台灣常用洪水預報模式在沖淤劇烈河川之適用性研究（1/2）

Abstract

鑑於國內中央管河川於颱洪期間，河道之沖淤變化劇烈，河川水理模式（含降雨-逕流模組）預測水位不易且差異相當大，以致後續之預警與防災決策困難。為避免決策者發佈洪水警報過於頻繁，準確的預測降雨及洪流量實為相當重要之課題，因此洪水預報模式之準確性與適用性，成為幫助決策者的重要工具。本計畫之研究重點如下： 1. 相關基本資料之收集與彙整； 2. 河川斷面穩定及變化劇烈之分類： 3. 水文、水理洪水預報之比較； 4. 定床與動床數模之比較； 5. 河道沖淤劇烈變化對洪水位影響之評估； 6. 辦理洪水預報模式之技術移轉與教育訓練等事宜。 本計畫第一年度將對河川進行床形變化劇烈之分類；針對床形變化不大及劇烈之河川，評估颱洪期間結合降雨預測模式之預測降雨量，以SOBEK模式、淡水河模式、台大NewC水理模式、調適性網路模糊論系統（ANFIS）水文模式以及模式交大EFA定床模式模擬颱洪期間，此五種模式洪水預測之準確性及適用性進行評估。 第二年以交大EFA動床模式與第一年四種定床洪水預報模式，進行沖淤劇烈河川之洪水預測比較，評估其準確性與適用性。並檢討河道劇烈沖淤變化對洪水位的影響，最後將各洪水預報模式之結果做分析與評估後，辦理現況檢討與教育訓練的實用講習會，將技術移轉於水利署相關單位。

Abstract Due to the bed form of Central government rivers change much during the typhoon period in Taiwan, flood forecasting models do not forecast water level of channel easily and the result is not accurately. In order to avoid the government announce flood alarm frequency, to forecast rainfall and flood discharge accurately is very important study. So the accurately and applicability of flood forecasting model are important tool helpful to government. There are some goals for this project to present.

1. To collect and analyze basic information. 2. To assort the cross section of channel that is stable or rapidly migrating. 3. To compare the hydrology and hydraulic flooding forecasting models. 4. To compare the fixed bed and movable bed models. 5. To estimate the rapidly migrating channel to effect the flood water level. 6. To handle technology transfer of the flood forecasting models and education training.

First year, we will assort the cross section of channel that is stable or rapidly migrating. Then, use SOBEK model, Tanshui River model, ANFIS model, NewC model and EFA model to calculate the flood water level and estimate the forecasting results for these five models.

Second year, use EFA moveable bed model and fixed bed model at first year to calculate the flood water level at rapidly migrating river. Then, estimate the rapidly migrating channel to affect the flood water level and analyze the result. Final, devolve the technology of the flood forecasting models to Water Resources Agency.