Understanding Surface Air Temperature Cold Bias Over China in CMIP6 Models

Abstract

The systematic simulated cold biases in models for surface air temperature (SAT) over China have been under discussion for a long time, but the related attribution is still unclear. In this study, we investigate the main contributors and relevant physical processes of SAT biases over China based on 31 models participating in the sixth phase of the Coupled Model Intercomparison Project (CMIP6) from a surface energy budget perspective. On annual and seasonal scales, less downward clear‐sky longwave radiation in the models, which is due to the underestimated tropospheric temperature and water vapor amount, is the main cause of cold biases over the Tibetan Plateau, Tarim Basin, Sichuan Basin and most of eastern China. Meanwhile, the simulation biases in surface albedo and sensible heat flux induce cold biases depending on regions and seasons. In winter and spring, the overestimated surface albedo in association with more snow cover leads to prevalent cold biases over the western Tibetan Plateau and Northeast China. Excessive sensible heat release caused by the overestimated ground–air temperature difference facilitates SAT underestimation in the Sichuan Basin during summer and autumn. In addition, the cold biases in winter and spring are larger than those in summer and autumn, due to the stronger snow cover overestimation and tropospheric temperature underestimation.

Difference in boreal winter predictability between two dynamical cores of Community Atmosphere Model 5

“…Numerous studies in climate modeling, including this study, have consistently reported cold bias characteristics in surface temperature across different seasons and regions (Chen et al 2017, Jun et al 2018, Fan et al 2020. This issue remains a subject of ongoing investigation, as also highlighted by Wang et al (2023), who noted that systematic cold biases in surface temperature simulation across various models have been a long-standing concern, with the underlying causes still unclear. Significant biases are also identified in atmospheric circulations.…”

Improving CMIP6 Atmospheric River Precipitation Estimation by Cycle‐Consistent Generative Adversarial Networks

“…Huang et al. (2023) proposed a convolutional neural network‐based model for bias correction of CMIP6 precipitation data based on monthly precipitation data. However, CMIP6 GCMs are based on climate simulation data, meaning they have different internal variabilities, and there is no one‐to‐one correspondence between observations and model simulations, and they are not synchronized at daily time scales (François et al., 2021; Hess et al., 2022, 2023; Pan et al., 2021).…”