摘要
云南是我国森林火灾的多发区和重灾区,火灾风险预警需求紧迫。综合运用影响森林火灾发生发展的基本原理,利用MODIS数据、DEM数据及VIIRS热异常数据集提取火险因子,构建森林火灾动态危险性指数模型,并经过ROC检验对模型预报结果的精度分析。结果表明:1)模型ROC曲线最佳划分阈值0.527、AUC值0.811,预测精度较好。2)森林火灾频次与NDVI呈正相关关系,NDII7、TVDI值在0.6~0.8时对森林火灾影响最显著;森林火灾频次与3 000 m以上高程呈负相关关系,与坡度呈负相关关系,坡向对于森林火灾的发生没有显著的影响。3)利用模型预报结果通过自然断点法将火险等级分为5级,得出云南省高火险区主要分布在昆明市、迪庆州、丽江市、大理州、楚雄州、曲靖市、昭通市、玉溪市、红河州、文山州等北部、中部及东南部地区。
Yunnan is the frequently occurring and severely affected area of forest fires in China,and the demand for early warning of fire risk is urgent.Based on the basic principles affecting the occurrence and development of forest fire,the fire risk factors were extracted from MODIS data,DEM data and VIIRS thermal anomaly data set,and the dynamic risk index model of forest fire was constructed.The accuracy of the prediction results of the model was analyzed by ROC test.The results show that:1)The optimal threshold of ROC curve is 0.527,AUC value is 0.811,and the prediction accuracy is good.2)There was a positive correlation between forest fire frequency and NDVI,and ndii7 and TVDI had the most significant impact on forest fire when the value was 0.6~0.8;The frequency of forest fire is negatively correlated with the elevation above 3000 m,and negatively correlated with the slope.Slope aspect has no significant effect on the occurrence of forest fire.3)Based on the prediction results of the model,the fire risk level is divided into five levels by using the natural breakpoint method.It is concluded that the high fire risk areas in Yunnan Province are mainly distributed in Kunming,Diqing,Lijiang,Dali,Chuxiong,Qujing,Zhaotong,Yuxi,Honghe,Wenshan,which are in northern,central and southeastern regions in Yunnan Province.
作者
蔡珊珊
赵璠
邓小凡
邓航
赵俊帆
CAI Shanshan;ZHAO Fan;DNEG Xiaofan;DENG Hang;ZHAO Junfan(College of Big Data and Intelligent Engineering,Southwest Forestry University,Kunming,Yunnan 650224,China)
出处
《贵州师范大学学报(自然科学版)》
北大核心
2025年第1期69-78,共10页
Journal of Guizhou Normal University:Natural Sciences
基金
云南省基础研究计划项目(202301AT070223)
国家自然科学基金项目(32160374)
云南省中青年学术和技术带头人后备人才项目(202405AC350034)。
