面向水文模拟的空间离散指将地表划分成独立的计算单元，在单元内部计算产流，单元之间计算汇流，它是水文模型的构建基础。不同于流域水文模型主要基于自然地形的离散化（如使用自然子流域），城市水文模型离散需要考虑地形与人工设施的双重影响。目前基于手动的空间离散方式不仅只能有限的描述城市地表水流过程，而且其过低的离散效率直接导致传统流域水文模型重点研究的一些核心问题（如区域、数据与模型间的相互制约关系）很难在城市区域开展。本项目拟构建面向不同结构城市水文模型的地表空间自动离散方法，重点研究综合自然及人工水流通道的城市汇水网络的生成方法、基于空间自动离散方法的多重精细化程度的城市水文模拟、空间离散单元与城市水文模拟系统的耦合方法等。在此基础上，研究面向城市水文模拟的最优数据尺度选择策略。课题研究可有效提高城市地表空间建模的效率与质量，促进多源多尺度数据在城市水文模型中的应用，从而推动城市水文模型的发展。

Spatial discretization for hydrological simulation means to divide the surface to individual computation units, calculate runoff generation for each unit and simulate flow among units, which is the foundation of the urban hydrological model. Different to the spatial discretization of river hydrological model, which is mainly based on the natural surface topography, urban hydrological must consider the impact of both natural surface topography and artificial facilities. The manual spatial discretization can only describe the urban surface water movement with a limited accuracy, and more importantly, due to its low efficiency, some core issues in traditional river hydrological model (e.g. the restrictive relation among study area, data and model) are hard to study in urban areas. For this reason, this project aims to design an automatic urban surface spatial discretization method for urban hydrological simulation with different model structures. The main studies include the method to generate urban catchment network with consideration of natural and artificial water channel, urban hydrological simulation with different complexities based on the automatic urban surface spatial discretization method, and the combined way between the discretization unit and the urban hydrological simulation system. Based on this, we will analyze the best data scales for urban hydrological simulation. This project can considerably improve the quality and efficient of urban surface modeling, promote the application of data from different sources with different scales on the urban hydrological model, and contribute to the development of the urban hydrological model.