Assessment of sewer flooding model based on ensemble quantitative precipitation forecast

Short duration rainfall intensity in Taiwan has increased in recent years, which results in street runoff exceeding the design capacity of storm sewer systems and causing inundation in urban areas. If potential inundation areas could be forecasted in advance and warnings message disseminated in time, additional reaction time for local disaster mitigation units and residents should be able to reduce inundation damage. In general, meteorological–hydrological ensemble forecast systems require moderately long lead times. The time-consuming modeling process is usually less amenable to the needs of real-time flood warnings. Therefore, the main goal of this study is to establish an inundation evaluation system suitable for all metropolitan areas in Taiwan in conjunction with the quantitative precipitation forecast technology developed by the Taiwan Typhoon and Flood Research Institute, which can be used for inundation forecast 24 h before the arrival of typhoons. In this study, information for the design capacity of storm sewer throughout Taiwan was collected. Two methods are proposed to evaluate the inundations: (a) evaluation based on the criterion of sewer capacity (CSC), and (b) evaluation based on the percentage of ensemble members (PEM). In addition, the probability of inundation is classified into four levels (high, medium, low, and no inundation). To verify the accuracy of the proposed system, Typhoon Megi and Typhoon Nanmadol were used as test cases. Four verification indices were adopted to evaluate the probability of inundation for metropolitan areas during typhoons. The inundation evaluation results basically match the observed data on flooding, which demonstrate that this flood evaluation system has an effective grasp on the probability of inundation for storm sewer systems.