Comparison of CMIP6 climate models and quantile mapping bias correction methods in the simulation of precipitation

Due to inherent limitations of global climate models, their outputs are significantly biased in comparison to observed values which could provide unreliable climate projections. This study evaluates the performance of 10 global climate models of the Coupled Model Intercomparison Project Phase 6 (CMIP6) for simulating precipitation in the Rafsanjan study area over calibration (1986-2005) and validation (2006-2014) period. For correcting simulated precipitation, various quantile mapping-based bias correction methods applied in these two periods. Evaluating the performance of various climate models and quantile mapping-based bias correction methods and approaches is carried out through multiple statistical metrics including NSE, PBIAS, MAE, and KGE as well as Taylor's diagram. Finally, simulated precipitation of selected model extracted for projection period under SSP1-2.6, SSP2-4.5 and SSP3-7.0 scenarios and corrected by suitable bias correction method. Results showed that the MPI-ESM1-2-LR model has better performance in simulating precipitation over calibration and validation periods compared to other climate models. The results of evaluating the performance of quantile mapping-based bias correction methods in both periods also showed that bernlnorm method performs better than others for the correction of simulated precipitation by climate models. In addition, the evaluation results of quantile mapping approaches including NTP, PT, and DDT in these periods demonstrated that NTP and PT have an acceptable performance compared to the DDT approach. Present study can help to improve the credibility of future climate projections using CMIP6 climate models.