基于FloodArea模型的成都主城区内涝风险评估

本文以成都主城区为例，运用气象数据、地理信息数据、社会经济统计数据及内涝灾情资料，通过多种常用分布函数的对比，选出重现期降水估算的最优函数，采用Pilgrim & Cordery法推求研究区的小时雨型，然后结合改进的基于FloodArea内涝模型，开展了24 h历时20、30、50、100 a一遇降水情景内涝模拟，并利用修订的内涝公路风险等级标准和财产损失曲线，探讨100 a一遇降水情景下内涝交通风险等级和居民室内财产损失风险。结果表明：①GEV（Generalized Extreme Value Distribution）分布函数是成都主城区重现期降水估算的最优函数；主城区24 h历时小时雨型呈双峰型, 且峰值出现在降水过程前部。②基于FloodArea模型，通过对输入数据或参数的改进，能够较好模拟城市内涝空间分布；各降水情景模拟结果显示高新南区、高新西区、青羊区内涝淹没范围占比相较其他地区偏高。③24 h历时100 a一遇降水情景内涝可造成成都主城区86.1%公路长度占比出行困难，其中一级风险公路长度占比为105%，二、三级风险公路长度占比分别为27.5%、28.4%，成华区内涝公路风险最高。④24 h历时100 a一遇降水情景内涝可造成居民室内财产潜在损失约占主城区GDP（Gross Domestic Product）的0.8%，其中武侯区财产损失风险最大，潜在损失占其GDP的1.6%。

Flood Risk Assessment in Main Urban Area of Chengdu Based on FloodArea Model

In recent years, the problem of urban waterlogging is becoming increasingly serious, and urban waterlogging risk assessment is becoming one of the hotspots and challenges in urban waterlogging disaster research. This article takes the main urban area of Chengdu as an example. Meteorological data, geographic information data, socio economic statistical data, and waterlogging disaster information are used. The optimal function for estimating precipitation during the return period is selected by comparing multiple commonly used distribution functions. The hourly rainfall pattern in the study area is calculated with the Pilgrim & Cordery method. Then, an improved FloodArea waterlogging model is developed to simulate waterlogging scenarios with a 24 hour rainfall period of 20, 30, 50, and 100 years. Based on the revised risk level standards for waterlogging highways and the revised loss curves for property, the levels of waterlogging traffic risk and the risk of indoor property loss for residents are discussed under the 100 year return period precipitation scenario. The results show that: (1) The GEV (Generalized Extreme Value) distribution function is the optimal function for estimating precipitation at the return period in the main urban area of Chengdu. The 24 hour hourly rainfall pattern in the main urban area of Chengdu presents a bimodal pattern, and the peak appears at the front of the precipitation process. (2) Based on the FloodArea model, the spatial distribution of urban waterlogging can be well simulated by improving the input data or parameters. The simulation results of various precipitation scenarios show that the proportion of waterlogging inundation areas in Gaoxin South Zone, Gaoxin West Zone, and Qingyang District is higher than in other areas. (3) The 24 hour 100 year rainfall scenario of waterlogging can cause 86.1% of the road length in the main urban area of Chengdu to be difficult to travel. Among them, the length of first level risk roads accounts for 10.5%, and the length of second and third level risk roads accounts for 27.5% and 28.4% respectively, with the highest risk of waterlogging roads in Chenghua District. (4) The potential loss of indoor property caused by waterlogging during a 24 hour 100 year rainfall scenario accounts for approximately 0.8% of the GDP (Gross Domestic Product) of the main urban area. Wuhou District has the highest risk of property loss, with potential losses accounting for 1.6% of its GDP. The evaluation results can provide support for the prevention and reduction of waterlogging disasters in Chengdu, and the established methods can provide technical reference for urban waterlogging risk assessment.