Abstract
In previous studies, reanalysis data has been used instead of observational data for bias correction of GCM outputs in observation data scarce basin. However, reanalysis data does not match the real weather characteristics (e.g. precipitation, temperature) in a local scale since it is targeted at a global scale. Therefore, this study aims to develop a bias correction method for re-analysis data based on observational data. And also, we made an attempt to apply the method for two stations and to validate by splitting data into two periods, to clarify the applicable range. As a result, it is revealed that time-series pattern of precipitation in a reanalysis data can be corrected using three relations of monthly precipitation in reanalysis data and 1) number of precipitation events in each month, 2) average and 3) standard deviation of 3 hourly precipitation in observational data. Furthermore, despite the observational data with and without significance the trend, it is suggested that precipitation pattern of reanalysis data in validation period can be corrected by the developed statistical creation method of pseudo observational data in calibration period.
