BIAS CORRECTION OF RUNOFF DATA IN AGCM3.2S FOR UPPER CHAO PHRAYA RIVER BASIN, THAILAND

Abstract

This study investigates the applicability of bias correction for runoff generation data from AGCM3.2s over the northern part of Thailand. A land surface model generates 20 years of reference runoff data after being carefully calibrated and validated by observation discharge to ensure the performance of the runoff reference data. Two simple bias correction methods, namely linear scaling factor and empirical quantile mapping bias correction, were tested to reduce biases. The linear scaling factor method performs better in reducing bias than the empirical quantile mapping in the monthly mean of runoff generation for each grid and river discharge. In contrast, both bias correction methods are not effective in adjusting to the quantile of monthly river discharge. This study provides an understanding of the characteristics of biases in GCM runoff and leads to the development of new bias correction methods.