It is often the case that at the end of one of my talks about some aspect of climate research or about the development of tools for the analysis of climate I get asked questions regarding global warming. Whether global warming is “happening” or whether the claims on either side of the issue are true or false. Or more to the point, I am asked whether we should or should not be concerned about warming. I know full well that my answers to this line of questioning is never satisfying; sometimes this is purposely so. I do have something to say about global warming and global change, but it is nothing more than a personal opinion. That of a concerned citizen, rather than of an expert on this issue.
I have taken the time and the effort to examine the data used by others to demonstrate a global warming trend, I understand how this data has been processed and what are the challenges involved in making the analyses. The upshot is that I have not seen any data or analysis that demonstrates that warming is not occurring. Like the myriad of people who work in global climate trends, I also see a very significant correlation between human activities and warming that make the industrial-era warming trend unlike any other change in climate before then. Further, I have not seen any reason to conclude that the best and most authoritative scientists working on global warming trends would not readily modify their conclusions if they were presented with data that shows a different picture of what’s happening.
The idea that climatologists and the governments that fund them have a self interest in promulgating global warming is so childish and yet so amazingly distracting to even thoughtful people: besides the fact that not all climatologists are climate change experts, ”climate” is a natural phenomenon, whether it is warming or not, and thus climate scientists will always have something to study. If anyone has a self interest in what climate does, however, it is those whose livelihood is affected by the state of climate itself, the groups from where most of the people who use this line of thinking come from. This is not a laughing matter: half a century of opposition to anything nuclear has had the effect of seriously slowing down research into the safe use of nuclear energy sources. I was not keen on how the nuclear industry does its business, but to nearly kill nuclear research was a terrible thing: we lost precious time in finding ways to safely use the stuff.
So what’s the fuss over a little bit of warming? If you live in the South West of the US, as I do, you are familiar with pre-Columbian communities that disappeared because of sudden drought, or post-Columbian communities that were wiped out not by war but by the rapid introduction of new diseases.
The main reason I cannot give you more than a personal opinion on the implications of global warming is that it is a risk analysis problem and this is not within the range of my expertise, not by a mile: sure, the more we know about climate dynamics, the more informed any risk analysis decision is. But that I know a little bit of probability theory and a little bit of climate dynamics is not adequate, just as someone who might know something about probability and know how to design cars will be poorly equipped to design liability insurance instruments.
There are parallels with the risk analysis problem associated with cigarette smoking: in the climate problem there is compelling correlation between human activities and global warming. There is a compelling correlation between cigarette smoking and some forms of cancer; but to date, no one knows exactly how cigarettes cause cancer. The causation relation between human activities (burning hydrocarbons, among other things) and global warming is not fully understood either. Decisions need to be made now (we cannot wait to know everything on climate). In the cigarette case, actions were taken to curb smoking, because the risk analysis made it clear that it was better to curb smoking than not. No one waited for the causation mechanism to be fully elucidated.
The cigarette risk analysis is not a perfect analogue to the climate change risk analysis problem: once you remove cigarettes, you remove the problem of cigarettes and cancer. On the other hand, climate will and does change, whether it is due to human activities or not. Climate change can have huge impact on some or many of us. Natural or man-made, some of the people who stand to loose a lot from the change are those very people who rely extraordinarily on cheap amounts of energy to make it through a change that can have monumental effects on society and the economy.
To characterize this problem as non-existent is stupid. But to think that it is just a bit more complicated than what we need to do to keep the Mississippi river flowing in the same place, always, is at best, irresponsible. Aside from being one of the most complex risk analysis problems ever, it’s not clear what should be done or could be done to reduce the impact of climate change. It is perfectly valid to ask, as some climate change debaters do, whether putting in the (presumed) huge resources to tackling climate change right now would be better spent on wiping out hunger or ravaging diseases, for example.
No one really knows what could be done to weather or climate change. Some scenarios include climate engineering. Climate engineering gives some people the shivers, since it requires us to know a lot about climate (I think far more than we know at present), but research in it should be pursued. Virtually every risk analysis scenario involves two aspects of human activity: energy use and population growth. Both issues are correlated with warming and go beyond climate change. Did I also mention that some of it is unpalatable, if you are the country that uses about one quarter of all energy worldwide, or if you are a developing nation, or if you are neither but your country is slowly disappearing under the waves and you have zero influence on the rest of humanity?
The risk analysis folks don’t even have a good managed-risk plan, at present. What are the key challenges in climate dynamics that need to be handled in order to produce serious climate-change risk analysis strategies? You can mention obvious things: more data, the problem is that for a minimally 10^7 degree of freedom problem, there’s no expectation that we can circumvent using a Bayesian approach that combines models and data in some way. Significantly improved means on bounds on uncertainties and better means to obtain these. In models they need significantly more reliable ways to capture how local effects affect global ones and vice versa. They need faster ways of producing different climate scenarios. They need better and more complete notions of sensitivity analysis. But frankly, the most important thing is a better understanding of the physics of climate: a global climate model is nothing more than a compendium of dynamics that agree with our expectations of outcomes; there are no theorems in this business.This is how most science, outside of mathematics is done: via compelling evidence, not necessarily evidence beyond a shadow of doubt. Insight into climate is the key to understanding how to parameterize microscales, challenge the curse of dimensionality of the system, produce better data assimilation strategies, create better sensitivity tools.
The writer is Professor in the Mathematics Department at The University of Arizona. He has been doing research in climate dynamics, ocean dynamics, sensitivity analysis, and data assimilation for about 20 years. He is also Professor of Physics and of Atmospheric Sciences at Arizona.
Prof. Juan M. Restrepo
Group Leader, Uncertainty Quantification Group
Department of Mathematics
University of Arizona
Tucson AZ 85721, U.S.A.