近年來,由於氣候變遷與都會區的急速發展而造成熱等效應的影響,使得氣溫不斷的上升,加上建築物與不透水面積的增加,使得雨水不易排出而造成生命財產的損失,然而藉由雨水下水道的建置能避免都市積淹水的情況,因此藉由不同淹水模擬模式的評估可以事先做好防災準備以降低災害造成的損失。國內外學者針對都會區淹水已發展出許多種之方法與模式,而本研究將藉由三種模式探討短延時降雨事件對地表淹水雨水雨下水道水位之影響,地表淹水以準確度與偵測率來評估模式結果,下水道則檢驗其水位歷線與實測值之差距,此三種模式分別為先以雨水下水道收集雨水再由人孔溢流到二維淹水模式之分離式演算模式(模式A)、先以雨水下水道收集雨水系統且可與二維淹水模式迭代之半交互演算模式(模式B) 及以二維地表收集雨水且可與雨水下水道迭代之交互演算模式(模式C)。 本研究以雙園排水系統為研究區域,其橫跨台北市萬華區、中正區及大安區三區,並以2009年8月12日與2012年8月12日之短延時強降雨事件做為淹水模擬事件,在地表淹水部分以模式C有較佳之模擬結果,其準確度在99%以上,偵測率亦在60%以上;下水道水位方面,模式A和B有高估的情況,模式C則低估,但以模式C之結果較為接近。
In recent years, the temperature has been rising due to climate change and rapid development of urban cities. Heavy rainfall often cause severe damages to lives and properties due to the increase of impervious areas and buildings which leads to a difficult discharge of rainwater; however, by constructing rainwater sewer systems in urban areas, we can avoid such flooding situations; therefore, we can evaluate different flooding situations by using simulation models and thus be prepared for disasters and reduce losses. Researchers around the world have developed many kinds of methods and models to deal with urban inundation, and in this study, I will use three different rainwater collection system models to evaluate the impacts of short duration rainfall events on surface flooding areas and sewer water levels. Surface flooding areas will be evaluated by “Accuracy” and “Probability of detection(POD)”; simulated sewer water levels will be compared to the measured datas. The three models include: the storm sewer systems model combining with two-dimensional overland flow model by using separated algorithm (model A), the storm sewer systems model combining with two-dimensional overland flow model by using half-interactive algorithm (model B), the storm sewer systems model combining with two-dimensional overland flow model by using interactive algorithm (model C). In this study, the Shuang-wuang drainage system is selected as the study area, which spans the Wanhua District , Zhongzheng District and Da-an District. Rainfall events in August 12, 2009 and August 12, 2012, are selected for the simulation. In terms of surface flooding, model C has the best results, with an accuracy above 99 % and POD above 60% ; while in sewer water levels , model A and B are overestimated compared to measured datas. Model C is a bit underestimated but has the closest results.