基于CADDIES-2D模型的北京城区暴雨洪涝模拟及验证分析

Flood Simulation and Verification in Beijing Urban Area Based on CADDIES-2D Model

针对极端降水引发的洪涝灾害严重阻碍城市发展的问题,利用CADDIES-2D洪水模型,基于不同年份或不同空间分辨率的输入数据,模拟了北京五环内城区在2012年"7·21"特大暴雨降雨后的内涝情况,并将其结果与FloudArea模型和BUW模型结果进行比较。结果表明,分别基于2010、2017年的10m分辨率土地利用数据进行模拟,平均误差均为0.10m,模拟精度高且鲁棒性强;基于10、20、30m分辨率输入数据的模型模拟平均误差分别为0.10、0.16、0.18m,随着数据粒度变粗尽管模型存在一定精度损失,但结果仍较为可靠,且运行时间大幅降低;CADDIES-2D模型模拟精度优于FloodArea模型、稍逊于BUW模型,但相比于BUW模型,CADDIES-2D模型对地下管网资料要求低且在不同地区的迁移性较强,在缺乏精细建模资料时依然可提供较为准确的二维洪水模拟结果。可见CADDIES-2D模型鲁棒性强、运行效率高且对建模资料要求低,研究成果可为城市雨洪模拟与预警提供参考。

The flood disaster caused by extreme rainfall seriously hinders urban development.Based on the data of different years or different spatial resolutions,the CADDIES-2 Dflood model was applied to simulate the waterlogging situation of the inner urban area of Beijing after the heavy rainfall event on 21 July 2012,and compared the results with other flood models.Based on the land utilization data of 2010 and 2017 with 10 mresolution are quite close,an average error is 0.10 m,which shows high accuracy and strong robustness of the model.The average error of the model based on the data of 10 m,20 mand 30 mresolution is 0.10 m,0.16 mand 0.18 m,respectively.As the data resolution becomes rough,the model running time is greatly reduced with compromise of accuracy,but the results are still reliable.The simulation accuracy of the CADDIES-2D is better than that of the Flood Area model and slightly worse than that of the BUW model.However,compared with the BUW model,the CADDIES-2D model has lower requirements for underground drainage network data so that it has stronger extensibility in different areas.The CADDIES-2D can provide relatively accurate flood simulation results even lack of fine modeling data.To sum up,the CADDIES-2D has strong robustness,high calculation efficiency and low requirements for modeling data.The results can provide a reference for simulation and early warning of urban flood in Beijing.