The Intersection

i-26a8c4c69bb6e36898b270ed42aaa626-3 storms.jpg

As I noted yesterday, a very important paper (PDF) has just come out on hurricanes and global warming, by Jim Kossin of the University of Wisconsin-Madison’s Cooperative Institute for Meteorological Satellite Studies and his colleagues. The paper was published in Geophysical Research Letters. Here’s how the University of Wisconsin-Madison’s press release describes Kossin’s results:

HEADLINE: New evidence that global warming fuels stronger Atlantic hurricanes

MADISON — Atmospheric scientists have uncovered fresh evidence to support the hotly debated theory that global warming has contributed to the emergence of stronger hurricanes in the Atlantic Ocean.

The unsettling trend is confined to the Atlantic, however, and does not hold up in any of the world’s other oceans, researchers have also found….

Now look at how these same results are described by World Climate Report:

Global hurricane intensity not increasing…so concludes a just-published paper by University of Wisconsin atmospheric scientist Jim Kossin and colleagues. In order that we can’t be accused of misrepresenting the authors’ meaning, here is the complete conclusion section of their Geophysical Research Letters paper [LONG QUOTATION FROM THE PAPER FOLLOWS]

All we can say is “oh my.” Can it be possible that the small band of global warming alarmists who is going around pushing the concept that global warming has led to measurably more dangerous hurricanes is wrong? Horrors.

Over at Prometheus, meanwhile, Roger Pielke, Jr., voices his view that the original Wisconsin-Madison press release is guilty of “completely misrepresenting the science in the paper that it is presenting,” which has led to some interesting exchanges.

Because I discuss this paper in detail in Storm World and don’t wish to steal my own thunder, I’m going to remain mostly mum for now about my own view of what’s happening here. However, I will make a few fairly obvious points.

First, because the new paper’s satellite-based reanalysis suggests that hurricane intensity is indeed increasing in the Atlantic but not globally, it’s very easy to selectively “frame” this result to support widely disparate interpretations. All you have to do is pick which of these two conclusions to lead with.

Second, it’s worth recognizing that this is not the first time these results were made public. (I could hardly have covered them in Storm World if they’d only come out just now.) In fact, Kossin gave a talk about his findings back in October (PDF). A key to reconciling the two different “spins” on these findings can be found in his conclusions there:

Similar warming trends are found everywhere in the tropics. Why is the Atlantic behaving so differently?

If the data are not good enough to accurately measure long-term hurricane behavior, then our path to understanding how hurricanes will change in a warming world must be through better physical understanding. This is our present research challenge.

In the University of Wisconsin press release, Kossin goes beyond the present paper and tries to do just that–add some “physical understanding”:

Sea-surface temperatures may be one reason why greenhouse gases are exacting a unique toll on the Atlantic Ocean, says Kossin. Hurricanes need temperatures of around 27 degrees Celsius (81 degrees Fahrenheit) to gather steam. On average, the Atlantic’s surface is slightly colder than that but other oceans, such as the Western Pacific, are naturally much warmer.

“The average conditions in the Atlantic at any given time are just on the cusp of what it takes for a hurricane to form,” says Kossin. ” So it might be that imposing only a small (man-made) change in conditions, creates a much better chance of having a hurricane.”

The Atlantic is also unique in that all the physical variables that converge to form hurricanes — including wind speeds, wind directions and temperatures — mysteriously feed off each other in ways that only make conditions more ripe for a storm. But scientists don’t really understand why, Kossin adds.

Now, Prometheus is right to observe that this kind of stuff is not in Kossin’s latest paper. But I suspect that it may be found in other work of his, such as this paper listed on Kossin’s website: “Kossin, J. P., and D. J. Vimont, 2007: A more general framework for understanding Atlantic hurricane variability and trends. Bull. Amer. Meteor. Soc., in review.” In any event, I’ll leave you with this question: Is it really wrong for a press release to go beyond a single study to provided a broader interpretation of that study’s results? Don’t we want scientists to contextualize their findings?

P.S.: Left this additional comment at Prometheus:

It seems to me that one issue raised by Roger’s post is whether it’s appropriate for a press release to go beyond the findings of the latest study so as to provide a broader context for interpreting those findings, perhaps drawing upon other research to do so. Press releases would be sterile indeed if they had to be purely technocratic recaps of the latest findings, and nothing more.

P.P.S.: Real Climate has now done a post and opened a thread on the Kossin study. I’m sure there will be much more valuable commentary over there, so for anyone who wants more depth…..

UPDATE: Indeed, Jim Kossin, Kevin Trenberth, and Judith Curry are discussing the paper over at RealClimate now, so while this thread provides some good one-stop shopping about press releases and spin, and Prometheus provides more, that’s where the scientific dialogue is now happening. I might add that this is a real achievement on RealClimate’s part…a few years ago most scientists wouldn’t have been caught dead on the blogs. Now, it’s almost routine for them to comment there.

Comments

  1. #1 Roger Pielke, Jr.
    March 1, 2007

    Hi Chris-

    You’ve got this just about right. Three quick comments:

    1. Context and accuracy are not trade-offs. If the press release was indeed referring to Kossin’s other work (which I doubt, but lets just assume it was), then it should simply have said so, right?

    2. The press release, as written, misrepresents the paper, especially the headline and open paragraphs. There isn’t a single scientist who questions whether the North Atlantic has become more active over the past 3 decades … to suggest that Kossin provides “fresh evidence” is just wrong. Further, Kossin’s paper deals with the statistics of trends and says absolutely nothing about attribution (it even says that 23 years is too short for attribution!).

    3. The most important conclusion of this paper is that it completely contradicts Webster et al. 2005 around the world. I know you know this, but reporters reading the press release easily would not.

    For someone who parses language very closely when President Bush is speaking, you are giving a lot of leeway on this issue, which in my view is a slam dunk. It is a badly written press release that mischaracterizes the paper it is promoting. Period. Why not simply admit that and move on? Whatever one’s views of hurricanes global warming or political preferences, surely we should all agree that the public communication of science by universities and the NSF should not mischaracterize the work being represented (?)

    Finally, let me note that Jim Kossin is a great researcher and this is an important paper.

  2. #2 Chris Mooney
    March 1, 2007

    Roger,

    By centally highlighting the support for a trend in the Atlantic rather than the lack of support for a trend everywhere else, I would agree that the press release downplays the finding from the paper that is the most salient for those of us following the hurricane-global warming debate–namely, this paper challenges the notion that there is a global trend toward more intense hurricanes. So we’re in agreement there (your point 3).

    We’re also in agreement on your point 1.

    However, we may diverge on whether “misrepresentation” is really the right word. I would go with something closer to “selective representation” or “selective framing.”

    The study does present “fresh evidence” in the sense that it’s a new dataset confirming an older dataset. And it does show an Atlantic trend. The question is whether showing such a trend can be “support” for the Atlantic hurricane-GW hypothesis even if no formal “attribution” is done….

  3. #3 Chris Mooney
    March 1, 2007

    To add, I think the real point is that this study is kind of a Rorschach. Let’s turn the tables and look at how World Climate Report interpreted it: “Can it be possible that the small band of global warming alarmists who is going around pushing the concept that global warming has led to measurably more dangerous hurricanes is wrong? Horrors.”

    I am guessing you wouldn’t agree with that interpretation either…

  4. #4 Roger Pielke, Jr.
    March 1, 2007

    Chris- The misrepresentation hinges on this question that you raise:

    “The question is whether showing such a trend can be “support” for the Atlantic hurricane-GW hypothesis even if no formal “attribution” is done….”

    I know that you know the answer to this, so why be so coy? Kossin et al. also answer this question explicitly in their paper. It is not selective to represent a paper as saying something that it does not (when in fact it actually says the opposite, i.e., that it offers nothing new on attribution).

    Thanks!

  5. #5 Fred Bortz
    March 1, 2007

    I just want to say thanks to Chris and Roger for having this discussion in public, where we all can benefit from your views and perspectives.

  6. #6 Roger Pielke, Jr.
    March 1, 2007

    Chris- On the ink blot test … absolutely. This is the same dynamic that Sarewitz explains as the “excess of objectivity”. I’d fully expect advocacy groups like World Climate Report and DeSmogBlog to put their respective spins on the paper. Were it only those groups interpreting the paper I wouldn’t have raised the issue. From where I sit is another thing altogether when it is the University of Wisconsin and NSF engaging in misrepresentation.

