Projections of the Earth's future climate at a finer scale are of great importance in climate-related studies. However, the typical spatial resolution of CMIP6 models is approximately 100 km, which is still not sufficient for resolving fine-scale orography, land cover and dynamics of the atmosphere, hindering their ability to simulate extreme weather and climate events. Dynamical downscaling method with a regional climate model is an important approach to obtaining fine-scale weather and climate information; whereas the traditional dynamical downscaling simulations are often degraded by biases in the global climate model (GCM). Therefore, GCM bias corrections have recently become an important topic in dynamical downscaling studies in the recent decade.

Recently, an article, published in Scientific Data, reported a novel GCM bias correction method. The method takes advantage of the non-linear long-term trend of ensemble mean of 18 CMIP6 models to reduce the uncertainty of future projection generated by a single GCM. Moreover, both the GCM mean and variance biases were corrected based on the ERA5 reanalysis data. Using this GCM bias correction method, the authors developed a set of bias-corrected large-scale forcing data with a grid spacing of 1.25 longitude by 1.25 latitude based on the ERA5 reanalysis and CMIP6 data. The bias-corrected dataset includes three surface variables and eight upper air variables for three sets of bias-corrected CMIP6 data, the historical data from 1979 to 2014, and SSP245 and SSP585 from 2015 to 2100.

"The bias-corrected GCM data shows much better quality than individual CMIP6 models and can provide high-quality large-scale forcing for dynamical downscaling projections of the Earth's future climate, atmospheric environment, hydrology, agriculture, wind power, etc," says Dr. XU Zhongfeng, the first author of the paper from the Institute of Atmospheric Physics, Chinese Academy of Sciences.

Fig. 1 Performance of the CMIP6 models and bias corrected data (MPI-ESM1-2-HR_bc) in simulating the climatological mean (1979-2014) of multiple variables against the ERA5 data. Lighter colours represent a better model performance. (Image by XU Zhongfeng)

Relevant references and data:

 
Xu Zhongfeng *, Ying Han, Chi-Yung Tam, Zong-Liang Yang, Congbin Fu, 2021: Bias-corrected CMIP6 global dataset for dynamical downscaling of the historical and future climate (1979–2100). Scientific Data, https://doi.org/10.1038/s41597-021-01079-3

Xu Zhongfeng*, Ying HAN, and Zongliang YANG, 2019: Dynamical downscaling of regional climate: A review of methods and limitations. Science China Earth Sciences 129. https://doi.org/10.1007/s11430-018-9261-5

Dataset: http://www.doi.org/10.11922/sciencedb.00487