Wednesday, September 05, 2007 ... Deutsch/Español/Related posts from blogosphere

Warming, hurricanes, and legends

It is interesting to observe what kind of "evidence" seems enough for certain people to convince themselves about far-reaching scientific assertions. We have finally seen two Category 5 hurricanes, Dean and Felix, so the people who have been silent about the hurricanes for 18 months suddenly started to be visible again. Their scientific theories are as variable as the weather.

asks whether Dean and Felix are two brothers whose father is global warming. The answer "remains unclear" but it is enough to ask the question to propagate the meme (or myth), isn't it?

The first sentence says

Despite growing consensus that global warming may spawn stronger tropical cyclones, weather experts believe it is too soon to blame climate change for the unprecedented punch of back-to-back monster hurricanes.
Well, this sentence is a contradiction. If it is too soon to tell, there can't be any growing "consensus" that the answer is "Yes".

But let us look at the third paragraph:

It was the first time on record that two Atlantic hurricanes had made landfall as Category 5 storms in the same season, and only the fourth time since records began in 1851 that more than one Category 5 had formed in a year (after 1960, 1961, 2005).
Well, this is really extraordinary evidence. We have the fourth year in which more than one Category 5 hurricane was formed (and detected, something that they don't say: do you think that scientists in 1851 had the full Atlantic ocean under control?). Is it something statistically interesting? Well, I don't think so. As some papers have shown, a combination of more comprehensive detection techniques and pure chance is enough to explain the data in a satisfactory way.

But even more awkward is their argument that "it was the first time on record that two Atlantic hurricanes had made landfall as Category 5 storms". The assumption behind this sentence is that global warming not only makes hurricanes stronger but it makes sure that they will be stronger exactly before they make landfall so that the damage is maximized.

Perhaps, global warming is also optimized to cause damage in Nicaragua, a country of a heroic communist leader Daniel Ortega who has described his revolution as the twin revolution of the Iranian Islamic revolution (too bad that Felix couldn't destroy their capital city). That's a similar theory as the theory about Elvis Presley who is going to be found in Tora Bora.

Many people just can't estimate how likely it is to find something seemingly unique or special about one year. I claim that the following true statement is comparably scientifically relevant as their statement:

The year 2007 is the first year in one thousand years whose last three digits encode the most famous secret agent and the first year ever when people realized that.
Well, that's perhaps too funny but do you see my point? If you combine various conditions to construct your criteria, you may almost always design a criterion that will make the current year special, especially if you only have approximately 100 competing years. They could have also written that

1960 and 1961 were the only two consecutive years on record in which more than one Category 5 hurricane was formed.
Apparently, they don't find such statements interesting enough if they're not about the recent era. The statement above could support the professional cassandras in 1961 who were predicting a doom 46 years ago. But it seems that these cassandras were wrong so we must find some support for the next choice, the 2007 cassandras, right?

Last year, they could have also written that
2006 was the first year in the history in which the number of Category 5 storms decreased by more than 2 storms relatively to the previous year: it was by 4!
Well, no one has found this observation useful enough to be printed, either. Shockingly enough, William Connolley argues against the overinterpretation of these contrived statistical criteria and agrees with your humble correspondent: he even independently writes that people usually blip over the caveats in articles.

After the comments discussed above, Reuters declares that the theoretical maximum speed could be 10% higher when the sea surface temperatures are higher. It might be true or not: the data don't provide us with much evidence that it's true - for example neither 1960 nor 1961 were terribly warm years. But many other reasons behind particular hurricanes are known and no scientist is gonna promote a direct link (unfortunately for the author of the essay, as you can see). They make sure that these careful comments sound boring and they're hidden because, I claim, the real purpose of the article is to make the readers think that the link is there.

I think that people should learn how to "smell" propaganda and bias that tries to manipulate with their opinions. Contrived criteria that are manufactured to make the recent era (or the most recent year) look more special than it is are a great example of such a bias. People should learn how to define analogous criteria in the same universality class and figure out whether they still lead to the same conclusions.

People should learn how to ask the right neutral and maximally relevant questions about science and find the relevant data themselves. For example, the most relevant hurricane-related quantity for Americans is probably the number of U.S. landfalling hurricanes and there seems to be no trend in the last 156 years. Once people rely on either side of any dispute to invent interpretations and slogans for them, they're already being had.

And that's the memo.

Add to Digg this Add to reddit

snail feedback (0) :

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');