May 8, 2013 10:49AM

The Science vs. the Pseudoscience of Extreme Weather

Over at Capital Weather Gang, the always-perceptive Jason Samenow details a recent Twitterspat between Dot Earth’s (aka The New York Times’) Andrew Revkin and Penn State’s Michael Mann over attributing extreme weather events to anthropogenic climate change—tornadoes, in particular.

Revkin tweeted to ask whether the folks who were alluding to anthropogenic greenhouse gas emissions being behind the major (and deadly) tornado outbreak during the spring of 2011 were willing to attribute the record lack of tornado occurrences during the past 12 months to the same cause.

Revkin could have very well asked this same question about all kinds of bad weather—blizzards, hurricanes, droughts, floods, record heat, record cold, summer in Washington, winter in Chicago, etc.

Used to be, when the weather was bad, folks would logically cite Mark Twain’s “if you don’t like the weather in New England now, just wait a few minutes.”  Now, someone will show up on TV who is quick to point out that this sort of thing “is consistent with” expectations of global warming.  These same folks tend nap when the weather is hunky-dory, and to go into hibernation when the extreme weather category of their previous pronouncement has a hiatus.

Since the bang-up year of 2011, the number of tornadoes has dropped off the table, with the last 12 months showing the fewest since systematic observations began in the 1950s.

And like tornados, major hurricane strikes have also become scarce, in fact, they are so in remission that someone might soon announce they have been cured.  It has currently been more than 7 years since a Category 3 made landfall in the U.S., the longest time in more than 100 years—and all this when overall hurricane activity in the Atlantic basin has been elevated.  Maybe there is something to research that finds that while anthropogenic climate change may increase the frequency of major hurricanes in the Atlantic, it changes the circulation patterns such that they are more likely to remain offshore (see page 30-32 of our comments on the draft National Assessment Report)

But we digress…

Apparently the folks who rally around the anthropogenic climate change/extreme weather linkage don’t like being awoken when all is calm.

In his response tweet, Michael Mann accused Revkin of “concern trolling”—which Samenow kindly defined for us as, in this case, “manufacturing a question to score points with climate change skeptics.”

Mann followed up with “Perhaps Andy ‪@Revkin is claiming that the background state of the atmosphere has NOT changed? Is that what he is arguing?”

We are sure that is not what Revkin is arguing at all.

It is true, and Revkin knows this, that human-activity in all its forms exerts some influence on the base state of the atmosphere. So, what?

This does not mean that such an influence is noticeable or even detectable, much less that every time the weather is bad, human activities are to blame.

In fact, in virtually all cases, the magnitude of natural variability is still much larger than the magnitude of the human influence. Nor is the specific impact of the human influence on any weather event ever clear.  For example, see our comments about “superstorm” Sandy.

What this does smack of, is pseudoscience—an explanation of things that is not refutable by any conceivable event. Karl Popper famously drew this distinction in his 1957 essay (republished in 1963) Science: Conjectures and Refutations.

If the explanation is that bad weather events serve as evidence that humans have altered the atmospheric base state in such a way as to perceptively influence (make worse) extreme weather events, but that the absence of such events does not serve as counter evidence, then we have a situation where the theory is untestable. Pseudoscience.

Before we should even start to consider taking seriously claims about global warming making things worse, we need to start seeing some predictions about individual events.

Next time there is a discussion about a potential tornado outbreak somewhere in the U.S., we want to see a priori discussions of how anthropogenic climate change will influence the event.  For instance, will it lead to 25 tornadoes instead of 23?  Will the storms stay on the ground for 15 minutes instead of 14?  Will their path length be 12 miles instead of 11?  Will they hit the Lazy Acres Trailer park or just skirt by? Will they occur when kids are gathered in the school cafeteria or just after lunch lets out?  All these factors, and many orders of magnitude more, collectively determine the impact of the event. And this completely leaves out the confounding influence that there are now more people, with more stuff in potential harm’s way.

Now, you might rightly point out that such predictions are impossible—the atmosphere is too chaotic and our models too coarse to pin down events with such detail.

But until such forecasts are issued—it is also impossible to assess whether the event acted in a manner expected by the global warming alarmists.

Despite this lack of predictive ability, there is no shortage of ex post facto postmortems.  Until that changes, no autopsies please!

True science offers up expectations in advance (i.e. “hypotheses,” which, in the weather/climate business are called “forecasts”), and then uses observations to verify (or reject) those hypotheses.  Pseudoscience accepts all observations as support of its theories.

Isn’t it time we take a greater interest in the science of climate change?


Karl Popper:  Conjectures and Refutations, Routledge & Kegan Paul, 1963, reprinted 2008, pp 43-86.