Detecting and Attributing Hurricanes

[Stop Press! Chris Rapley says "in the long term, cities such as London may have to be relocated" - heard on R4 10pm news. There will be more, no doubt...]

Well, obviously enough detecting hurricanes isn't too hard. But the question exercising many people minds is *attribution*: has GW lead to an increase in hurricanes? Attributing individual hurricanes is difficult/impossible (the line taken by RC); we're talking about over all trends & stats.

JA (when not blogging about pet food) has a nice article about attribution here, though he means attribution in general. And there is a nice article about an interview with Judith Curry (top quote: William Gray is one of some "hurricane scientists who don’t know a lot about global climate").

But the point I wanted to make is that failure to attribute to GW does *not* amount to an attribution to anything else: natural cycles, etc etc. And failing to determine a human signal is not the same thing as ruling out a human signal (BTW, I'm not actually asserting that this "failure" has occurred). Which brings me to RPs beastly rough post Slouching Toward Scientific McCarthyism (also read the comments). He quotes "NOAA attributes this increased activity to natural occurring cycles in tropical climate patterns near the equator" (my emphasis). RP believes that disagreeing with "The increased activity since 1995 is due to natural fluctuations and cycles of hurricane activity" (again, my emphasis) is bizarre, since the statement is "fully supportable by peer-reviewed science". But is it? For it to have been so, there would have to be a proper attribution to natural causes. If there is any such papers, what are they?

[Minor updates: at least one cultured person has complained that my calling RPs post "beastly rough" was a bit over the top. They have missed the allusion, which I think RP originally intended in his title, to "Slouching to Bethlehem": to quote the second stanza:

A shape with lion body and the head of a man,
A gaze blank and pitiless as the sun,
Is moving its slow thighs, while all about it
Reel shadows of the indignant desert birds.
The darkness drops again; but now I know
That twenty centuries of stony sleep
were vexed to nightmare by a rocking cradle,
And what rough beast, its hour come round at last,
Slouches towards Bethlehem to be born?

I like that; and that was the bit I remembered. But it starts "Turning and turning in the widening gyre / The falcon cannot hear the falconer / Things fall apart; the centre cannot hold;" which fits a hurricanes discussion quite nicely :-)

And secondly, my somewhat flippant "detecting hurricanes is easy enough" is not really right; there are problems with detection in remote regions in the early days.]


Anonymous said...


You've misstated my views.

James Annan and I had a discussion about this on his blog, and I'll summarize here. If you accept the IPCC's definition of and criteria for detection and attribution, then the null hypothesis in all cases is indeed that observed climate behavior can be explained by natural processes.

James correctly pointed out that one need not start out with such a null hypothesis to which I agreed.

Nonetheless, this is the approach of the IPCC and it does set a rather high bar for attribution to be achieved. It also provides a default assumption for situations in which attribution has not been achieved.

Being a policy scientist, I do not have expertise relevant to taking sides in predicting where future TC research will go. I can however read the existing literature (which we summarized in our BAMS article). To date there are no peer-reviewed papers on tropical cyclones that meet the IPCC's threshold for attribution. Emanuel's may not even make the threshold for detection, though Webster et al. might. Both TC Emanuel and Webster et al. fully expect D&A to be achieved at some point in the future.

Soon there will undoubtedly be a body of papers offering competing theories to explain observed trends (e.g., Bell, Trenberth, Emanuel will each offer very different perspectives), and even then we may not reach the IPCC threshold for attribution.

Recent and forthcoming studies are indeed suggestive of a linkage. However, non-TC experts who decide to place their bets on one side or orther of this debate are doing so for non-scientific (e.g., personality, ideological) reasons.

And as I have said before, as interesting an issue as this is scientifically, it is not so important from a policy standpoint as the same policies make sense no matter who winds up winning this debate.

William M. Connolley said...

You're still making the same mistake, which probably has a clever latin tag about false or unfalse dichotomies or somesuch. Failure to dismiss the null hypothesis doesn't amount to attribution at all: that was the point I was trying to make.

Anonymous said...

William- Your beef is with the IPCC, not me. I've accurately presented their prespective on attribution.

William M. Connolley said...

Nope, you've failed to understand the meaning of not rejecting the null hypothesis. Have you actually read the IPCC D+A chapter? In fact, looking at that again, the entire defn of attribution is how-to-attribute-to-anthro. Attribution to natural causes isn't mentioned (in my brief skim).

Obviously some events *can* be attributed to natural causes: the T changes after Pinatubo, say. But hurricane trends? No.

For reasons of your own you've decided to come down hard on one side of this debate, whilst defending the "Gray" sides indefensible claims.

Anonymous said...


You arejust being onery, I am sure;-) But even so, here are some unambiguous quotes from the D&A chapter:

"Detection is the process of demonstrating that an observed change is significantly different (in a statistical sense) than can be explained by natural internal variability."

Natural variability *is* the null hypothesis.

" ... attribution of observed climate change to a given combination of human activity and natural influences requires another approach. This involves statistical analysis and the careful assessment of multiple lines of evidence to demonstrate, within a pre-specified margin of error, that the observed changes are:

* unlikely to be due entirely to internal variability;
* consistent with the estimated responses to the given combination of anthropogenic and natural forcing; and
* not consistent with alternative, physically plausible explanations of recent climate change that exclude important elements of the given combination of forcings."

Ditto for attribution.

You can keep trying to put me in a tribe on this, but if you are interested in accuracy not black hats/white hats it'd be best instead to read my papers! ;-)

William M. Connolley said...

Roger - this is IPCC stuff, so I don't see why reading your papers would help. As your quotes demonstrate, from the IPCC defn you cannot attribute anything to natural causes, and you are (all together now) failing to understand the difference between attribution, and not rejecting the null hypothesis.

Anonymous said...


I'll let you have the last word on this thread, but I'll simply note that that this statement of yours is just bizarre:

"you cannot attribute anything to natural causes"

How about all of climate history for billions of years?

And I will repeat once again for you, I am not in the Bill Gray camp, as you have incorrectly asserted. Please read my papers for my views. Thanks.

William M. Connolley said...

Well Roger, having yet another repeat is an odd way of letting me have the last word...

And yet again you've read things wrong. Of course you can attribute things to natural causes in some cases. I gave an example: Pinatubo cooling. Or another would be the ice age cycle. But by the IPCC defn, you can't *attribute* things to internal-variability causes, since that is the null hypothesis.

I know you say you're not in the Bill Gray camp. But for hurricanes, based on your words, it looks to me like you are.

Anonymous said...

Off-topic, but the poem's actual title is "The Second Coming." :)