基于模糊聚类因子分析的区域洪灾风险评价

Regional flood risk assessment based on fuzzy cluster factor analysis

在从洪灾风险形成机理出发建立区域洪灾风险评价指标体系的基础上,综合模糊C-均值聚类、因子分析和聚类有效性函数,建立基于模糊聚类因子分析的区域洪灾风险评价方法,并将其应用于怀洪南片防洪保护区15个乡镇的洪灾风险评价中。结果表明,怀洪南片防洪保护区的洪灾风险与社会经济因子、土地覆盖物因子和洪水因子等3个主要因子关系密切,被评价的15个乡镇中,小蚌埠镇为特高风险区,怀远县城关镇为高风险区,吴小街镇为较高风险区,新集镇、头铺镇、大新镇、曹顾张乡、梅桥镇、曹老集镇和沫河口镇为中等风险,临北回族乡为较低风险区,新马桥镇、王庄镇、浍南镇和魏庄镇为低风险区。研究结果揭示了影响怀洪南片防洪保护区洪灾风险的主要因子及防洪保护区洪灾风险的空间分布,可为区域防洪减灾活动提供指导。

Based on the formation mechanism of flood risk, the regional flood risk assessment index system is established, on this basis a regional flood risk assessment model is put forward which couples fuzzy C-means clustering, factor analysis and clustering validity function. As an example apply the proposed model to evaluate flood risk degree of 15 towns in the Huaihongnanpian flood protection area （simplified as the HHNP）. The research resuhs show that the flood risk of HHNP is closely related to three main factors, they are socioeconomic factor, land cover factor and flood factor respectively. The risk degree is divided into six zones objectively, namely especially high, high, relatively high, medium, relatively low and low. In the 15 towns, Xiaobengbu, Chengguan and Wuxiaojie fall into especially high, high, relatively high zone, respectively. Xinji, Toupu, Daxin, Caoguzhang, Meiqiao, Caolaoji and Mohekou fall into median risk zone. Linbeihuizu belongs to relatively low zone and Xinmaqiao, Wangzhuang, Kuainan and Weizhuang are in low zone. The research results revealed the main driving factors and the spatial distribution of flood risk in HHNP, it is highly significant for us to understand the main risk sources and pro- vide guidance for flood control and management in the region.