Effect of the Horizontal Resolution of Climate Simulations on the Hydrological Representation of Extreme Low and High Flows

The advantages of dynamical downscaling at a finer resolution have been underestimated for hydrological simulations. The main objective of this study is to investigate the effect of the horizontal resolution of climate data on streamflow simulations in terms of low and high flows using flow-duration curves (FDCs). For this study, observed climate information and two downscaled climate datasets with different horizontal resolutions (5 km and 20 km) generated by the Weather and Research Forecasting (WRF) model forced by ERA-Interim reanalysis data are used. To generate streamflow, the Soil and Water Assessment Tool (SWAT) is used. This study focuses on three dam basins in the Han River basin, Korea. Although the annual precipitation in the Han River basin derived from WRF20 is higher than the observed precipitation and the results of WRF05, the magnitude and number of extreme events are underestimated compared to those of the observed and WRF05 precipitation. In addition, the annual maximum daily streamflow and 95% flow driven by the WRF05 exhibit notable improvements compared to those driven by WRF20. Although dynamical downscaling demands significant computational costs, the production of fine-scale climate simulation is necessary to improve the precipitation characteristics that can critically influence low and high flows in Korea in areas with complex geographical features.