I spend most of my time doubting what it is that I’m doing. Most? Pretty much all of my time. Much is made of doubt in science. The deep mystery of the universe. Peering into the unkown.

I’m not talking about that sort of doubt. No, I’m talking about the fact that nearly every day I have the continual feeling that I have absolutely no idea what I’m doing. Feyman talks very eloquently about doubt and ignorance¬† in science. The forever frustration in trying to understand the rules of the game that nature is playing. I don’t even know what game I’m looking at most of the time. Every now and again the crowd cheer and some people move about or swap places. Something significant has happened I conclude. Then it’s back to white noise.

The reasons for this are (mainly) twofold. First and foremost I’m not very smart. I wish I was. But there it is and that’s how it goes. I’ve just got to get on with it. Second, I do ‘multidisciplinary’ research. I spend a lot of my time talking to people who really do know things. And who are also smart. There’s a wonderful passage in an E. O. Wilson book (title escapes me) in which he describes crouching down at the edge of a tropical rain forest, looking at a small wolf spider. Staring into its tiny, unblinking eyes he considers how little we know about this particular species and how he could happily imagine spending many years or decades discovering how it moves, hunts, breeds and makes its living among the giant trees.

I appreciate this sort of scientific obsession. To continually learn more about something and so fill in more of the blanks. The ability to write a richer story about some little facet of nature. But try as I might (and OK, I haven’t tried that hard) I always seem to get distracted by something else. Sometimes it’s not the something else per se, but that way that this new thing may interact with some other things.

I used to play a game as a child. I’m sure many others played/play this sort of game too. You name two different things. Then you connect them by telling a story. Telephone and conkers. Blanket and abseil. Points awarded to the best fabulist. If you’re not careful, you can see connections everywhere. That’s a common feature of conspiracy theories. But doubt should limit the sort of epistemological damage that unconstrainted connections can do.

If I want to study connections and their system properties then I’m doomed to a life of doubt. But I always hold out for those moments when I make out the cheers and shouts, a pattern emerges and I feel I’ve learnt something.

New paper in press

I’m a co-author on a paper that will soon be published in the Journal of Biogeography. You can download a PDF version of it here. It’s about how we can make better predictions of the response of plants, trees and other vegetation to things like climate change. This is something that a lot of people are spending a lot of time and money trying to do.

Why such interest? Take the change in forest cover for example. Climate models predict potentially profound changes in temperature and rainfall in the Amazon region. If significant numbers of trees in rain forests were to die in response to such changes, they would release many millions of tons of carbon dioxide into the atmosphere which, along with other effects, would have large impacts on the local and global climate.

In this paper we have suggested how vegetation models (in particular, models called Dynamic Global Vegetation Models – DGVMs) could be improved by using the ever-increasing amounts of new data being produced by field studies, aircraft and satellites. We outline how this data can improve model performance via Bayesian statistics. Currently, data is used to determine the best values for model parameters. For example, people conduct experiements to try to determine how photosynthesis is affected by changing temperature. But this leaves a lot of data unused. By¬†inverting DGVMs it should be possible to produce better model predictions because more of the data can be employed to refine model output. Depending on how you feel about statistical methods and inference this may be a great idea or something really rather unsettling. To those experience the latter emotion I would say: relax, it doesn’t matter what colour the cat is, only that it catches mice! Actually that may make it worse. Let’s talk about it over coffee.