Sensitivity of Asian Summer Monsoon Precipitation Simulations to Perturbed Parameters in FGOALS‐f2 Seasonal Hindcasts

Abstract

Improving the simulation of the Asian summer monsoon (ASM) precipitation pattern remains a critical challenge in global climate models, which often exhibit persistent regional biases. In this study, a targeted parameter tuning strategy is introduced to identify and optimize key physical parameters in the FGOALS‐f2 model, with the specific goal of enhancing its ASM representation quickly and economically. We employed a series of experiments that included two types of perturbed parameter ensemble (PPE) from a total of 65 parameters under a seasonal forecasting framework; more than 500 seasonal forecasting experiments were carried out, and the experiments were ultimately tested in long‐term Atmosphere Model Intercomparison Project (AMIP)‐type and Coupled Model Intercomparison Project (CMIP)‐type simulations. Our results indicate the importance of parameters in deep convection and cloud microphysics schemes, specifically governing entrainment and subgrid hydrometeor fall speeds, as the most sensitive parameters for ASM simulation. Optimizing these parameters could substantially reduce long‐standing precipitation biases over East Asia, the South Indian Ocean, and the Maritime Continent. However, our experiments also reveal a quantifiable skill limit for parameter tuning alone. The precipitation simulations for the Bay of Bengal and western Pacific exhibited persistently low skill and uncertainty. This work demonstrates that PPE experiments based on seasonal hindcasting can be used to quickly test model sensitivity to different atmospheric parameters, while the influences of different parameters are similar in most of the monsoon regions in both the seasonal hindcasting and AMIP‐type simulations.