Computer Models Simulate the Present-Day Climate Well
22 Apr, 2008 11:40 am
Information about climate and how it responds to increased greenhouse gas concentrations depends heavily on insight gained from numerical simulations by coupled climate models. The confidence placed in quantitative estimates of the rate and magnitude of future climate change is therefore strongly related to the quality of these models. In this study, we test the realism of several generations of coupled climate models, including those used for the 1995, 2001, and 2007 reports of the Intergovernmental Panel on Climate Change (IPCC). By validating against observations of present climate, we show that the coupled models have been steadily improving over time and that the best models are converging toward a level of accuracy that is similar to observation-based analyses of the atmosphere.
A new study by meteorologists at the University of Utah shows that the latest generation of climate models is quite accurate when compared against real observations. The study entitled “How Well do Coupled Models Simulate Today’s Climate?” appeared on the Bulletin of the American Meteorological Society. In this study, Thomas Reichler and colleague Junsu Kim, a graduate research assistant, investigate how well climate models simulate our present climate.
In order to evaluate the models, the authors developed an objective measure for the agreement between the output of the models and the observational datasets for present climate. They apply this measure to about 50 different national and international models that have been developed over the past two decades at major climate research centers in Australia, Canada, China, France, Germany, Italy, Japan, Korea, Norway, Russia, the United Kingdom, and the United States. The authors included in their comparison also the latest generation of climate models that was used for the recent (2007) report of the Intergovernmental Panel on Climate Change (IPCC).
Reichler and Kim investigated how much climate models have improved over time and how much we can trust the latest model generation. The authors tested the models on current climate assuming that when models succeed in predicting the present-day climate they will also succeed in predicting the future climate.
The study finds that the latest model generation is much more realistic than earlier models. Some models reach an unprecedented level of realism. This improvement arises from the enormous progress in model development that took place over the past, which is due to a better theoretical understanding of how climate works and from an increase in computational resources. The authors conclude that one can have now more confidence in predictions of future climate.
Thomas Reichler and Junsu Kim, How Well Do Coupled Models Simulate Today's Climate? Bulletin of the American Meteorological Society, vol. 89, iss. 3, 303-331. (March 2008). Abstract available here.