APPLICATION OF TIME-FREQUENCY ANALYSIS METHOD FOR BIAS CORRECTION OF DAILY PRECIPITATION OUTPUT FROM CLIMATE MODELS WITH COMPARIING VARIOUS BIAS CORRECTION METHODS

摘要

This study focused on precipitation outputs from regional climate models and compared the properties of precipitation products which were obtained by using several bias correction methods. Result indicated future change of variance and continued heavy rainfall (i.e. 5 days total rain) could largely vary according to the choice of bias correction methods and regional climate models. This study also suggests a bias correction method which aims to correct temporal variation of daily precipitation obtained from the climate models. The proposed method showed relatively better performance in reproducing maximum 5 days total precipitation, which statistics would be important in evaluating the potential for water disasters. Result also indicated although spatial distribution of daily precipitation is improved by bias correction methods, the frequencies of each precipitation pattern are different according to regional climate models and employed bias correction methods.