Enhanced Driving Data for Regional Climate Models: Investigating the Systematic Improvements with GCM Run-time Empirical Bias Correction

A novel runtime empirical bias correction (EBC) has recently been developed and applied to enhance the Canadian Center for Climate Modelling and Analysis' (CCCma) global earth system model CanESM, demonstrating significant improvements in future climate projections, particularly under strong climate change scenarios. The application of EBC to CanESM provides enhanced driving data for dynamical downscaling through regional climate models (RCMs).

This project aims to assess the impact of the improved EBC driving data on two RCMs, namely CanRCM5 (CCCma) and CRCM5 (Ouranos), in order to evaluate the systematic improvement of meteorological variables. Multiple 10-member ensembles are utilized to investigate the added value of employing EBC in driving the RCM simulations. The ensembles consist of three sets: the first set utilizes the original CanESM5 as driving data, the second set incorporates EBC on sea surface temperature (SST) and sea ice concentration (SIC) using the original CanESM5, and the third set employs bias-corrected atmosphere, SST, and SIC data. All three ensembles are compared against ERA5 data as a reference for the historical period.

Results indicate a clear advantage of using EBC, particularly in cases where the initial bias is substantial. For instance, significant improvement in modeling key meteorological phenomena, notably the North American monsoon and the northeasters (extratropical cyclones). These improvements can be attributed not only to the refinement in addressing climatological biases in land and ocean data but also to an enhanced representation of cyclonic activities due to a better representation of overall circulation in our region. Ultimately, this research seeks to contribute to the scientific community by providing a methodology to mitigate uncertainties in downscaled projections of future climate change through the utilization of EBC.