Evaluation of different bias correction methods and Projection of Future Precipitation Changes Using GFDL-ESM4 model in Lake Urmia basin

The outputs of general circulation models (GCMs) usually have a bias compared to observational data, and some corrections must be made before using them to develop future climate scenarios. The bias correction methods are the standard statistical methods for processing the output of climate models. In this research, the effect of five bias correction methods on the projected precipitation of the GFDL-ESM4 model in the Lake Urmia basin has been evaluated. The methods used in this research include linear scaling (LS), local intensity scaling (LOCI), power transformation (PT), distribution mapping (DM) and delta change factor (DC). Statistical metrics such as the correlation coefficient, root mean square error (RMSE) and percentage bias (PBias) have been used to evaluate the accuracy of the corrected data in the period of 1990-2014 compared to the observational data and to choose the best method for correcting the data of future scenarios. research results showed that the delta change method significantly improved the raw estimates after correction; Therefore, this method was used to correct the data of scenarios SSP1-2.6, SSP2-4.5 and SSP5-8.5. In addition, the projection of the mean annual precipitation shows a decrease between 2 and 9 percent in SSP1-2.6, between 5 and 17 percent in SSP2-4.5, and between 8 and 26 percent in SSP2-8.5 compared to the observed data.