耦合蚁群算法和SCS-CN水文模型的城市不透水面空间格局优化

近年来,中国城市暴雨内涝频繁发生,已经发展成为一类严重的“城市病”。城市不透水面密度及其空间格局是形成暴雨内涝的一个重要影响因素。本文提出一种耦合蚁群算法和SCS-CN水文模型优化不透水面空间格局的方法,从而实现通过增加地面雨水渗透量达到减缓城市内涝发生的目的。首先应用Williams公式计算基于坡度修正的CN值,在此基础上计算地表径流量;然后设定径流系数最小化目标,耦合水文模型和蚁群算法对径流小区尺度的不透水面空间格局进行优化配置;最后应用景观格局指数对不透水面空间格局进行分析。研究结果表明：面对1年、5年、10年、20年、50年以及100年一遇重现期的1 h持续降雨事件,研究区优化后的不透水面空间格局可以分别减少径流系数21.19%、19.58%、19.38%、18.93%、18.41%和17.25%,在一定程度上缓解城市暴雨内涝的发生。在此基础上,提出面向暴雨内涝防治的城市更新优化措施建议：① 通过增加草地、花园、树木等植被绿化减少高不透水面类型的面积,并划分成更多中高不透水面类型的斑块;② 集聚低、中低等不透水面类型,从而加大连通性,形成更多的中高不透水面类型;③ 增加每个径流小区内斑块数量,增大斑块密度,减少其蔓延度和聚集度。

Optimization of Spatial Pattern of Urban Imperviousness based on the Integration of SCS-CN Hydrological Model and the Ant Colony Algorithm

In recent years, the frequent occurrence of waterlogging in China has been one of the serious urban diseases. Spatial pattern of urban impervious surface density is an important factor affecting the waterlogging. This paper aims to provide a new method to optimize the spatial pattern of impervious surface in order to reduce urban waterlogging by the integration of SCS-CN model and the Ant colony algorithm. Firstly, the density of urban impervious surface was estimated by remote sensing images through the method of linear spectral mixing modeling. Secondly,the CN value was corrected by using the Williams formula. Then, the modified SCS-CN model was used to calculate the surface runoff. Thirdly,according to the goal of minimizing runoff coefficient, the spatial pattern of impervious surface of 18 runoff plots was optimized by Ant colony algorithm. Fourthly, landscape pattern indices were used to analyze the spatial pattern of impervious surface. The results show that: in rainfall reappearing periods of 1 year, 5 years, 10 years, 20 years, 50 years and 100 years, the optimized impervious surface pattern could reduce the runoff coefficient by 21.19%, 19.58%, 19.38%, 18.93%, 18.41% and 17.25%, respectively. Based on the experimental results above, this research puts forward three suggestions for the optimization of urban renewal. ① Increase the area of grassland, garden, trees and other vegetation types to reduce the high impervious surface area which can be further divided into patches of lower levels of impervious surface. ② Gather low and medium-to-low types of impervious surface to increase the connectivity and the medium-to-high levels of impervious surface. ③ Increase the quantity and density of patches in each runoff plot and reduce the degree of spread and aggregation.