Using feedbacks derived from a modified energy balance framework to validate climate models

Abstract:

Given the importance of global climate model (GCM) simulations in climate science and policy, finding effective ways to test models against observations is a key goal of the scientific community. In this presentation, we show a better way of evaluating feedbacks in models with observations. We use a modified energy balance framework described by Dessler et al. [Atmos. Chem. Phys., 18 (2018), pp. 5147–5155], which ties changes in top-of-atmosphere flux to changes in 500-hPa tropical temperatures rather than the normal approach of using global average surface temperatures. This study evaluates the performance of Coupled Model Intercomparison Project (CMIP) phase 5 and phase 6 GCMs and compares feedbacks derived from them to those inferred from observations. Overall, the CMIP5 and CMIP6 GCMs both perform well, and there is no evidence that the GCMs have errors in their feedback physics or are missing any important processes. Compared to the traditional way of defining feedbacks, we find that the modified energy balance framework allows a more robust test of the GCMs. We are further able to identify problematic GCMs and rank them by how well they do in reproducing the observed feedbacks.