Assessment of Sponge City Flood Control Capacity According to Rainfall Pattern Using a Numerical Model after Muti-Source Validation

Autor: Haichao Li, Hiroshi Ishidaira, Yanqi Wei, Kazuyoshi Souma, Jun Magome
Jazyk: angličtina
Rok vydání: 2022
Předmět:
Zdroj: Water, Vol 14, Iss 5, p 769 (2022)
Druh dokumentu: article
ISSN: 2073-4441
DOI: 10.3390/w14050769
Popis: Urban floods are a common urban disaster that threaten the economy and development of cities. Sponge cities can improve flood resistance ability and reduce floods by setting low-impact development measures (LID). Evaluating flood reduction benefits is the basic link in the construction of sponge cities. Therefore, it is of great significance to evaluate the benefits of sponge cities from the perspective of different rain patterns. In this study, we investigated the urban runoff of various rainfall patterns in Mianyang city using the Strom Water Management Model (SWMM). We employed 2–100-year return periods and three different temporal rainfall downscaling methods to evaluate rain patterns and simulate urban runoff in Mianyang, with and without the implementation of sponge city measures. After calibration, model performance was validated using multi-source data concerning flood peaks and inter-annual variations in flood magnitude. Notably, the effects of peak rainfall patterns on historical floods were generally greater than the effects of synthetic rainfalls generated by temporal downscaling. Compared to the rainfall patterns of historical flood events, the flood protection capacities of sponge cities can be easily overestimated when using the synthetic rainfall patterns generated by temporal downscaling. Overall, an earlier flood peak was associated with better flood sponge city protection capacity. In this context, the results obtained in this study provide useful reference information about the impact of rainfall pattern on urban flood control by LID, and can be used for sponge city design in other part of China.
Databáze: Directory of Open Access Journals