  7. #7 Chip Knappenberger
    March 1, 2007

    Hold on a second here…do either Chris or Roger think what we wrote at http://www.worldclimatereport.com (quoted above) is misrepresenting the implications of Kossin et al.’s GRL paper?

    Again, here is the quote:

    “Can it be possible that the small band of global warming alarmists who is going around pushing the concept that global warming has led to measurably more dangerous hurricanes is wrong? Horrors.”

    Reading beyond some literary flourishes, where have we even come close to misrepresenting the findings of Kossin et al.? Is it not true that there are some people (scientists included) that push the notion that anthropogenic global warming has already led to a detectable increase in the intensity of strong hurricanes? And don’t the results of Kossin et al. GRL 2007 suggest that there has not been a detectable increase in the intensity of strong hurricanes, when aggregated globally? So, might not the folks claiming that “global warming has led to measurably more dangerous hurricanes” be wrong? Don’t the Kossin et al. 2007 results at least suggest that possibility?

    Where is the misrepresentation?

    -Chip

  8. #8 The Science Pundit
    March 1, 2007

    For someone who parses language very closely when President Bush is speaking, you are giving a lot of leeway on this issue, which in my view is a slam dunk.

    I realize this is a complete aside, but what exactly is wrong with holding POTUS to a higher standard?

  9. #9 Fred Bortz
    March 1, 2007

    Oh, come on, Chip!

    You call serious scientists who raise interesting questions a “small band of global warming alarmists,” and then you want us to consider that a “literary flourish.”

    I haven’t seen anyone claim that the evidence for relationships between global warming and hurricane strength has led to a firm conclusion. It isn’t alarmist to say that studies tend to support one interpretation over another. We all know the jury is still out here.

    We need to pay attention to the possibilities. That doesn’t mean those who do so are alarmists.

    As I read this particular report, the evidence supports (to a limited extent) a relationship between global warming and the strength of Atlantic tropical storms but finds no similar correlation in other ocean basins. If further research supports that conclusion, we will have some very interesting questions to ask about the phenomenon.

    Floridly yours,
    “Dr. Fred”

  10. #10 Chris Mooney
    March 1, 2007

    Roger,

    I really wasn’t trying to be coy. As you know, the IPCC also “supported” the notion of human-induced hurricane intensification without formal attribution….

    That said, I don’t particularly want to defend the press release; I too found its opening–and first sentence especially–surprising. Interestingly, the version on the Madison website has a byline, suggesting this was done by one person….but I don’t want to speculate on why it came out the way it did.

    In any event, although the IPCC report didn’t really reawaken much controversy in the hurricane-climate area, it looks like Kossin’s study may have done so.

    Chip–The study finds a trend in the Atlantic, no trend in other basins. You use it to question “alarmists” who think global warming has led to “measurably more dangerous hurricanes.” Well, as Fred Bortz points out, a lot of those “alarmists” are leading scientists who now think a human-induced trend is emerging in the Atlantic but not necessarily elsewhere (or at least, who acknowledge that the data elsewhere are far more disputed than the data for Atlantic).

    There are many ways to spin the Kossin study, but you can’t use it to debunk this more nuanced position, with which it is perfectly consistent. But by writing, “Can it be possible that the small band of global warming alarmists who is going around pushing the concept that global warming has led to measurably more dangerous hurricanes is wrong?” it seems to me that’s precisely what you’re trying to do.

  11. #11 Roger Pielke Jr.
    March 1, 2007

    Chip-

    I just had a look at your commentary. I agree with Chris about the “literary flourishes” but that aside, the following excerpt from your commentary is factually misleading:

    What about the high profile results from papers in Nature (Emanuel, 2005) and Science (Webster et al, 2005) that made a lot of noise that intense hurricanes were increasing the world over and that anthropogenic global warming was to blame? Well, according to the new results by Kossin et al., and as has been suggested by Landsea (2006) and others, it looks like they were the result of a reliance on data of poor quality data.

    Kossin et al. 2007 and Emanuel 2005 on the Atlantic (NATL) are completely consistent with one another. Landsea’s critique of Emanuel involves pre-1983 (i.e., pre-Kossin) NATL data, so Landsea’s comments are irrelevant here. It is true that Kossin et al. challenges Webster et al. and Emanuel in the Pacific, but by lumping in Emanuel 2005 without this explanation, you are suggesting something that isn’t the case.

    Thanks!

