Tuesday, June 25, 2013

"Did climate change cause storm X" is the wrong question to be asking.

Treehugger has a story titled, "Did Climate Change Cause the Alberta Flooding?" Although the short article was decent, it leads with a question - the general form of, "Did climate change/global warming cause weather catastrophe X?" - that is misguided in its construction. Let me explain.

"Climate change" is talking about the shift from the historic climate regime into a different (ahistorical) one. That is all. Different areas (e.g., Toronto vs. Denver) have different climates, but they still do share some weather events (e.g., snow, wind, rain). However, the likelihood of one type of weather event happening is more an issue of climate.

However, what's makes this complicated is a conflation of the actual incidence versus the likelihood of the incidence. For example, if there is a 10% chance of rain for a particular date, and then it does rain on that date, the statistical likelihood of rain remains at 10%, even though the actual incidence of rain is 100%. Why? Because these are measuring different things. To use a baseball analogy, let's look at the Blue Jay's Adam Lind who has a .337 batting average and 30 RBIs for 202 at-bats so far in 2013. This means - all things being equal - he will get on base about 1/3 of the time and - of those times that he gets on base - his hit will allow someone(s) to score about 1/3 of the time. (That's pretty darn good.) Okay, now let's say that at his last at-bat, Lind hit a grand-slam home-run. (Congrats to Lind in this hypothetical case!) Now the analogous question: "Did his batting average and RBIs cause the grand-slam home-run?" That's a nonsensical question, and the only way you can really answer it is, "no, Lind's batting average and RBIs did not cause the grand-slam home-run." And - with baseball - we understand why that question doesn't make sense to ask. When it comes to climate change, though, we don't understand (yet) why the question is the wrong question and why the answer given by climate change deniers (as well as climate change scientist) are both, "No," or, "We can't say for sure." (This last response more often from scientists.)

A better question is, "How much has the likelihood of something like storm X occurring changed from its historical condition?" To use the baseball analogy one more time, this form of the question is more similar to the question of, "How have Lind's batting average and RBIs changed compared to his average before he was sent down for a time in the the Triple A's (i.e., his "historic batting average")?"

Previous entries explaining the difference between climate and weather: here (also using the Blue Jays), here, and here.

Previous entries looking at the implications of a change in the underlying incidence rate of floods: here and here.

No comments: