Author(s): Jin-Cheng Fu; Ming-Hsi Hsu
Linked Author(s):
Keywords: River flood forecasting; Feedback routing; Tanshui River; Taiwan 1
Abstract: Taiwan is threatened by the disastrous storms that result from tropical cyclones or typhoon every year. The torrential rainfall of these storms, together with Taiwan’s special geographical features such as the steep slope and short length of its river systems, often result in the disastrous losses. The Tanshui River Basin flood forecasting system was used to guard against flood damage for the metropolitan city of Taipei in northern Taiwan. However, during typhoon Nari (2001), the inaccuracy results of flood forecasting were caused the excessive preparations in flood damage mitigation and a surfeit of civilians were evacuated (about 24, 000), these measures not only formed the local government’s economic loss but also decreased the populace to government's confidence. For solving this issue, providing a reliable river stages forecasting before planning proper actions is a critical task. Therefore, an attempt to improve the precision with real-time river stages was conducted in the study. Based on the dynamic wave theory of unsteady flow in open channels and a four-point implicit finitedifference method, a river flood forecasting model with real-time feedback routing was developed for the Tanshui River Basin. During processes of the real-time feedback routing, there are two updating steps using the observed river stages in the present time. In the first step, the least-squares method is employed to solve the over-determined problem because of the number of discritized equations adding the observed river stages is more than flow variables to update the water stage profiles and discharges as the flood forecasting initiation. As for the second step, a roughness updating technique taking observed stages as the targets was developed. The optimal method is used to find the new Manning n and to minimize the differences between observed and forecasting river stages in each time step of the routing processes. The typhoon Nari (2001) was used to examine the validity of the feedback routing. The rootmean-square of dimensionless water stage differences between the simulated and the observed are compared to evaluate the accuracy of the proposed model. The results show that the proposed model achieves better accuracy for 1~3-hr ahead predictions of river stages than 4~6-hr ahead predictions. The new approach of river flood forecasting model with real-time feedback routing really provides a useful tool for flood forecasting.
Year: 2007