以大兴安岭地区2000—2010年卫星火点数据为基础,运用逻辑斯蒂回归模型建立基于气象、植被、地形、社会经济和基础设施因子的林火预测模型,并划分火险等级。此外,基于火险区划对大兴安岭地区的林火监控与扑救效率进行评价分析。模型预测结果表明,坡度、月平均降水、月平均温度等多种因素对大兴安岭地区林火发生有显著影响;火险区划表明,高火险区域面积占研究区域总面积的24.47%;林火监控效果的评价结果显示,对比现有瞭望塔选址方式,网格布设法(20km×20km网格和10km×10km网格)的监控布点可以显著提高林火监测效果,高火险区域监测面积分别提高36.97%和60.52%,同时,站点需求数量也相应减少;林火扑救可达性研究表明,大兴安岭地区东、南、西部地区对高火险区域的可达性较好,中、北部地区则相对较差,而全区仅有1/3的森林消防站处于可达性高的区域,因此建议在可达性差的高火险区域适当增设森林消防站,从而全面提高该区域林火扑救效率。

A logistic regression model was used to predicte Daxing'an Mountain's forest fire based on the climate,vegetation,terrain,socio-economic and infrastructure factors combing satellite fire point dataset from 2000 to 2010,the fire risk division was generated accordingly.In addition,an analysis of efficiency of fire ignition monitoring and suppression was also conducted based on fire risk division.The results showed that thirteen factors including slope,monthly average precipitation,monthly average temperature etc.,are significantly correlated with forest fires.High fire risk zone accounts for 24.47% of Daxing'an Mountains.Compared to current allocation of fire monitoring,the gridding methods on location selection of fire tower (20 km×20 km grid and 10 km×10 km grid) can significantly improve the efficiency of fire monitoring.The visible rates on the high fire risk zone improved 36.97% and 60.52% respectively and the required number of fire towers reduced meanwhile.On the other hand,the accessibilities of fire suppression to high fire risk zone are higher in East,South and West of Daxing'an Mountains than the central and Northern areas, however,only one third forest fire station in the region can send fire fighters to high fire risk areas in a short time.We therefore suggest that more fire stations need to be added in the poor accessibility area in order to improve the efficiency of forest fire suppression of Daxing'an Mountains.