中文摘要：

草原火灾是一种突发性强、破坏性大、处置救助较为困难的自然灾害,对草原资源危害极为严重。以呼伦贝尔草原牧业6个旗为研究区,利用1994-2005年呼伦贝尔草原火灾统计月报表和相关气象资料,依据自然灾害风险形成理论、区域灾害系统理论、灾害预警理论建立草原火灾风险预警模型。利用Logistic回归模型进行警源识别确定影响草原火灾风险预警的关键因子,采用层次分析法通过专家打分计算各个指标的权重,用网格GIS技术结合回归分析对选区的指标进行空间展布使得空间评价尺度更加精确,利用加权综合评分法分析了呼伦贝尔草原火灾的内生警源和外生警源的警兆,用最优分割法对1994-2004年火灾案例进行最优分割确定预警阈值并划分为蓝色、黄色、橙色、红色警报。以2005年全年呼伦贝尔草原火灾案例为例对呼伦贝尔草原火灾进行风险预警。经过检验发现火灾风险高预警区与火点吻合较好,从一定程度上检验了该模型的准确性。

英文摘要：

Grassland fire disaster is one of a sudden,destructive,and hard to disposal grassland disasters,it caused a great threat to the pastoral people＇s lives and property in Hulunbeier grassland fire statistics on statements and related meteorological data from 1994 to 2005 and theories of risk of natural disasters,and of regional disaster system,an astoral.In this research,early warning of grassland fire disaster risk was modeled through the study of six livestock counties of Hulunbeier grassland,based on data of disaster early warning.Logistic regression model was used in the identification of the key factors influence the early warning of grassland fire disaster risk;in the calculation the weight of individual indicators,analytic hierarchy through scoring by experts was used;grid GIS technology combined with regression analysis was used in the spatial distribution of indicators of the constituency which will make space rating scale more accurate;warning source of endogenous and exogenous sources of warning signs of Hulunbeier grassland fires was analysis through weighted comprehensive analysis;with the optimal partitioning of the fire cases from 1994 to 2004,early warning threshold was identified and was divided into the blue,yellow,orange and red alert.Grassland fire risk warning has been calculated on Hulunbeier,take all the fires on Hulunbeier grassland in 2005 for instance.It shows a good agreement between high risk fire areas of early warning and fire point,which,in terms,tested the accuracy of the model in a certain degree.