近年来, 随着极端气候事件频发与城市化进程不断加快, 城市内涝的问题越来越严酷, 严重威胁到了居民的生命财产安全。城市内涝模拟是进行内涝预警与排水系统优化的重要技术手段, 其中暴雨洪水管理模型 (SWMM) 因其开源性与动力模拟能力被广泛应用。然而, 该模型通常依赖实测水文数据进行参数校准, 而多数中小城市受限于监测设施, 缺乏可靠实测数据, 制约了模型的精度评估与实际应用。为解决上述问题, 本研究以安徽省太和县城北片区为研究区域, 提出一种基于SAR影像辅助校准SWMM模型参数的方法, 为中小城市防洪规划与排水设施优化提供了技术支撑和方法参考, 采用SAR影像来计算双极化水体指数 (SDWI) , 并结合大津法 (Otsu) 提取实际淹没区域, 通过空间重叠分析与召回率、准确度等指标对模拟结果进行参数校准。利用SAR影像提取的淹没范围作为验证, 所构建的SWMM模型在典型区域召回率分别达到79. 35%和84. 69%, 准确率分别达到71. 89%和72. 53%, 这验证了模型的适用性与可靠性。随后开展5年、10年和20年重现期降雨情景下的内涝模拟, 模拟结果揭示, 随着降雨重现期的增加, 研究区淹没范围及中重度积水节点显著增加。通过对比分析积水范围空间分布与管网、节点超载情况, 系统地揭示了研究区的城市内涝积水特征。
In recent years, with the frequent occurrence of extreme climate events and the acceleration of urbanization, the problem of urban waterlogging has become increasingly serious, which seriously threatens the safety of residents' lives and property. Urban waterlogging simulation is an important technical means for waterlogging warning and drainage system optimization. Among them, the storm water management model (SWMM) is widely used because of its open source and dynamic simulation ability. However, the model usually relies on measured hydrological data for parameter calibration, and most small and mediumsized cities are limited by monitoring facilities and lack reliable measured data, which restricts the accuracy evaluation and practical application of the model. In order to solve the above problems, this study takes the northern area of Taihe County, Anhui Province as the research area, and proposes a method for calibrating SWMM model parameters based on SAR images, which provides technical support and method reference for flood control planning and drainage facilities optimization in small and medium-sized cities. The SAR image is used to calculate the Sentinel-1 Dual-Polarized Water Index (SDWI) , and the actual submerged area is extracted by Otsu method. The parameters of the simulation results are calibrated by spatial overlap analysis, recall rate and accuracy. Using the submerged range extracted from SAR images as a verification, the recall rates of the constructed SWMM model in typical areas reached 79. 35% and 84. 69%, respectively, and the accuracy rates reached 71. 89% and 72. 53%, respectively, which verified the applicability and reliability of the model. Subsequently, the waterlogging simulation under 5-year, 10-year and 20-year return period rainfall scenarios was carried out. The simulation results revealed that with the increase of rainfall return period, the inundated area and moderate to severe water accumulation nodes in the study area increased significantly. By comparing and analyzing the spatial distribution of water accumulation range and the overload of pipe network and nodes, the characteristics of urban waterlogging and water accumulation in the study area are systematically revealed.