An Analysis of CMIP6 Climate Model Performance in Simulating Climatological Conditions Compared to CMIP5


  • Catherine Liang Aspiring Scientists' Summer Internship Program, 2019
  • Alexander Hall Aspiring Scientists' Summer Internship Program, 2019
  • Abdullah al Fahad Center for Ocean-Land-Atmosphere Studies, College of Science, George Mason
  • Dr. Natalie Burls



The Coupled Model Intercomparison Project (CMIP), a worldwide multi-model framework for climate experiments and projections, recently began releasing output from the CMIP6 coordinated effort to the scientific community for analysis. The climate models used to undertake CMIP6 represent the state-of-the-art, with the latest scientific understanding included in this generation of climate models. Consequently, one expectation is that this latest generation of models can more accurately simulate climatological conditions in temperature and precipitation across the globe than the previous CMIP effort, CMIP5. However, the skill of these “improved” models has not yet been deeply investigated. We compared the outputs of models under CMIP6 against observations through multi-model mean plots and individual calculation of the Root Mean Square Error (RMSE) to both identify any possible biases. The plots indicate visually apparent improvements, while density distributions for RMSEs show shifts in cross-model performance between CMIP5 and CMIP6. In this project, RMSEs in CMIP6 are noticeably lower for precipitation and sea surface temperatures both globally and in the Pacific equatorial regions. Our results indicate a possible improvement in climate models from CMIP5 to CMIP6, with a reduced wet bias around the equator, and reduced warm biases in the Pacific.  We anticipate that our findings will add to the scientific community’s understanding of ongoing developments in climate modeling groups and will help promote future study.


Author Biography

Dr. Natalie Burls

Department of Atmospheric, Oceanic, & Earth Sciences, College of Science, George Mason University     





Abstracts from the 2019 Aspiring Scientists' Summer Internship Program