  12. #12 Benny
    March 1, 2007

    I thought Roger Pielke Jr. is a “political scientist” who never comments on the science, just the policy?

    Must be my mistake…..

  13. #13 Chip Knappenberger
    March 1, 2007

    Roger,

    The Landsea article I was referring to is this one:

    Landsea et al., 2006. Can we detect trends in extreme tropical cyclones? Science, 313, 452-454.

    (not, as I think you thought, his 2005 response to Emanuel)

    Would you care to re-evaluate your opinion of our article in light of that reference since that seemed to be the basis of your “factual misleading” finding?

    -Chip

    PS. I can’t remember the number of times that we have been referred to as part of “a small, but vocal band of skepctics,” so I thought it would be fun to turn it around. Apparently, it is OK for it to be applied to some groups, but not to others! :^)

  14. #14 Steve Bloom
    March 1, 2007

    Chip, considering that virtually every World Climate Report post not only contains at least one intentional distortion of the science but is often premised on one or more distortions, I have to say that it demonstrates considerable gall for you to accuse someone else of the same thing.

  15. #15 Roger Pielke, Jr.
    March 1, 2007

    Chip- Thanks … Landsea et al. 2006 assert that there is not trend in the Atlantic when the record is extended back to 1960. Landsea’s Figure 1b in his Nature reply to Emanuel shows a clear trend from the mid-1970s in the Atlantic.

    So I’ll stick with my assertion that you should have distinguished Emanuel’s NATL and WPAC analyses with respect to Kossin et al., rather than suggesting that Kossin et al. somehow completely impeaches Emanuel 2005.

    Thanks!

  16. #16 Fred Bortz
    March 1, 2007

    Chip, you are a vocal skeptic, right? So you might object to the use of the words “small” and “band,” but being skeptical and vocal are qualities I appreciate in my scientific colleagues. So why do you feel so insulted by the term “small, but vocal band of skepctics”?

    The equivalent to “small band of global warming alarmists” might be “small band of denialists” or more extremely, “small band of stubborn denialists.” Now those are insults, even if you deserve them (and I don’t know your work well enough to judge that).

    The “literary flourishes” in your article tell me that you want to feed denialism, not skepticism.

    And that’s all the squabbling I intend to do over language and who is insulting whom.

  17. #17 Eli Rabett
    March 1, 2007

    I think you guys need to figure out how NSF releases are generated. This is a great example of free fall speculation.

    NSF encourages grantees to send them Research and Education Highlights: (Taken from some boilerplate on my machine):

    Research and Education Highlights are a crisp summary (~150 words) with an interesting and informative image highlighting your NSF funded work. Include a title, list of authors with affiliation(s), an appropriate color image (avoid graphs), and an acknowledgement of support….The text and graphics should be at the level of a press release, explaining briefly and in non-technical language what has been accomplished and why it is significant.

    More at http://www.nsf.gov/about/career_opps/orientation/highlights.jsp

  18. #18 Roger Pielke, Jr.
    March 1, 2007

    Eli- Thanks … Jim Kossin explained in detail the origins of the UW and NSF press releases on the Tropical Storms list-serv. TS list rules prevent discussion of specific content shared there, but do know that I am well aware of the specifics in this case, and thus obviously comfortable with my views expressed on the releases. Thanks!

  19. #19 Eli Rabett
    March 1, 2007

    Roger, I presume you are referring to what Peter Webster said at your place:

    “Don’t you think it is the scientist him/herself who should determine if the paper was misprepresented? You are aware of emails about this issue.”

    and of course, it would be interesting in light of your comment above to reply to Dan Hughes who said:

    “I think that the author of the paper should have been closely involved with writing the ‘public consumption’ aspects of the press release. Maybe even actually writing it. At the least the author of the paper should have been given an opportunity to peer-review the press release.”

    and one of our greek friends who also commented

    “The whole topic of public affairs offices editing or blocking press releases has been well covered, but what about the flip side of the argument, which is ensuring that what information is put out by a Government Agency (in this case the NSF release) accurately portrays the information presented?

    NSF specifically chose an image of Katrina, and one that was clearly labeled as such (as you know there are hundreds of images that you could pick from). To me, someone was looking to accentuate and create an alarmist story. Yet nowhere in the paper is there a mention of Katrina.”

