More Katrina Misattribution?

It took me a few seconds to realize that 'GW' does not stand for 'George W'. LOL

You don't think the Gulf has been warmer than usual? And that warmer-than-usual is what global warming models would predict?

Chris, you are making a serious error here, and I am shocked that you would even mention it. The authors' statement is more than reasonable.

The probability that higher than average sea surface temperature in the Gulf of Mexico was a significant variable fueling Katrina's intensity would be on the order of at least ninety-five per-cent - Max Mayfield and William Gray notwithstanding. Moreover, the probability that the increased SST was significantly caused by anthropogenic global warming (yes, anthropogenic) would be better than ninety per-cent. Such basic inferences from powerful empirical data are the heart of scientific hypothesis formation and testing.

David,

I think Chris makes his point clear enough... but I think your question is also fair. The smart guys at Real Climate tried to articulate this issue of attribution in their post just after Katrina:

"Due to this semi-random nature of weather, it is wrong to blame any one event such as Katrina specifically on global warming - and of course it is just as indefensible to blame Katrina on a long-term natural cycle in the climate."

So, in a way, you're damned if you do and damned if you don't. A warmer world increases the likelihood that SST in the Gulf of Mexico during any given week will be above average. It follows that such conditions make more powerful storms more likely as well. And, since Giller & Wroth don't explicitly say that global warming caused katrina... it would appear that they may have walked this fine line with some degree of skill.

Folks,
I'd be glad to be proved wrong on this, that's what blogging is for...but the authors appear to be making a direct attribution of Gulf SST on a particular week to global warming....I'm expressing skepticism that you can go from the average to the extremely specific like that, but maybe I'm mistaken.

I think Chris may be right on this... The authors write:

"Why were the Gulf's waters warmer than usual? You guessed it -- "

In the context off this piece, it's clear that the answer they are pointing too is "global anthropogenic warming." And they are not talking about Gulf waters on average. They are talking about SSTs at a very specific point in time... which is where they cross the line.

As an aside, I have not yet seen the movie... but I fear that it crosses this same line with regularity. I hope I'm wrong... but I think we're likely to start seeing more and more people missing this point regarding (mis)attribution.

Thanks, James...so now that we have the point of contention here clearly framed, does anyone think that James and I are wrong on this specific point, and that we *can* attribute SSTs in a specific place at a specific time to a global average background temperature change?

I see two attribution steps. The first between global warming and a specific SST increase, and second between a specific SST increase and a hurricane. With each step you lose some certainty (unless you're completely uncertain to begin with). The RealClimate argument crossed both attribution steps, but if you knew SST were greater, your predictive power increases. That said, I can't answer either question, but I feel like I helped by doubling someone's workload.

I may be out of my depth here, but could you say something like this: Since 1920 (or whatever year) the average SST in the gulf has risen by N degrees per season, to NN degrees. On the day that Katrina came through, it was significantly higher, XX degrees.

In 1920 (or again, whatever year) the chance of the Gulf's SST being XX degrees was Y %, but in 2006, the dice are loaded-- the chance of it being XX is raised to Z %. And therefore the window allowing Katrina-like phenomena is much wider (more days of optimal conditions during the season).

You might then find a familiar analogy that corresponds to this statistic. For instance, there are many times that drinking does not directly cause a car accident, but "A" number of drinks increases the chance by "B" % that you'll have an accident, etc. ...So it can't be said that drinking directly caused all these accidents, but the government is justified in making a certain blood alcohol level illegal. (Again, I may be out of my depth, I kind of pulled this analogy out of the air.)

The trouble is that statistics are so abstract, but the problem is very concrete. Warm water raises the lethality of "hurricanes." How do you get this across, and still keep "hurricanes" plural? It's not surprising that journalists reach for the newspaper headlines to get their point across, because when they do, they're working with what's already in the public's mind.

By the way, if I had to choose between reporters occasionally slipping and coloring outside the lines, or being really timid and not coloring enough? I'll pick occasionally coloring outside the lines. You do your best, but this is a complex subject and most reporters lack advanced degrees...

It seems to me the opposition is playing a completely different game. Their eyes really aren't on these kinds of details. How many pundits did a close reading of the first paragraph of the Time story a while back? They were too busy brandishing the cover and blustering in basso profundo. And if they did start getting into the details? I think that would be good, because those details do not favor them...

Hey Jon,

"I'll pick occasionally coloring outside the lines."

I think that many people who take this issue of global warming very seriously agree with you on this. Particularly, as you say, given what "the opposition" is up to. Especially after Katrina, many of my friends and family members felt that it may be ok to push the hurricane-GW attribution argument and I ended up in general disagreement with many of them over this point. Yet, I share your frustration with the thickheadedness of the punditocracy... so, I'm sympathetic to your view.

However, nearly all of the scientists and journalists that I know are generally unwilling to force the hurricane-GW issue by "coloring outside the lines," as you suggest. Our value to society depends too much on our credibility for this to be a wise or productive course of action. It's one thing to openly advocate a general policy solution (e.g., reduce emissions, one way or another)... but it is another thing altogether to distort the facts in doing so.

Best Regards,

James

James-- Thanks for your thoughtful comment.

I think what I'm saying is that a piece that communicates effectively with a high degree of accuracy may still contain flaws. For instance, RealClimate says that Al Gore's movie handles climate science "admirably", but at the same time remarks that there are some "scientific errors that are important in the film":

http://www.realclimate.org/index.php/archives/2006/05/al-gores-movie/

It would have been nice if Al Gore's movie had been error-free. But it's also nice that the movie didn't stay in scientific committee, so to speak, into perpetuity.

So I guess what I'm afraid of is a perfectionist streak that hobbles trying to tell the story. Compelling stories about complex matters are hard to tell. And it seems like a thankless job, because when it's done well it tends to look effortless.

But I agree that it wouldn't be good if reporters began to deliberately "color outside the lines", or if they stopped giving their material due dilligence. And of course I'd agree that reporters' credibility is the top priority. But wouldn't telling stories well and getting them to press be a priority as well? Even if later, when an expert closely parses it, there's a flaw? Again, this is a complex subject, so errors like this are bound to happen.

I probably should add a disclaimer that I personally am not a reporter or a scientist, so it's possible that I lack perspective.

As a scientist trying to find out how tell my stories outside the commmittee, I am keenly looking for this elusive third way, of colouring inside the lines and using just the right amount of colour. I feel it's like an Asterix cartoon (Belgian cartoonist; Gaulois characters), which appeals to so many people and on so many levels (eg. the druid's name was Getafix). But hey, scientists are human, we make mistakes - we also colour outside the lines either on purpose or by accident.

Hey Jon,

Thanks for your response and for the link. Over at RC, I noticed that Roger Pielke Jr. piped in almost immediately to give his two cents about the hurricane-GW connection, suggesting that Gore may (or may not) be coloring outside the lines and that this would be scientifically indefensible. I guess this is different from scientifically indefensible statements made by Senator James Inhofe (Chairman of the Committee on Environment & Public Works) on the floor of the U.S. Senate, on national television and in committee hearings. I'm still not clear on why one is intolerable and the other is just "politics as usual."

So, your points are very well taken and I appreciate the thoughtful exchange.

Best, James