基于WCA2D与SWMM模型的城市暴雨洪涝快速模拟

Rapid simulation of urban rainstorm flood based on WCA2D and SWMM model

为探讨计算高效的元胞自动机模型(WCA2D)与传统一维管网模型耦合的机制以及计算效果,尝试将WCA2D与SWMM模型耦合(SWMM/WCA2D),以广州市长湴片区为例探究一种暴雨洪涝快速二维模拟技术,对比实测积水数据以及SWMM/LISFLOOD-FP模拟,结果表明:SWMM/WCA2D模拟结果与"20180607"实测积水数据相近,表明模型精度良好;根据多指标评估结果,综合考虑主干渠道淤积以及建筑物阻挡情景的RTPR、RPPV、F1值分别达到0.8、0.6、0.7,模拟精度最高,最能反映区域实际情况;通过与SWMM耦合,WCA2D和LISFLOOD-FP的模拟结果差异小(最大水深差值基本低于0.1 m)、相关性强(相关系数基本超过0.7),但前者计算效率是后者的3~5倍,表明WCA2D能够耦合SWMM且计算效率更高,为复杂城市化地区暴雨洪涝快速模拟提供了一种新方法。

This study attempts to combine the cellular automata inundation model(WCA2 D) with SWMM(SWMM/WCA2 D) to explore the mechanism and computational efficiency of WCA2 D coupling the traditional one-dimensional pipeline network model. For this purpose,we developed a rapid flooding simulation technology with the Changban catchment(a typical catchment features complicated in Guangzhou) as a study case. By comparing with the measured inundation data and the simulation results performed by SWMM/LISFLOOD-FP model,we could obtain the following results:the simulation results are close to measured inundation data related to the "20180607" rainstorm,suggesting that the coupled model presents good precision. According to the results of model precision indices,the scenario considering the blockage of main pipe and buildings shows the highest simulation precision with the RTPR,RPPV and F1 of approximately 0.8,0.6 and 0.7,respectively,which best characterizes the actual conditions of the study area. By coupling with SWMM,the simulation results of WCA2 D and LISFLOOD-FP shows little difference(the maximum depth difference lower than 0.1 m) and higher correlation(the correlation coefficients almost larger than 0.7),however,the computational efficiency of WCA2 D is 3-5 times higher than that of LISFLOOD-FP,indicating that the WCA2 D could be feasibly and effectively combined by SWMM. Therefore,the coupled model(SWMM/WCA2 D) proposed by this study could provide a novel way to implement rapid urban flooding simulation in the complicated urbanized areas.