湖南省森林火险天气等级预测模型研究

摘要

To improve Hunan Province weather grade forecasting precision of forest fire risk and reduce the empty and false negative ratio, based on forest fire data from 2005 to 2015 in counties in Hunan province and corresponding to the meteorological data from county meteorological station, obtained the segmented regions by CV, set the regional scores of the highest temperature, wind speed, relative humidity, rainfall and other weather factors on of forest fire in Hunan Province. Got the influence weight using principal component analysis method in succession. And finally Established a weighted forest fire weather index model and determined weather grade for forest fire risk. By the forest fire data in 2015 verify the accuracy of the model, it reached 74.2%, indicating that the weather grade prediction model has practical value.%为提高湖南省森林火险天气等级预测精度，降低空报和漏报比例，基于湖南省各县2005—2015年森林火灾数据及各县气象站对应的同期气象资料，利用变异系数方法确定了日最高气温、风速、相对湿度、降雨量等天气因子的分段区域，设定了各区域的得分值，并运用主成分分析方法确定了不同分段区域对湖南省森林火灾的影响权重，建立了加权森林火险天气指数模型，根据天气指数确定森林火险天气等级。经2015年森林火灾数据验证模型精度达到了74.2%，表明了该森林火险天气等级预测模型具有适用价值。