Regional disparities in summer precipitation variability over Northern China and the role of anthropogenic forcing from 1961 to 2014

Abstract

Based on CN05.1 dataset and Coupled Model Intercomparison Project 6 (CMIP6) model outputs under different forcing experiments, this study systematically evaluates model performance, analyzes sources of simulation biases, and performs detection and attribution analyses of summer precipitation across northern China from 1961 to 2014. Results indicate that most CMIP6 models can reasonably reproduce the dipole pattern of precipitation, with 28 models achieving a comprehensive Taylor skill score above 0.6. However, notable discrepancies exist between simulated and observed precipitation trends. While the models effectively replicate the increasing trend observed in the western part of northern China (WNC), they fail to simulate the decreasing trend observed in the eastern part of northern China (ENC), instead exhibiting an opposite increasing trend. Analysis of these discrepancies reveals significant model deviations in simulating key atmospheric circulation systems and their associated moisture and dynamic conditions influencing precipitation. Additionally, the models inadequately reproduce the predominant roles of evaporation and horizontal thermodynamic processes on precipitation highlighted by observation, contributing to the divergence in simulated versus observed precipitation trends. Further detection and attribution analyses demonstrate that the signal of anthropogenic greenhouse gas (GHG) forcing is detectable in the increasing precipitation in WNC and can be distinguished from other external forcings. However, model deficiencies in reproducing precipitation trends in ENC prevent the detection of any external forcing signals in this region.