Evaluating the Role of Kernel Choice on Radiative Feedback Estimates

Résumé fourni par la source

Abstract Radiative feedbacks are important for understanding climate sensitivity, yet estimates vary widely across climate models. Feedbacks are often computed using the radiative kernel technique, which uses pre‐calculated radiative sensitivities (kernels) derived from different models and base states. While much attention has been given to intermodel differences, the impact of kernel choice has received less scrutiny. Here, we quantify the influence of 10 radiative kernels on feedback estimates from 48 CMIP5 and CMIP6 models. We find that kernel‐related spread is comparable to the intermodel spread in global water vapor and albedo feedbacks, and in polar regions, the kernel‐induced spread for cloud feedbacks is also comparable to the model spread. We show that differences in the base climate state used to derive kernels are an important source of this spread. Our results suggest that we should account for uncertainties due to kernel choice, particularly in polar amplification studies.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.