每當颱風來襲時,經常降下大量與高強度之豪雨,降雨的特性常造成臺灣許多災害,並導致複合性的災害之發生。因此如何即時且有效率地提供淹水預警及雨量預報,讓防災應變人員能即早進行決策判斷,在臺灣防災管理是一重要課題。但由於水情資料來源多且複雜,在臺灣現有水情預警系統中,常需經過繁瑣之演算分析程序及操作,才能提供防汛應變人員預警資料。為整合預警之流程及提高產生預警圖資之效率,本研究開發一淹水預警作業資訊平臺,有效地整合及介接雨量預報資料及淹水模式,以減少大量的操作時間及人力成本,並提供即時的視覺化展示介面,以供防汛應變人員可進行最新淹水預警模擬資料之查詢及比對,讓防汛應變人員縮短資訊介接及彙整所需之時間,達成即時決策之目的。本系統亦定期且自動地產生模擬圖資,提供防汛應變人員進行歷史預警資料及真實災情之比對,作為改善及加強未來防災策略之參考依據。本研究實際進行原有操作與系統實作之預警流程比較,結果顯示使用淹水預警作業資訊平臺,有效地大量節省了原有作業流程所需花費的時間,且降低了人為疏失導致模擬資料錯誤的風險,因此確實能增進防汛應變及決策的效率。
Whenever a typhoon strikes with torrential rainfall, it tends to bring exceeding amount and high intensity precipitation in a very short period of time. With such condensed rainfall in a short period of time, it may bring catastrophes and result in compounded disasters. For the disaster prevention countering personnel to perform early decision making judgment, instantaneous and effective flooding alert and precipitation forecast become a vital issue in Taiwan. However, the sources of water information are many and complex and that they often require complicated procedures and operation within the current pre-warning system in Taiwan to provide early warning information for the flood prevention countering personnel. In order to integrate the pre-warning procedures and elevate the efficiency in generating pre-warning graphic data, this study develops an Operational Information Platform for Inundation Forecasting to integrate and interconnect the precipitation forecast data and inundation forecasting. This operational platform can reduce the massive operation time and the cost of manual labor, which provides visualization interface for the flood prevention countering personnel to conduct the latest flood pre-warning simulation data inquiries and comparison. This platform also cut the time required in data connection and compiling to achieve the goal of real-time decision making. It runs periodically and automatically to generate the simulation graphic data to provide the flooding prevention countering personnel to conduct the comparison between archived pre-warning data and actual disaster situation to improve future disaster prevention strategies. This study also conducts the comparison between the original operation and the present system operation on the pre-warning procedures. The results indicate that the platform reduces the time that is normally required by the original semi-automatic early pre-warning procedures as well as the risk of simulation data error resulted from the human negligence. It is proven that this platform increases the efficiency of flood inundation forecasting in flood prevention contingency and decision making.