    On the other hand, your answer was essentially like the joke about the engineer and physicist in a balloon who shouted down to a fellow on the ground, asking where they were. 100 feet in the air was the response. The engineer turned to the physicist and said: Must be a mathematician. Why said the physicist. Well the answer is 100% correct and just as useless. Care to point to or quote the response in the list-serv???

  20. #20 Roger Pielke, Jr.
    March 1, 2007

    Eli- Nope, sorry (really) .. I won’t violate the list rules, which are there to protect exactly this type of exchange. So when I say that I am comfortable with my views expressed on this, you can either decide to trust me or not .. but we’ve interacted long enough for me to have confidence in your decision ;-)

    On the substance of the paper itself Real Climate takes a look at it here:

    http://www.realclimate.org/index.php/archives/2007/03/hurricane-heat/

  21. #21 Eli Rabett
    March 1, 2007

    Not quite the point Roger. If what Peter Webster clearly implied is true, you should point out that those folk who commented on your blog are barking up the wrong tree. You could do it by pointing out what the normal procedure is without betraying any confidence.

    Since the default assumption is that normal procedures were followed at NSF (and normal procedures at universities are about the same) it seems to me that anyone who claims that the press releases went well beyond what the PI was claiming and he never saw the press release has the strong burden of proof to show that. Ball’s in your court Roger.

    Moreover, the bit about being comfortable is wonderfully constructed. You ARE a mathematician, aren’t you?

  22. #22 Roger Pielke, Jr.
    March 1, 2007

    Eli- Thanks … I’ll simply point out that your line of argument is premised on a single “if” statement, and you can connect the dots from there. Reread the last sentence of my post.

    And FYI my focus was mainly applied math (probability, stats, PDE, numerical analysis, etc.), the useful kind;-) Though I will admit that my favorite class was non-Euclidian geometry. … Enough on this, the horse is dead … what else you got? ;-)

  23. #23 Benny
    March 1, 2007

    So it’s not just science policy AND science, but also science media relations which now falls under Pielke’s range of expertise. Sort of like one-stop-shopping for climate change punditry.

    Journalists, take notice.

  24. #24 Eli Rabett
    March 1, 2007

    Roger: A single if argument either is or it ain’t. Let us consider yours. It starts with:
    ———————————-
    “A press release put out by the University of Wisconsin today is a case of a press release completely misrepresenting the science in the paper that it is presenting.”
    ————————-
    Of course, the press release said
    ————————–
    “This new dataset is unlike anything that’s been done before,” says Kossin. “It’s going to serve a purpose as being the only globally consistent dataset around. The caveat of course, is that it only goes back to 1983.”

    Even so, it’s a good start. Once the NCDC researchers recalibrated the hurricane figures, Kossin took a fresh look at how the new numbers on hurricane strength correlate with records on warming ocean temperatures, a side effect of global warming.

    What he found both supported and contradicted previous findings. “The data says that the Atlantic has been trending upwards in hurricane intensity quite a bit,” says Kossin. “But the trends appear to be inflated or spurious everywhere else, meaning that we still can’t make any global statements.”
    ——————————
    Which is what the paper says.

    The press release does go beyond the paper in quoting Klossin’s speculation:
    ——————————
    “The average conditions in the Atlantic at any given time are just on the cusp of what it takes for a hurricane to form,” says Kossin. ” So it might be that imposing only a small (man-made) change in conditions, creates a much better chance of having a hurricane.”
    ——————————-
    But it is clear that the UWis press release does not misrepresent the paper or put words into the mouth of Klossin, et al. Your argument fails.

  25. #25 Roger Pielke, Jr.
    March 2, 2007

    Geez Eli, and I took you for a smart bunny;-) Yes, there are many correct statements in the press release. And there are also some that are completely wrong. Anyone who reads the first few paragraphs of the press releases can see the misrepresentation. Saying that it doesn’t exist is not a strong point for you! As this blog debate has devolved into semantics, the horse is really dead now … you are free to have the last word, as this one is mine;-)

  26. #26 Eli Rabett
    March 2, 2007

    Roger, thank you for allowing a mere hare the last word. While the temptation to point out that at worst a wise (me) reader would find some things in the press release that are not in the paper, but are the opinions of Dr. Kossin, is strong, being a Rabett of high principle I too will resist and let your statement end this once and for all.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.