摘要
为精确分析小流域暴雨洪水坡面汇流过程,以湖南省宝盖寺小流域为研究对象,基于高精度激光雷达点云数据,构建二维水动力学模型,对2012年场次洪水过程进行模拟,并将模拟结果与分布式水文模型模拟结果和实测结果进行对比分析.同时为解决网格数量多、运算速度慢的问题,该模型采用GPU并行算法.研究结果表明:1)采用的高分辨率地形数据,可以较好地模拟小流域坡面降雨径流过程,且可得到整个流域上的洪水汇流过程;2)二维水动力学模型的计算结果和分布式水文模型计算结果与实测值吻合较好,水动力学模型计算洪峰出现的时间误差1 d,洪峰流量误差为23%,且可应用于高分辨率地形数据;3)采用的GPU并行算法,可使模型的计算效率提高近50倍,使得计算时间小于小流域汇流时间.该模型可用于暴雨山洪的预报预警和三维推演工作.
To analyze the formation of confluence on a slope due to storm floods in Baogaisi,a small watershed in Hunan Province,a two-dimensional hydrodynamic model was constructed from high-precision LIDAR point cloud data.The hydrodynamic model was used to simulate flood events in 2012.The simulation was compared with the hydrological simulation,and the actually measured data.To solve the problem of large number of grids and slow computing speed,GPU parallel algorithm was applied to the model.It was found that with high-resolution topographic data,the rainfall runoff process on the slope of small watershed could be simulated rather well,obtaining the flood convergence process in the whole basin.The hydrodynamic simulation was in good agreement with the hydrological simulation,and with the measured data.Peak time error and flood peak error as calculated by the hydrodynamic model were found to be 1 h and 23%respectively.The hydrodynamic model could be applied to high resolution terrain data.The use of GPU parallel algorithm was found to improve computation efficiency by 50 times,shortening the calculation time significantly.This model could be used for forecasting,warning and three-dimensional deduction of storm floods.
作者
魏永强
刘昌军
赵伟明
WEI Yongqiang;LIU Changjun;ZHAO Weiming(Hunan Institute of Water Resources and Hydropower Research,410007,Changsha,Hunan,China;China Institute of Water Resources and Hydropower Research,100038,Beijing,China)
出处
《北京师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2020年第6期838-845,共8页
Journal of Beijing Normal University(Natural Science)
基金
湖南省水利科技资助项目(湘水科计[2016]194-33)
创新团队资助项目(JZ0145B2017)
湖南省重大水利科技计划资助项目(湘水科计[2017]230-35)
广西重点研发计划(桂科AB19245054)。
