摘要
【目的】目前我国完成了以县级行政区划为单元的森林火险区划,火险区划的结果在一定时期内相对不变,只适用于县级行政区划或以上区域开展森林火灾预防工作。以卫星遥感数据为基础,对山头地块的森林火灾风险进行基于栅格尺度的动态实时预警,为森林防火工作提供有效的预防信息。【方法】参考国内外各类研究文献从植被、气候、地形等角度构建森林火灾风险预警指标体系,利用2009—2019年森林火灾卫星热点数据、MOD02数据、DEM数据提取预警因子;以森林火灾动态危险性指数模型为基础,利用自然风险灾害指数模型、火险综合预报指数等模型对其进行改进得到森林火灾风险预警模型;绘制各预警因子的栅格图层,并进行叠置分析得到区域森林火灾风险预警图。【结果】以广东省为研究区。选取温度植被干旱指数(TVDI)、归一化植被指数(NDVI)、归一化红外指数(NDII7)、地形因子、历史森林火灾因子作为预警因子构建预警指标体系;构建的风险预警模型相对运行特征(ROC)值为1.0,大于0.7,拟合效果较好;利用2021年2月14日的数据根据模型计算得到的广东省2月15日的森林火灾风险预警图,可发现在2月15日广东省西部北部区域存在较少的高风险区,东部南部区域存在较多的高风险区。【结论】使用遥感数据可以进行森林火灾风险预警因子的提取工作;使用遥感数据可对区域森林火灾风险实现动态预警,可为森林防火工作提供科学依据。
【Objective】At present,China has completed the forest fire risk zoning with county-level administrative divisions as the unit.The results of the fire risk zoning are relatively unchanged within a certain period of time and only appropriate for the forest fire prevention work in county-level administrative divisions or above.Based on satellite remote sensing data,this paper provides real-time early warning of forest fire risk in hilltop plots to provide effective prevention information for forest fire prevention work.【Method】The forest fire risk early warning index system was constructed from vegetation,climate and terrain perspectives with reference to various domestic and foreign research literature,and the early warning factors were extracted using the forest fire satellite hot-spot data,MOD02 data and DEM data from 2009 to 2019;the forest fire risk early warning model was constructed based on the forest fire dynamic hazard index model,natural risk hazard index model and fire risk comprehensive forecast index,etc.The raster layers of each factor were drawn,and the regional forest fire risk early warning map was obtained by superposition analysis.【Result】Take Guangdong province as the study area.Temperature vegetation drought index(TVDI),Normalized vegetation index(NDVI),Normalized infrared index(NDII7),terrain factor and historical forest fire factor are selected as early warning factors to construct early warning index system;the Relative operating characteristic(ROC)value of the risk early warning model is 1.0,greater than 0.7,and the fitting effect is good;using the data of February 14,2021,according to the model calculation of February 15,Guangdong province according to the forest fire risk early warning map of February 15,it can be found that there are less high-risk areas in the northern part of western Guangdong province,and more high-risk areas in the southern part of eastern Guangdong province.【Conclusion】Remote sensing data can be used to extract early warning factors of forest fire risk;remote sensing data can be used to realize dynamic early warning of regional forest fire risk,which can provide scientific basis for forest fire prevention.
作者
李哲全
张贵
谭三清
吴鑫
LI Zhequan;ZHANG Gui;TAN Sanqing;WU Xin(College of Forestry,Central South University of Forestry&Technology,Changsha 410004,Hunan,China)
出处
《中南林业科技大学学报》
CAS
CSCD
北大核心
2021年第7期26-33,共8页
Journal of Central South University of Forestry & Technology
基金
湖南省科技创新平台与人才计划项目(2017TP1022)
湖南省应急管理科技项目(2020YJ007)
中南林业科技大学青年科学基金项目(2018YC006)。
关键词
卫星遥感
森林火灾
风险预警
森林火灾动态危险性指数
satellite remote sensing
forest fire
risk early warning
forest fire dynamic risk index
