城市暴雨内涝水文水动力自适应耦合建模方法研究

The hydrologic and hydrodynamic processes of urban water logging are simultaneously occurred on the same urban surface in terms of water flow movement, so the hydrologic process and the hydrodynamic process are tightly coupled with very complex relationship. According to the drawback including relatively low elaborate level and poor prediction accuracy of the traditional urban rainstorm water logging model, by which the hydrologic and hydrodynamic processes are still not accurately described, the methods of data analysis, theoretical modelling, numerical simulation and case study are adopted in this work, and the response laws of some key hydrological parameters to the local inundation state are studied. New formulas for computing the outlet flow of the drainage area in different positions in terms of point-, line-, and area- boundaries are derived. Then the adaptively coupled hydrologic-hydrodynamic mechanisms for urban rainstorm water logging are meticulously described. Considering the bi-directional coupled effect between the hydrologic model and hydrodynamic model, various coupling approach are proposed to develop a new urban rainstorm water logging model. Different computing patterns, including the mesh up-scaling, first-order accurate scheme, grid dynamic control, local time stepping and GPU high performance computing, are introduced for resolving the two dimensional shallow water equations. And then a high-speed computing method for simulating the urban rainstorm water logging is developed. The expected research results of the project should provide important basis to improve the national capability on pre-warning of urban rainstorm water logging, as well as to enhance the reliability of model application, thus are of momentous scientific significance and practical application value.

城市地表是暴雨内涝水文与水动力过程的共同载体，水流运动是水文与水动力过程的共同形式，故暴雨内涝水文与水动力过程存在极为复杂的耦合关系。本项目针对已有城市暴雨内涝模型尚不能准确刻画水文水动力过程，且精细化程度与预测结果精度较低的问题，通过资料分析、理论建模、数值模拟和实例研究，解析子汇水区关键产汇流参数对汇水区内局部淹没状态的响应关系，推导汇水区出口流量在点状、线状、面状等不同形式耦合位置的计算公式，阐明城市暴雨内涝水文水动力自适应耦合机理，提出不同形式耦合方法，建立水文与水动力双向耦合的城市暴雨内涝数学模型，引入网格尺度粗化、低阶精度格式、网格动态控制、局部时间步长、GPU高性能计算等模式，发展城市暴雨内涝水文水动力耦合模型高速计算方法。本项目研究成果对于提升我国城市暴雨内涝预测预警能力、增强模型应用的可靠性具有重要科学意义和实际应用价值。

本项目针对已有城市暴雨内涝模型尚不能准确刻画水文水动力过程，且精细化程度与预测结果精度较低的问题，通过资料分析、理论建模、数值模拟和实例研究，解析了子汇水区关键产汇流参数对汇水区内局部淹没状态的响应关系，推导出汇水区出口流量在点状、线状、面状等不同形式耦合位置的计算公式，阐明了城市暴雨内涝水文水动力自适应耦合机理，提出了不同形式耦合方法，建立了水文与水动力双向耦合的城市暴雨内涝数学模型，并通过引入网格尺度粗化、低阶精度格式、网格动态控制、局部时间步长、GPU高性能计算等模式，发展了城市暴雨内涝水文水动力耦合模型高速计算方法。相关成果已应用于广州市典型区域洪涝实时预报预警，在2021年防暴雨应急中得到成功检验。本项目研究成果对于提升我国城市暴雨内涝预测预警能力、增强模型应用的可靠性具有重要科学意义和实际应用价值。