Evaluation of CMIP6 models performing on rainfall seasons and moisture tracking simulation in Yangtze River Basin

Abstract

This study evaluates the performance of the CMIP6 models in simulating monsoon rainfall and moisture tracking in the Yangtze River basin. The findings reveal varying degrees of accuracy across different regions of the basin during the monsoon period. Downstream and midstream regions demonstrate higher accuracy, whereas upstream areas exhibit lower precision, along with an overall trend of overestimation. The evaluation encompasses the timing of monsoon months, as well as the peak month, while analyzing the simulation’s accuracy for rainfall. It also entails an overarching examination through a Taylor Diagram and Taylor skill scores, which spotlight models with superior and inferior performance. EC-Earth3 exhibits commendable performance, whereas models like IITM-ESM showcase poorer results. Furthermore, moisture tracking assessments, utilizing the WAM2layers model, identify limitations within the CMIP6 model in terms of replicating water vapor sources and pathways, especially in proximity to geographical features such as the Himalayas and the coastline. In addition to the basin itself, the CMIP6 model simulates central Asia as the main source of evaporation, rather than the Indian Ocean, according to the results of
ERA5. However, no obvious pattern differences are shown between the different CMIP6 GCMs. Persistent challenges stem from data availability and numerical inconsistencies, necessitating enhancements in both the CMIP6 models and the WAM2layers code.