中国安全科学学报 ›› 2023, Vol. 33 ›› Issue (12): 53-59.doi: 10.16265/j.cnki.issn1003-3033.2023.12.1736

• 安全工程技术 • 上一篇    下一篇

气候因子对我国森林火灾的影响及预测

鲁义1,2(), 周钦云1, 邵淑珍1, 王伟2, 戴玉清3, 韦雪娥4   

  1. 1 湖南科技大学 资源环境与安全工程学院,湖南 湘潭 411201
    2 应急管理部 上海消防研究所,上海 200032
    3 福建省森林消防总队,福建 福州 350003
    4 山东鲁泰控股集团有限公司 鹿洼煤矿,山东 济宁 272350
  • 收稿日期:2023-06-10 修回日期:2023-09-15 出版日期:2023-12-28
  • 作者简介:

    鲁义 (1986—),男,江西新干人,博士,教授,博士生导师,主要从事火灾科学与技术方面的研究。E-mail:

  • 基金资助:
    湖南省重点研发计划项目(2022GK2042); 湖湘青年英才项目(2020RC3047); 湖南省教育厅优秀青年项目(20B230); 湖南省科技创新人才计划大学生科技创新创业项目(2021RC1012)

Influence and prediction of climatic factors on forest fires in China

LU Yi1,2(), ZHOU Qinyun1, SHAO Shuzhen1, WANG Wei2, DAI Yuqing3, WEI Xue'e4   

  1. 1 School of Resource Environment & Safety Engineering, Hunan University of Science and Technology, Xiangtan Hunan 411201, China
    2 Shanghai Fire Research Institute of MEN, Shanghai 200032, China
    3 Fujian Provincial Forest Fire Brigade, Fuzhou Fujian 350003, China
    4 Luwa Coal Mine, Shandong Lutai Holding Grope Co., Ltd., Jining Shandong 272350, China
  • Received:2023-06-10 Revised:2023-09-15 Published:2023-12-28

摘要:

为探究气候因子对森林火灾事故的影响机制,首先基于我国2011—2020年森林火灾事故数据,选择年均气压、气温、降水量和日照时间等4类气候因子作为特征变量;其次采用层次聚类法得出森林火险区划结果,并计算灰色关联度系数,得出气候因子对森林火灾的影响程度;然后建立我国4类气候因子线性回归模型,得出气候因子对森林火灾影响程度的排序,并预测2021年的险情值。研究结果表明:我国火险区域在气候因子的影响下,可划为10类;气候因子对单位面积森林火灾发生次数影响排序为:年均降水量>年均气压>年均气温>年均日照时间;线性回归分析和灰色关联度系数分析的排序结果相似;线性回归模型预测的2021年险情值与实际值相对误差不超过5%。

关键词: 气候因子, 森林火灾, 火险区, 聚类分析, 关联度, 线性回归

Abstract:

In order to explore the influence mechanism of climatic factors on the weight of forest fire accident, firstly, based on the data of forest fire accidents in China from 2011 to 2020, four kinds of climatic factors, including annual average pressure, temperature, precipitation and sunshine time, were selected as characteristic variables. Secondly, the results of forest fire danger zoning were obtained by the hierarchical clustering method, and the grey correlation coefficient was calculated to obtain the influence degree of climate factors on forest fire. Then, the regression model of four types of climatic factors in China was established to obtain the weight ranking of the impact of climatic factors on forest fires, and the risk value in 2021 was predicted. The results show that the fire danger areas in China could be divided into 10 categories under the influence of climatic factors. The influence of climate on the number of forest fires per unit area is ranked as follows: annual average precipitation > annual average pressure > annual average temperature > annual average sunshine time. Linear regression analysis and grey correlation degree analysis have similarities in the ranking results of fire danger zone. The relative error between the predicted risk value by linear regression analysis and the actual value in 2021 is less than 5%.

Key words: climatic factors, forest fire, fire hazard area, hierarchical clustering, correlation, linear regression