首页|社交媒体数据支持的城市承灾体脆弱性评估——以深圳极端天气为例

社交媒体数据支持的城市承灾体脆弱性评估——以深圳极端天气为例

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
世界范围内极端天气频度和强度的显著增强,对高密度城市中公众和基础设施等承灾体造成严峻威胁,揭示和评估城市承灾体脆弱性对提升城市韧性具有重大意义.相对传统数据在极端突发灾害事件中的应用局限,社交媒体数据包含了实时且丰富时空信息而呈现出巨大潜力.研究以2018年"山竹"台风灾害3个时间序列阶段和深圳76个街道的空间分布为例,获取并预处理灾害期间相关微博数据后,分别基于LDA模型和SnowNLP模型训练得到公众灾害感知脆弱性和公众情绪脆弱性,并基于语料库构建进一步细分得到6种类型的基础设施脆弱性.研究发现这3种脆弱性的集中峰值呈现依次滞后的时间序列特征以及南北圈层差异和"哑铃状"的空间分布特征.最后针对性提出提升城市承灾体脆弱性时空协调性的相关建议,对城市灾害韧性提供一定规划指引.
VULNERABILITY ASSESSMENT OF URBAN DISASTER-AFFECTED BODIES BASED ON SOCIAL MEDIA DATA:TAKING EXTREME WEATHER IN SHENZHEN AS AN EXAMPLE
The frequency and intensity of extreme weather events have significantly increased worldwide,posing severe threats to the public and infrastructure in high-density cities.Therefore,it is crucial to assess the vulnerability of urban disaster-affected bodies,as it plays an essential role in enhancing urban resilience.Different from the traditional data,which is often limited in evaluating sudden-onset disasters,social media data offers real-time and rich spatio-temporal information,making it a valuable source.This paper focuses on the spatial distribution of 76 sub-districts in Shenzhen during the three time periods of the 2018 typhoon Mangkhut.After preprocessing the relevant Weibo data during the disaster period,the paper identifies and classifies public perception vulnerability and public sentiment vulnerability using the LDA model and the SnowNLP model.Furthermore,the infrastructure vulnerability is subdivided into six categories based on corpus construction.The analysis reveals that three types of vulnerabilities exhibit concentrated peaks with successive lags,demonstrating distinct spatial patterns between the northern and southern regions and a"dumbbell-shaped"distribution.Based on these findings,the paper proposes relevant planning suggestions to improve the spatio-temporal coordination of the vulnerability of urban disaster-affected bodies,in order to enhance urban disaster resilience.

vulnerability of disaster-affected bodiessocial mediapublic perceptionpublic sentimentinfrastructure vulnerability to disaster

范雅婷、李越凡、涂伟、李云

展开 >

深圳大学建筑与城市规划学院

亚热带建筑与城市科学全国重点实验室

深圳大学建筑与城市规划学院城市空间信息工程系

深圳大学建筑与城市规划学院城市规划系

深圳市建筑环境优化设计研究重点实验室

中国城市规划学会青年工作委员会

展开 >

承灾体脆弱性 社交媒体 公众感知 公众情绪 基础设施承灾脆弱性

国家自然科学基金委员会青年项目广东省自然科学基金项目面上项目深圳市孔雀计划科研启动基金

523080262023A1515011274827/000297

2024

10.11819/cpr20231627a

城市规划
中国城市规划学会

城市规划

CSTPCDCSSCICHSSCD北大核心
影响因子：2.515
ISSN：1002-1329
年,卷(期)：2024.48(8)
  • 19