欢迎访问《中国农学通报》,

中国农学通报 ›› 2022, Vol. 38 ›› Issue (32): 119-127.doi: 10.11924/j.issn.1000-6850.casb2021-1089

所属专题: 农业气象

• 农业信息·科技教育 • 上一篇    下一篇

基于主成分分析的湖北省洪涝灾害风险评估模型构建

卢梦瑶1(), 刘德虎1, 鲁雪丽1, 梁衡1, 孙媛媛1, 刘亚林1, 宋廷强1(), 范海生2   

  1. 1青岛科技大学信息科学技术学院,山东青岛 266000
    2珠海市岭南大数据研究院,广东珠海 519000
  • 收稿日期:2021-11-12 修回日期:2022-01-21 出版日期:2022-11-15 发布日期:2022-11-09
  • 通讯作者: 宋廷强
  • 作者简介:卢梦瑶,女,1998年出生,山东济南人,硕士在读,研究方向:洪涝灾害评估、卫星影像处理。通信地址:266100 山东省青岛市崂山区中韩街道松岭路99号 青岛科技大学信息科学技术学院,E-mail: lmy_qust@163.com
  • 基金资助:
    山东省重点研发计划“星陆双基协同反演的洪水遥感监测预警系统关键技术研究”(2019GGX101047)

Construction of Flood Disaster Risk Assessment Model Based on Principal Component Analysis in Hubei Province

LU Mengyao1(), LIU Dehu1, LU Xueli1, LIANG Heng1, SUN Yuanyuan1, LIU Yalin1, SONG Tingqiang1(), FAN Haisheng2   

  1. 1Qingdao University of Science and Technology, Qingdao, Shandong 266000
    2Zhuhai Lingnan Big Data Institute, Zhuhai, Guangdong 519000
  • Received:2021-11-12 Revised:2022-01-21 Online:2022-11-15 Published:2022-11-09
  • Contact: SONG Tingqiang

摘要:

湖北省历年由洪涝灾害造成农作物受损严重,对湖北省进行洪涝风险评估十分必要。本文提出了一种定量化风险评估的模型建立方法,通过多源数据(气象、社会经济、地理特征等数据)提取到15个指标,采取主成分分析法确定各因子对于洪涝灾害的影响权重,建立风险评估模型,并运用地理信息系统(GIS)分析技术得出洪涝灾害风险区划图。在现有评估指标体系的基础上,通过网络爬虫方式获取更能反映防减灾能力的灾害应急指标;采用主成分分析方法降低模型建立中的主观因素。结果表明:(1)通过模型得到降雨与地势为湖北省洪灾发生的最主要因素;(2)湖北省中东部地区多为高风险区,其中东部武汉、黄石等长江干流途经地区处于重风险区;西南部多为中风险区,西北部在全省为低风险区。综上,该模型可为湖北省开展综合减灾、调整区域可持续发展结构、进行准确农情监测提供科学支撑和决策依据,具有重要的科学和实践意义。

关键词: 农业风险评估, 洪涝灾害, 爬虫, 主成分分析, GIS分析

Abstract:

Over the years, floods have caused serious damage to crops in Hubei Province, restricting agricultural economy and threatening social development. It is necessary to carry out flood risk assessment in Hubei. This paper proposed a method of modeling quantitative risk assessment in Hubei Province. Fifteen indicators were extracted from multi-source data (meteorological, socio-economic, geographical characteristics and other data), and the principal component analysis method was adopted to determine the weight of each indicator on flood disaster to establish a risk assessment model, and the geographic information system (GIS) analysis technology was used to get the flood disaster risk zoning map. On the basis of the existing evaluation index system, through the way of web crawler, we obtained better disaster emergency indicators to reflect the ability of prevention and reduction, and used the principal component analysis method to reduce the subjective factors in model building. The results show that: (1) rainfall and topography are the most important factors of flood occurrence in Hubei; (2) most of the central and eastern parts of Hubei are high-risk areas, among which, Wuhan, Huangshi and other parts along the Yangtze River basin in the east are in a heavy risk area; the southwest of the province is mostly in a medium risk area, and the northwest is in a low risk area. In conclusion, this model could provide scientific support and a decision-making basis for carrying out comprehensive disaster reduction, adjusting regional sustainable development structure and monitoring agricultural production in Hubei Province, which has great scientific and practical significance.

Key words: agricultural risk assessment, flood disaster, web crawler, principal component analysis, GIS analysis

中图分类号: