Author(s): Seon-Ho Kim; Deg-Hyo Bae
Keywords: Climate Change; Drought; Impact Assessment; Analogue Method; Statistical Downscaling
Abstract: In recent decades, climate change has caused unusual extreme climates all over the world. Therefore, climate change impact assessments have been implemented based on global circulation models. However, there is a significant limitation in the representation of local climate variations in the spatial scale of the global circulation model. To address the scale mismatch, the analog-based downscaling method has commonly been used. But there are few studies to find optimal predictors of analog downscaling in South Korea. And it is difficult to find an optimized downscaling approach for extreme climate. In this study, optimal predictor variables were searched that are suitable for downscaling daily precipitation in cases of normal and drought periods in South Korea and elaborated on the added value of the downscaling method for drought periods. As a result of the predictor search, the combinations of precipitation and circulation were most practical for normal and drought periods. However, circulation variables or pressure levels suitable for normal and drought periods were explored significantly differently. The predictor sets presented in this study can be used not only for analog downscaling but also for other downscaling approaches. The downscaling results, which optimized drought period, was compared to results of conventional method in calibration period and the suggested method showed added accuracy than conventional method at many stations in drought periods. In validation results, suggested method also showed added accuracy in some stations, but stability of results is relatively lower than conventional approach. In future works, the stability of this methodology should be improved and its applicability to climate change scenarios should be explored.