bloggers like this: