基于随机森林和组合赋权法的武汉市洪涝灾害风险评估研究

Research on flood risk assessment of Wuhan city based on random forest and combined weighting

开展城市洪涝灾害风险评估工作是防灾减灾的基础和决策手段之一,已成为当前研究的热点和趋势。首先从城市洪涝灾害的概念和形成机制入手,将致灾因子危险性、孕灾环境敏感性、承灾体脆弱性和防灾减灾能力作为一级指标。以年平均降雨量、水系距离、高程、地形起伏度等14个二级指标为基础,建立风险评估指标体系。采用随机森林算法、层次分析法和熵值法相结合的方法对指标进行排序和赋权,得到各评估指标的组合权重。以武汉市为研究区域,将各指标权重输入到风险评估模型中,在ArcGIS中进行空间分析和栅格计算,得到研究区危险性、敏感性、脆弱性、防灾减灾能力风险区划图以及综合风险等级分布结果。

Therefore,conducting urban flood disaster risk assessment is one of the foundations and decision-making tools for disaster prevention and mitigation,and has become a current research hotspot and trend.This paper starts by exploring the concept and formation mechanism of urban flood disasters,taking the risk factors of hazard severity,disaster-prone environmental sensitivity,vulnerability of disaster-bearing bodies,and disaster prevention and mitigation capability as primary indicators.Based on this,a risk assessment index system is established,with fourteen secondary indicators including annual average rainfall,distance to water systems,elevation,and terrain relief as the foundation.Using ArcGIS to spatially process the data of each indicator,the importance of hazard and sensitivity indicators is ranked using the random forest algorithm.The method of combining analytic hierarchy process(AHP)and entropy method is employed to assign weights to the indicators,determining the combined weights of each evaluation indicator.Taking Wuhan city as the research area,the weights of each indicator are inputted into the established risk assessment model.Using ArcGIS’s spatial analysis techniques and raster calculation functions,the evaluation model is subjected to fuzzy comprehensive evaluation to obtain the risk zoning map of hazard,sensitivity,vulnerability,disaster prevention,and mitigation capability,as well as the distribution of comprehensive risk levels.The validation of actual disaster data verifies the scientific and rationality of the model results,providing important references for subsequent urban flood disaster prevention.