What are Climate Models Good for?

Gavin A. Schmidt, NASA Goddard Institute for Space Studies

Lecture Summary

Click here to view presentation video

Models are widely used tools that can be employed to tease apart and examine many of the variables inherent to complex systems. They may appear abstract or overwhelming or even mind-boggling, but models actually play a role in the everyday lives of most people. Market trends, certain weather predictions, and changes in fish populations are based on models. However, few of these models have been met with as much public and political skepticism as those that predict global climate change.

“What are climate models good for?” was the question posed by Dr. Gavin Schmidt of NASA Goddard Institute for Space Studies in his Metcalf Institute lecture on Monday, June 4, 2012, at the University of Rhode Island Graduate School of Oceanography. The models he generates to examine climate change include links among the small-scale processes that, taken together, can lead to global impacts. Schmidt and his colleagues then compare their models, models from other researchers, and real observations to quantify the range of variability in their models. This approach enables them to address certain aspects of the model itself, such as figuring out which variables may have greater impacts and which could be negligible.

Schmidt explained that models make it possible for researchers to test hypotheses about climate change — for example, how global temperatures might be affected if human sources of carbon dioxide were removed — and about the impacts of past events — for example, whether past volcanic eruptions are still affecting modern climate. Models can be used to understand climates in both the distant past, known as hindcasting, and the not-so-distant future.

Although models can be very powerful tools with diverse applications, Schmidt acknowledged that there is a degree of uncertainty inherent in modeling. The variability, or how spread out or closely clustered data is, in models increases with time. Much like the widening cone of predicted hurricane paths, forecasts have less certainty as they project further into the future. Another issue with models is that the relative impacts of different natural or manmade forces driving large-scale changes may be unknown, understudied, or underestimated. Errors in the initial conditions or assumptions used to construct a model could be the result of insufficient or imperfect information. However, the uncertainty can be reduced by quantifying the variability and comparing the results to other models, predicted and observed trends, and common sense. And finally, Schmidt noted, the combination of more field data and increased computing power have resulted in improved modeling capability and accuracy.

Schmidt concluded that models are powerful tools for examining and predicting broad-scale, long-term trends and they do this fairly well. Models allow researchers to ask explicit and interesting questions about the past or future and even questions about the efficacy of the model itself. They help researchers to identify how interactions — physical, meteorological, chemical, biological — are significant in ways that were previously difficult, unobtainable, or simply thought of as unimportant. So, while not perfect, models are reasonably accurate, improving, and, Schmidt argued, important tools for pushing the boundaries of scientific understanding.

Gavin Schmidt is a climate modeller at the NASA Goddard Institute for Space Studies in New York and is interested in modeling past, present and future climate. He received a B.A. (Hons) in mathematics from Oxford University and a Ph.D. in applied mathematics from University College London. He has worked with the American Museum of Natural History, the College de France and the New York Academy of Sciences. He is the co-author with Josh Wolfe of “Climate Change: Picturing the Science” (W. W. Norton, 2009), a collaboration between climate scientists and photographers for the general public. He was cited by Scientific American as one of the 50 Research Leaders of 2004, and was awarded the inaugural Climate Communications Prize of the American Geophysical Union in 2011.

Lecture summary prepared by Al Nyack, Ph.D.

Easy Sign Up