Reliable projections of regional climate are crucial for assessing climate risks and facilitating human adaptation to climate change. However, systematic biases in general circulation models (GCMs) can be transmitted to regional climate models (RCMs) through initial and lateral boundary conditions (ICs and LBCs), compromising the reliability of dynamical downscaling simulations. To address this issue, we developed a set of bias-corrected data based on CMIP6 outputs and used it to drive RCM simulations. Using the Weather Research and Forecasting (WRF) model at a 25-km resolution over the Asian–Western Pacific region for the period 1981–2014, we conducted three experiments: one forced by the original GCM outputs (WRF_GCM), one by the bias-corrected GCM data (WRF_GCMbc), and one by the ERA5 reanalysis (WRF_ERA). The results demonstrate that bias correction substantially improves the simulation of regional climate across multiple variables and timescales. For instance, the root-mean-square error (RMSE) decreased by 50%–90% for the mean climate and by 30%–60% for interannual to interdecadal variability. Furthermore, extremes in temperature and precipitation, as well as the simulation of tropical cyclones (TCs), were also markedly improved. These enhancements in WRF_GCMbc—spanning climatological mean states, interannual variability, climate extremes, and TCs—are attributed to more realistic large-scale circulation and sea surface temperature. These, in turn, improve downscaled precipitation, 2-m temperature, and TC representation through processes involving advection, radiation, and energy exchange. Our findings indicate that correcting biases in GCM-derived ICs and LBCs is an effective approach for enhancing dynamical downscaling simulations and projections of TCs and other high-impact weather events.
 

GCM bias correction, Dynamical downscaling, Cliamte variabilities, Climate extremes