Climate Change and the Null Hypothesis

November 15, 2011

Brit Trogen

When faced with a scientific problem, one of the first steps that must be addressed is establishing the null hypothesis. If you intend to prove the efficacy of a new drug, your null hypothesis might be that the drug has no effect. If you intend to prove the intellectual superiority of cats over all other sentient lifeforms, you might settle on the H0 that cats are dumb as rocks and proceed in designing an experiment that has the potential to disprove this. 



In the realm of climate research, the null hypothesis has been fairly consistent for the past few decades: that humans are not responsible for our changing climate. Studies have been designed, therefore, to disprove this hypothesis and provide evidence to the contrary, which now exists in abundance. Such abundance, in fact, that some researchers are making the argument that the question itself needs to be reframed entirely.


In a recent paper by Kevin Trenberth in Wiley Interdisciplinary Reviews: Climate Science, the author has begun a provocative dialogue into changing the null hypothesis of climate science. Climatologists, Trenberth argues, shouldn't be trying to prove that humans are not changing our climate. On this point, there is no doubt whatsoever. Instead, they should be trying to disprove that we are. Not doing so, he argues, makes it too easy to falsely conclude that humans aren't to blame, particularly in political spheres, an assumption called a Type 2 error. 



Simultaneously, though, the journal has published two responses to Trenberth's paper, which appear alongside it and illustrate just how divergent views can become on a single issue. One response, by Judith Curry, rejects Trenbeth's premise on the grounds that his new null hypothesis would only be false to a trivial extent, and therefore impossible to disprove. She suggests instead abandoning hypothoses testing altogether, in favor of "scientific exploration" into the effects of human influence relative to natural causes of climate change ("how big is the effect?" rather than if it exists).


The other response, by Myles Allen, is more sympathetic, but still argues that the time is not yet right to change course. In Allen's view, there is not yet "overwhelming" evidence for human influence over extreme weather, so attributing anthropogenic causes to all weather events is not yet justified. In this case, however, it seems like a combination of time and further research could tip the scales in Trenberth's favor.


In the end, it seems unlikely that Trenberth's radical idea will find its way into the mainstream quite yet, but it has certainly struck a chord with many. Perhaps given some time to percolate, we will face a future in which the influence of human activity on climate is no longer even a (hypothetical) question.



Email (optional)


© 2010 Science in Seconds. All rights reserved.     Disclaimer  |  Contact  |  Subscribe
Friend Science in Seconds on Facebook Follow Science in Seconds on Twitter Science in Seconds RSS Feed