空間分布雨量資料與進水口落葉雜物堵塞對區域淹水之影響
Journal
農業工程學報
Journal Volume
66
Journal Issue
4
Pages
1月13日
Date Issued
2020
Author(s)
Abstract
In recent years, the rainfall intensity and extreme weather events have been both increasing due to climate change. Therefore, the frequency of flooding disasters has been no doubt increasing. At present, the flooding simulation usually sets up a single rainfall station for model simulation. Although this method is faster and simpler to construct, the single rainfall station may not represent the spatial distribution of rainfall. In addition, road drainage inlets are used to collect surface runoff from streets and other land surfaces. If fallen leaves and debris are clogged at drain inlets, the drainage system is unable to perform the effectiveness, which may lead to flooding events. In this study, we discuss the effects of rainfall data from different spatial distributions of rainfall data on flooding simulations in urban areas. Comparisons of single rainfall station, Thiessen's Method rainfall setting, Thiessenʼs Method Partition rainfall setting, and Inverse Distance Weighted Method rainfall setting are made. Three indicators, including the methods of Probability of Detection (POD), Precision or Predictive Value (PPV) and Threat Score (TS), are used to evaluate the correctness of flooding simulations. Then, different degrees of inlet clogging are set up in the flooding simulation model, and analyzed which kind of clogging is most suitable for the actual flooding situation. The present study selects Jieshang and Hougang drainage systems of Shilin District, Taipei city, as the study cases. We use two short durations and one long duration rainfall events for simulation and analysis, and then we set up various clogging rates for flooding simulation as well. The results show that the Inverse Distance Weight Method is the best choice. 20% clogging rate has the best performance, which can be used as a reference for the clogging factor setting of urban areas.
Publisher
社團法人台灣農業工程學會
Type
journal article