Abstract
The objective of this study is to develop a Korea Flash Flood Guidance (KoFFG) system over the Han River basin. For the purpose, threshold runoff of high resolution sub-catchment was estimated for each duration based on the delineated geometry from digital terrain elevation and the matching of surface runoff peak to bankfull flow. The model-based soil moisture status was calculated by the semi-distributed TOPMODEL. The stochastic logarithmic bias adjustment model by using Kalman filter scheme was used to minimize the large scale bias for optimal radar rainfall estimates over the region. FFG were computed by the meteorologic and hydrologic components during July of 2004, and the case studies for the possibility of future flash flood threat were performed with the forecasted precipitation of RDAPS (5 km) model. The real-time KoFFG system was designed and developed with pre-processing, core computation module, and post-processing. The case studies were performed with historical data for the flash flood occurrence. It is shown that the developed KoFFG system is useful to forecast a warning of flash flood by using radar data and the accuracy of RDAPS would be needed to increase for the forecast of future flash flood threat.