Policy

When is it BEST for Scientists To Trust Each Other's Results?

YouTube

This will, I promise, be the last post I do on climate for a little while. But I could not resist passing along this video from Big Think.

It's an interview with James Lawrence Powell about the Berkeley Earth Surface Temperature (BEST) study and Richard Muller. As many of you know Muller, a well-known physicist, considered himself a skeptic of some aspects of climate science. With funding from the Koch brothers (and others) he started BEST to do an independent study of the planet's temperature history over the last hundred years or so. The results, (to no one in the climate community's surprise), were exactly in line with all the previous studies, including some at the heart of the faux scandal dubbed Climategate.

So was Muller doing climate science a service?

Powell, a geochemist and executive director of the National Physical Science Consortium, thinks perhaps not. While Powell is happy to see any experiment repeated to confirm or deny an established result, there was something about Muller's attitude that seemed disingenuous in this politically charged issue.

"I think what we could learn from the case study that Muller did," says Powell, "is that [Muller] should have trusted the other scientists and the peer review process which had produced the data that he was questioning."

I have heard similar points from other climate researchers. Studying the subject for decades, they have done their absolute best to root out bias and error from their work. The fact that the BEST study got exactly the same result as everyone else confirms that those earlier studies constituted good science. Having Muller come along to tell the world he's now vetted that work so everyone else can believe it seems a little beside the point at this stage in the game.

Given the importance of climate science to our future, it is good to shine as much light as possible on its methods and conclusions. But at what point do we agree that the process has worked as it should and the results are trustworthy?

The answer can't be "never."