In recent years, hydrological characteristics in Taiwan have changed considerably. The precipitation often exceeds the design rainfall in some watersheds. If heavy rainfall occurs in a densely-populated metropolitan area, a great amount of damage will be consequently produced. To make sure the decision maker can properly announce flood warning, accurate prediction of rainfall and river stage becomes an important issue.
This project aims at improving the accuracy of flood forecasting models to rapidly
migrating rivers in Taiwan during typhoon seasons. The evaluation on the five fixed-bed models: Dan-shuei river model, SOBEK model, NewC model, EFA model and HEC-RAS model in rapidly migrating rivers is made. In the first year of this project, testing and identifying five fixed-bed models in Dan-shuei River and Chou-shuei River by employing the above five models is fulfilled. In the second year of this project, testing and identifying EFA mobile-bed model and comparing the results with those generated by the five fixed-bed models will be executed. According to above results, the difference range of flood water level caused by the fixed-bed model applied to rapidly migrating river then can be evaluated. These results will provide better information for the flood-warning decision.