基于精细地表信息建模的城市微地貌雨洪模拟研究

Intensive urbanization leads to perturbations of spatial distribution of atmosphere, water and heat around urban areas, which further induces storm rainfall over raining seasons. Flash floods happened in unprepared cities cause tremendous economic losses and even human casualties. Flood risk can be transformed to water resources via optimized urban planning and design based on thorough understanding of rainfall runoff and flood inundation models and effective mitigation measures. Available works developed in risk management, hydrology and climatology have been more effective on macro- and meso-scale, which are apparently not suitable for precise micro-scale inundation simulation over micro landscape units. This proposal is going to investigate micro-scale flood simulation by incorporating high resolution spatial data acquired from airborne and terrestrial platforms. A universal geospatial data model based on triangulated irregular network is designed to integrate all runoff-related geographic features including disparate land cover types, buildings, flyover bridges and high resolution characteristic terrain features. Assimilating algorithm is necessary for fusion of multiple data sources. Topology of water runoff flow paths over building rooftops and viaducts is vital to build an integrated runoff inundation model. Key algorithms for improved flow direction determination and depressions identification are also going to be investigated. Optimized numerical solvers are going to be designed for the shallow water functions. New interactive interfaces of 3D animated visualization of runoff simulation will be investigated. Scale effect will be analyzed between simulation accuracy and resolution of source data. Typical watersheds in Beijing urbanized area will be used for case studies and raw data collection was initiated last year. Two preliminary results were accepted in two international academic conferences.

高强度城市化对区域大气、水、热等周边自然环境要素具有强烈的扰动作用，促使夏季短历时暴雨频发，进而易形成洪涝灾害反作用于人口经济高度集中的城市。为了有效揭示小尺度城市环境下降雨径流和雨洪形成机理，优化城市设计，变“水害”为“水利”，开展雨洪模拟研究具有重要的理论意义和应用价值。本项目选择城市局部微水文地貌单元汇水区为研究区，研究多源高精度空间数据融合算法，首次提出面向城市雨洪模拟的全要素三角网地理空间数据统一描述模型；设计建筑物屋面和立交路面排水网络嵌入的水文几何拓扑结构和构建算法，研究针对不同地面覆盖的混合水流方向确定、真实洼地识别等关键算法；研究一体化雨洪模拟的优化数值求解模型，设计雨洪模拟高效三维动态可视化方法，并分析模拟结果与空间数据之间的尺度效应。本研究选择了北京城市区域典型汇水区为试验区，检验模拟过程和结果。本研究可以直接辅助提高宜居城市设计水平和水灾害风险抵抗能力。

抗风险能力（Risk Resilient）是智慧城市、宜居城市建设关键组成部分，其中，暴雨洪涝灾害是复杂城市系统高效管理、健康运行最大的挑战之一。目前，我国城市化在高速发展的同时，更加注重质量提升，预计到2030年将有10亿左右人口生活在城市里，城镇化率达到70%左右，因此在新区规划、设计、建设和管理中，探讨转“水害”为“水利（用于景观和绿化）”的信息化模拟具有重要的现实意义。另一方面，随着地理空间数据采集获取相关技术和装备的快速成熟，高精度时空大数据为城市洪涝风险模拟与可视化提供了基础。本项目以揭示城市微地貌水文单元中降雨径流形成和退水过程机理描述为目标，开展多源高精度基础地理空间数据融合处理，围绕区域内影响径流过程的土地覆盖类型、地物要素和地形等全部关联要素的高精度空间数据，以不规则三角网为基础，研究地理空间数据统一描述模型构建方法，并嵌入地表漫流、地下管网、建筑物屋顶和立交路面排水网络的几何网络拓扑结构，设计并实现一体化降雨径流过程模拟模型和工具。（1）设计实现了基于高精度点云数据的三维道路模型构建算法。（2）构建了数字表面高程模型（Digital Surface Model, DSM）混合格网高程的线性分解模型。（3）提出了适于城市地表汇流特征提取的高程模型构建算法。（4）设计实现了顾及地表形态的流向算法。（5）设计实现了基于Geohash多级空间剖分的矢量数据压缩算法。