摘要
应用1974年—2009年间人为火的空间地理坐标,结合研究地的基础地理信息及矢量化林相图,应用ArcGIS10.0中的空间分析工具和SPSS19.0的逻辑斯蒂回归模型深入分析地形环境因素对人为火发生的影响。研究结果表明"到铁路距离"对人为火的影响显著,其次,在1∶1和1∶3选取随机点条件下人为火发生分别与"优势树种"和"林型"显著相关。本文研究显示在运用"随机点"方法进行相关性分析时,随机点选取数量对模型拟合结果有一定影响。
In this paper, we used the spatial analysis tools in ArcGIS10.0 and Logistic regression model imbedded in SPSS 19. 0 to identify the impact of human-activity and geographical factors on forest fires, based on the coordinates of human-caused fires during 1974-2009 and the basic geographical map. The results showed that ＂the distance to railway＂ has a significant impact on the human-caused fires. Moreover, ＂the dominant tree＂ and ＂the forest type＂ are significantly correlated with human-caused fires under 1 ： 1 and 1 ： 3 conditions, respectively. It was also found that the rate of random sample selection affects the parameter estimation of the model.
出处
《火灾科学》
CAS
CSCD
北大核心
2015年第1期16-25,共10页
Fire Safety Science
基金
福建农林大学重点大学建设基金(6112C035)
国家自然科学基金(31400552)
关键词
北方针叶林
塔河地区
人为火
地形环境
逻辑斯蒂回归
Boreal forest
Tahe area
Human-caused fire
Geographical factors
Logistic regression
