基于泰森多边形的广州市林火空间分布规律研究

Spatial Distribution of Forest Fire Based on the Voronoi in Guangzhou

将泰森多边形特点引入林火空间分布规律研究,依据泰森多边形面积越大,林火发生次数越少,则林火发生概率越低,通过克里格插值生成广州市林火易发性分布图,在对道路距离、海拔和坡度等关键影响因素进行空间异质性分析的基础上,采用地理加权回归模型进行局部参数估计,并定量分析各影响因素回归系数在空间上的作用,以直观地反映林火各影响因素空间分布。结果表明,林火易发性变异函数揭示林火的发生主要由地理自然因素和人为随机因素共同作用;林火各影响因素空间异质性特征明显,局部存在较强的聚集格局;GWR模型分析结果明显优于OLS模型,各变量回归系数在空间上作用不均匀,存在显著的空间差异;广州市东北部林火易发性明显高于其他地区,海拔因素呈从北到南正负效应相反的分布,坡度因素呈块状分布,道路因素呈带状、块状空间分布。林业管理部门根据林火发生的规律和特点,可在林火防控工作上更具针对性,最大程度地减少林火损失。

Voronoi diagram was introduced to the analysis of the distribution of forest fire.Based on the fact that the larger the voronoi area,the lower the frequency of forest fire,and the lower occurrence rate of the forest fire,Kriging interpolation was used to generate distribution diagram of forest fire.On the basis of spatial analysis on key factors of road,elevation and slope,and GWR model was used to estimate the local parameters to quantitatively analyze the regression in the spatial role,and intuitively reflect the forest fire affect on the spatial distribution.The results showed that the variation of forest fire indicated that the occurrence of forest fire was mainly caused by the combination of natural factors and man-made random factors.The spatial heterogeneity of forest fire had strong clustering pattern.The results of GWR model analysis were superior to OLS model,and the regression coefficients of each variable were not uniform in space and significant spatial differences existed.The rate of forest fire in the northeastern part of Guangzhou was significantly higher than that in other areas.The elevation factors were distributed from north to south with positive and negative effects,the slope factors were block,and the road factors were banding and block.According to the law and characteristics of forest fire above mentioned,forestry departments in Guangzhou could adopt more targeted measures to minimize the loss of forest fire.