Application of Bias- and Variance-Corrected SST on Wintertime Precipitation Simulation of Regional Climate Model over East Asian Region

Abstract

AbstractIn this study, the regional climate of East Asia was dynamically downscaled using Hadley Centre Global Environmental Model version 3-Regional Atmosphere (HadGEM3-RA) forced by the historical simulation data (1979–2005) of HadGEM2-AO produced by the National Institute of Meteorological Sciences (NIMS). To understand the impact of corrected SST on regional climate simulation, we integrated the experiments using uncorrected (UC_SST) and Bias- and Variance-corrected (BCVC_SST) HadGEM2-AO SST and used the simulated data driven by the ERA-Interim reanalysis data and HadGEM2-AO data. Examination of the spatial distribution, statistics, and interannual variation on wintertime precipitation over East Asia indicates that BCVC_SST reduced the overestimation of the climatological mean precipitation. In order to understand the impact of corrected SST on variability, we investigated the relationship between winter snowfall in South Korea and SST over East Asia. The negative correlation coefficient between the winter precipitation and the SST of the seas surrounding Korea appears in the result of observation data. The experiment result using BCVC_SST simulated the negative correlation between the winter snowfall and the SST around Korea more realistically than that of the simulations using UC_SST and HadGEM2-AO data. These results indicate that corrected SST helps to improve the variability of snowfall and SST simulated by HadGEM3-RA. However, time lag about the years when had peak point of SST appeared in the results compared between BCVC_SST experiment and observation data. The peak years shown in the result of the BCVC_SST experiment were similar to that of HadGEM2-AO data. At these results, even though the corrected SST improves climatological mean and variability of simulated data, it has the limitation not to overcome the error such as time lag showed in GCM SST. Additionally, the analysis of the snowfall in South Korea describes that SST is passively used for the source of snowfall and atmospheric variables mainly lead the intensity and the amount of snowfall.