Comments on: Repeat After Me: Correlation Does Not Necessarily Imply Causation (Katrina and Global Warming)

By: Rob

Rob — Tue, 30 Nov 1999 00:00:00 +0000

Did you read the report analyzing the change in hurricane strength since the 1960s?

The trend exactly parallels ocean temperature, which has been rising. Ocean temperature supplies energy to hurricanes.

This holds true no matter where in the world. If it were simply a cycle, you'd see other areas at opposite ends of the cycle. Hurricane cycles throughout the world are either uncorrelated or anti-correlated. So it's not like we've been in one huge cycle since the '60s.

We can't say what percentage of Katrina was due to global warming, but we can say that the overall trend of fiercer hurricanes as global temperatures increase can only be attributed to global warming.

It also doesn't matter that some of the global warming is not caused by man. That we are partially responsible is sufficient. It's also worth noting that, since we're partially responsible, we have the ability to moderate the cycle as well. Not only can we undo the damage done by humans, but it's within human grasp to reduce the natural global warming.

We're humans. We modify our environment to improve our lives on small scales. Why not large?

By: Funky Dung

Funky Dung — Tue, 30 Nov 1999 00:00:00 +0000

On the whole, I believe that human-induced global warming is a real phenomenon. However, I don't buy a lot of the Chicken Little hype that I hear. My scientific training won't let me make the leap from statistics to blame so quickly.

From the page I quoted:

To clarify the above post, this is what I am and what I am not criticizing: I am criticizing the "jumping to a conclusion" that Gelbspan and others do in claiming that the severity of Katrina was due to global warming. I do not mean to criticize the premiss that the globe is warming, nor do I claim that the conclusion that Katrina's severity was due to global warming is false. Rather, my point is that it is impossible to know whether climate change contributed to Katrina's destructiveness.

Also:

People frequently assume that if one criticizes an argument one is thereby claiming that the argument's conclusion is false. This is presumably because one reason to criticize an argument is believing that its conclusion is false. However, logicians such as myself are trained to analyze arguments whether or not they believe the conclusion. Moreover, it is quite possible to argue badly for something that is true; it happens all the time! So, when I criticize an argument in this weblog, or in the examples throughout the files, I am not taking a position on the truth-value of the conclusion.

By: dlw

dlw — Tue, 30 Nov 1999 00:00:00 +0000

I think that due to the non-experimental nature of the data, and our limited records going back in time that there are ambiguities in its interpretation.

I think the real question is what should be our risk-function.

I haven't seen anything that explains why the fact we are entering a season of increased hurricane activity in and of itself would make the hurricanes stronger. As I understand it, global warming, via the higher temperatures in the gulf of mexico, is making the hurricanes stronger not more numerous and it really doesn't matter if there have been just as strong hurricanes in the past. Such hurricanes would be the most powerful of many, many hurricanes and the issue at hand is whether recent hurricanes are tending to be more powerful/destructive, so the comparison is not a very good one.

dlw

By: dlw

dlw — Tue, 30 Nov 1999 00:00:00 +0000

I think I meant to use the word loss-function, not risk function. The risk-function is derived from the loss-function or was it the other way around. Does FunkyDung believe in Bayesian Statistics?

dlw

By: Funky Dung

Funky Dung — Tue, 30 Nov 1999 00:00:00 +0000

Define "believe in".

You might be good and politics and economics, but you'd make a poor scientist. There's far too little historical data to comfortably say that the rising intensity of storms is *caused* by anything but normal cyclic changes. The best we can say with the data at hand is that there is a strong correlation between warming oceans and storm strength. That's a no-brainer because tropical storms feed off warm water. What's not so obvious is why the water is warmer. It's no trivial matter to deduce whether or not humans have contributed appreciably to global warming. My gut says we have, but I don't have an appropriate theory of the phenomena and stats to back it up.

To paraphrase the author I linked to, just because I pick aprat a bad argument, doesn't mean that the conclusion of the argument is wrong. It only means that the means of reaching that conclusion, the logical methodology, is flawed.

"There's nothing I like less than bad arguments for a view that I
hold dear." - Daniel Dennett

By: dlw

dlw — Tue, 30 Nov 1999 00:00:00 +0000

I define Bayesian Statistics as a non-positivistic approach to statistics that seek to clarify the a priori assumptions made in the interpretation of data with the end of improving decision-making using all the relevant information available.

And my point is that we may have to make decisions now based on less than ideal situations and less information than we would prefer.

I.e., the p-value may be high, but the more important question is how we should weigh the making of a type-one vs type-two error.

dlw

By: Funky Dung

Funky Dung — Tue, 30 Nov 1999 00:00:00 +0000

"I.e., the p-value may be high, but the more important question is how we should weigh the making of a type-one vs type-two error."

From a response policy standpoint, I can agree with you. However, I draw the line at throwing blame around. For instance, some have blamed the destructiveness of Katrina on Bush not signing the Kyoto Protocol (and other mus-slinging nonsense).

By: dlw

dlw — Tue, 30 Nov 1999 00:00:00 +0000

I will agree that causally I doubt signing the Kyoto Protocol and making the changes would have prevented Katrina from being destructive.

But from a political spin perspective, Katrina-style destructiveness is the reason the sorts of measures proposed in Kyoto were made.

In politics, a lot of hay is made by the manipulation of associations in people's minds. Hence, you see conservatives repeating the news that Mars is having its own global warming. I don't see this as mud-slinging, I see it as a somewhat opportunistic manipulation of the images/insecurities in the minds of a people who generally won't take the time to do a more careful study of the matter.

Ideally, we should call on people to deliberate, to read up on the facts of the matter, not to flush most of what they learned about science in high-school down the toilets of their minds upon graduation. But inasmuch as such exertion is not pleasant for most people, it's going to take some serious changes in our hyper-individualized culture for people to become better informed voters, more skeptical of the spin we are fed daily in our news.

dlw