상하류 수위 특성을 기반으로 한 홍수예경보 기법

For flood forecasting and warning in rivers, it may be a better way to use observed water levels between upstream and downstream,instead of using the rainfall-runoff models such as the storage function method, to minimize the error involved in floodforecasting. In addition, the advanced time should be acquired to prepare the disaster mitigation action to minimize flood damages.
For this purpose, in this study, we suggest a flood forecasting and warning methodology which is able to predict downstreamwater levels at the point of flood forecasting in short time period, based on the currently observed upstream water levels.
Applying the Artificial Neural Network to the currently observed upstream water levels, we can predict water levels at a floodforecasting region which may occur within 30 minutes. After the suggested method is applied to the upstream Nam-gang watershedin the Nakdong-River basin, it is concluded that the method can predict downstream water levels in certain accuracy and willbe used as a flood forecasting and warning system in the region.