A Quantitative Method to Evaluate the Performance of Climate Models in Simulating Global Tropical Cyclones

Abstract

The capability to reproduce tropical cyclones (TCs) realistically is important for climate models. A recent study proposed a method for quantitative evaluation of climate model simulations of TC track characteristics in a specific basin, which can be used to rank multiple climate models based on their performance. As an extension of this method, we propose a more comprehensive method here to evaluate the capability of climate models in simulating multi-faceted characteristics of global TCs. Compared with the original method, the new method considers the capability of climate models in simulating not only TC tracks but also TC intensity and frequency. Moreover, the new method is applicable to the global domain. In this study, we apply this method to evaluate the performance of eight climate models that participated in phase 5 of the Coupled Model Intercomparison Project. It is found that, for the overall performance of global TC simulations, the CSIRO Mk3.6.0 model performs the best, followed by GFDL CM3, MPI-ESM-LR, and MRI-CGCM3 models. Moreover, the capability of each of these models in simulating global TCs differs substantially over different ocean basins.