Urbanization has become an irreversible trend and is even more rapid in floodprone area. However, the responses of both channel runoff and surface runoff to the urbanization and the underlying mechanisms still deserve more in-depth exploration within urban extents with limited data availability. Accordingly, we conducted a comprehensive quantitative analysis of the urbanization effects on channel and surface runoff from both statistical and model-based perspectives over three major urban agglomerations (i.e., Chengyu (CY), middle region of Yangtze River (MRYR), Yangtze River Delta (YRD)) in the Yangtze River Basin. The long-term high-resolution channel and surface runoff under different urbanization scenarios were simulated using both WRF-Hydro and SCS-CN models by integrating multiple remotely sensed, reanalysis, and in-situ datasets. Results demonstrated that the channel runoff increased by 4.4%, 5%, and 12.08% in CY, MRYR, and YRD urban agglomerations, respectively. Meanwhile, the urbanization also contributed to 30%, 21.4%, and 17.7% of the surface runoff changes in three urban agglomerations. Based on the high-resolution local climate zone (LCZ), our results highlighted that the high-rise and middle-rise buildings and compact-arranged regions were more prone to accumulating surface runoff and trigger flooding events due to the weaker infiltration capacity of land surface with the densely distribution of impervious surface. Further machine learning based investigation and case analysis also explored the potential influencing factors and mechanisms of the impact of urbanization on channel and surface runoff. Results in the current research show further evidence that policy makers of urban planning and management need to consider and ensure societal preparedness for extreme runoff and flooding events.

中文翻译：

---

城市化增强河道和地表径流：利用长江流域物理模型和实证模型进行定量分析

城市化已成为不可逆转的趋势，在洪水多发地区城市化进程更为迅速。然而，在数据有限的情况下，河道径流和地表径流对城市化的响应及其潜在机制仍值得在城市范围内进行更深入的探索。据此，我们从统计和模型的角度对成渝、长江中游、长三角三大城市群的城市化对河道和地表径流的影响进行了全面的定量分析。 （YRD））位于长江流域。利用WRF-Hydro和SCS-CN模型，通过集成多个遥感、再分析和现场数据集，模拟了不同城市化情景下的长期高分辨率河道和地表径流。结果表明，长三角、长三角和长三角城市群的河道径流分别增加了4.4%、5%和12.08%。同时，城市化对三个城市群地表径流变化的贡献率分别为30%、21.4%和17.7%。基于高分辨率当地气候区（LCZ），我们的研究结果表明，高层和中层建筑以及布局紧凑的区域由于土地入渗能力较弱，更容易积累地表径流并引发洪水事件不透水面分布密集的地表。进一步基于机器学习的调查和案例分析还探讨了城市化对河道和地表径流影响的潜在影响因素和机制。当前研究的结果进一步证明，城市规划和管理的政策制定者需要考虑并确保社会对极端径流和洪水事件做好准备。