大风是造成城市灾害的主要气象要素之一，近年来大风吹倒树木引起车辆受损、交通受阻、人员受伤等灾害事件频发，开展大风灾害导致的行道树损失风险研究对于减轻城市损失、提升防灾能力具有重要的意义。然而，目前对大风灾害行道树损失的定量风险研究工作较为少见。本文选取经济发达、路网密集、易受大风影响的上海市作为研究区域，在开展大风危险性建模、行道树脆弱性分析的基础上，定量评估了大风行道树损失风险。
本文采用上海市30个气象站点2005–2019年的小时极大风数据，分析了大风活动危险性的时空格局变化规律及大风过程分布特征，通过日极大风频率和风速强度进行统计拟合，构建了上海市各区大风危险性评估建模，并结合2007–2019年历史灾情数据，获取大风强度—行道树受损数据对，拟合了上海市大风致行道树受损脆弱性曲线并进行了行道树脆弱性评估。在此基础上，建立了上海市大风行道树受损定量评估模型，并计算得出上海市各区大风行道树损失重现期曲线，绘制了不同年遇型行道树损失分布图。结论如下：
（1）上海市近年来大风日数、大风过程明显趋于减少，时间上以春季大风和夏季大风为主，空间上以东南沿海地区和东北沿江地区居多，内陆地区偏少。上海市行道树受损较为严重的年份为2012年和2019年，灾情主要出现在8月份，发生灾害的时间段主要集中在白天，灾情空间上呈现东部区域数量较多，西部区域数量相对较少的特征。
（2）上海市各区风速出现频率随着风速强度的增大而减小，以幂函数作为风速频率和风速强度关系拟合效果最好。各区大风致灾因子危险性存在较大差异，呈现沿江沿海大于内陆的分布规律。整体来看，上海市各区不易发生日极大风速超过17m/s（八级风）的大风事件，但各区都存在小概率、高危险性的极端大风事件的可能性。
（3）利用大风强度–行道树损失率数据，利用幂函数分别构建上海市市区和郊区行道树脆弱性曲线模型，市区和郊区的脆弱性拟合方程R2分别达到0.930和0.934。
（4）上海市各区行道树期望损失随着重现期的增大而逐渐变大，空间上呈现沿江沿海地区大于内陆地区、内陆市区大于内陆郊区的特点，不同年遇型单位面积行道树损失情况总体呈现东部地区高于西部地区分布特征。10a一遇，各区行道树损失基本在3棵以下；50a一遇，各区行道树损失基本在15棵以下；100a一遇，各区行道树损失基本在70棵以下，其中浦东新区最多，损失为69棵；300a一遇，各区行道树损失基本在1500棵以下，浦东新区最多为1464棵。总体来看，沿江沿海区域损失较大，其中浦东新区行道树损失风险最大，宝山区的单位面积行道树受损最大，由于沿江沿海各区的危险大、行道树总量多、浦东新区、宝山区脆弱性也较高而导致，但对于各影响因素的权重有待后续进一步研究。

Gale is one of the main meteorological factors causing urban disasters. In recent years, trees blown down by gale cause frequent disaster events such as vehicle damage, traffic obstruction and personal injury. It is of great significance to carry out risk research on street tree loss caused by gale for reducing urban losses and improving disaster prevention ability. However, the quantitative risk study on the loss of street trees caused by gale is rare. In this paper, Shanghai, a city with developed economy, numerous roads and vulnerable to gale, is selected as the research area. Based on the risk modeling of gale and vulnerability analysis of street trees, the loss risk of street trees caused by gale is quantitatively evaluated.
In this paper, hourly maximum wind data from 30 meteorological stations in Shanghai from 2005 to 2019 were used to analyze the spatial and temporal patterns of gale activity risk and the distribution characteristics of gale process. Statistical fitting was carried out by daily maximum wind frequency and wind speed intensity to build gale risk assessment models in each district of Shanghai. Combined with the historical disaster data from 2007 to 2019, the damage data pair of gale intensity and street tree was obtained, and the vulnerability curve of gale street trees in Shanghai was fitted and the vulnerability assessment of street trees was conducted. On this basis, a quantitative evaluation model for the damage of street trees caused by gale in Shanghai was established, and the return period curves of street tree loss in each district of Shanghai were calculated, and the loss distribution maps of street trees in different years were drawn. The main conclusions are as follows:
(1) In recent years, the number of gale days and Gale Process in Shanghai tend to decrease obviously, and the gale is mainly spring gale and summer gale. The gale is mostly in the southeast coastal area and the Northeast riverside area, and less in the northwest inland area. The years of serious damage to street trees in Shanghai are 2012 and 2019. The disaster mainly occurs in August, and the time period of disaster is mainly concentrated in the daytime. The loss of streets trees is characterized by a large number in the eastern region and a relatively small number in the western region.
(2) The frequency of wind speed in each district of Shanghai decreases with the increase of wind speed intensity, and the fitting effect of power function as the relationship between wind speed frequency and wind speed intensity is the best. There are great differences in the risk of gale disaster factors in each region, showing the distribution law that the risk along the river and the coast is greater than that in the inland. On the whole, all districts in Shanghai are not prone to gale events with maximum wind speed exceeding 17m/s, but all districts have the possibility of extreme gale events with low probability and high risk.
(3) Using the data of gale intensity street tree loss rate and power function, the vulnerability curve model of street trees in urban and suburban areas of Shanghai is constructed respectively, and the determination coefficients of the vulnerability fitting equation of urban and suburban areas reach 0.930 and 0.934 respectively.
(4) The expected loss rate of street trees in all districts of Shanghai gradually increases with the increase of return period. Spatially, it is larger in coastal areas along the river than in inland areas, and larger in urban areas than in suburbs. The loss of street trees per unit area of different annual occurrence types generally presents the distribution characteristics that the eastern region is higher than the western region. In case of 10-year return period, the loss of street trees in each district is basically less than 3; In case of 50-year return period, the loss of street trees in each district is basically less than 15; In case of 100-year return period, the loss of street trees in each district is basically less than 70, of which Pudong New Area is the most, with a loss of 69; In case of 300-year return period, the loss of street trees in each district is basically less than 1500, and the maximum number in Pudong New Area is 1464. On the whole, the coastal areas along the river suffer great losses, among which the risk of loss of street trees in Pudong New Area is the largest, and the damage of street trees per unit area in Baoshan district is the largest. It is caused by the high risk of coastal areas along the river, the large amount of street trees and the high vulnerability of Pudong New Area and Baoshan District, but the weight of various influencing factors needs to be further studied.