摘要
利用EOS/MODIS数据和内蒙古不同类型草地枯草期野外实测可燃物月动态数据,建立枯草期遥感估测模型。在此基础上,选择积雪覆盖、可燃物重量、草地连续度、日降水量、日最小相对湿度、日最高气温、日最大风速等7个指标,采用层次分析法确定各指标的权重,构建草原火险指数,建立草原火险等级短期预报模型。应用模型可将内蒙古草原的火险状态划分为不燃、难燃、可燃、易燃、极易燃5级,并对火险做出短期预报。利用数据抽样回代检验方法对草原火险等级短期预报进行精度检验,检验结果为落区预报准确率达96.42%,预报结果与实际基本相符,可以用于内蒙古草原火险短期预报的实际应用。
Remote sensing estimation model on the amount of combustible-substance over dry period based on EOS / MODIS data was constructed using monthly dynamic data of field survey combustible-substance of different types of grassland in Inner Mongolia during dry period. On this basis, snow cover, the weight of combustible-substance, the grassland continuity degree, the daily precipitation, the daily minimum relative humidity, the daily maximum temperature, and the daily maximum wind speed were select- ed. Grassland fire danger index was constructed adopting analytic hierarchy process to determine the weight of each index. Short-time forecast model of the grassland fire danger rating was build up and the fire danger status of the grassland of Inner Mongolia was divided into five levels using above model, the five levels namely non-flammable, difficult-flammable, flammable, inflammable, extremely inflammable, and the fire danger level has been forecasted simultaneously. The accuracy of the short-time forecast of grassland fire danger rating was tested with the test method of data sampling back substitution, the results showed that the accuracy of location prediction reached 96.42% and the forecast and practical were basically matched, so the model can be applied to the short-time forecast of the grassland fire danger rating of Inner Mongolia in practice.
出处
《中国草地学报》
CSCD
北大核心
2012年第4期87-92,共6页
Chinese Journal of Grassland
基金
农业部公益性行业专项项目"草原火灾应急管理技术"(200903041)资助
关键词
内蒙古草原
可燃物
层次分析法
草原火险指数
短期预报
Inner Mongolia grassland
Fuel
Analytic hierarchy process
Grassland fire danger index
Short-time forecast
