空气中可吸入颗粒物浓度的增加与众多综合因素相关, 其空间分散程度与高程DEM间也有一定的相关性。 为了研究雾霾的污染与高度的空间相关关系, 以环首都地区100 km范围内为研究对象, 利用矩形格网尺度法对所研究区域进行不同边长及不同尺度的格网划分, 通过无人机获取可见光影像数据和高光谱POS信息数据, 对所研究区内的空气污染因子和高程因子进行提取和整合。 同时利用地统计学GS+软件的克里格插值法对所提取的变量数据进行空间相关性研究, 并利用MODIS遥感影像数据和无人机获取的POS数据与实地调查相结合的方法对地形和环境数据进行非线性回归拟合分析。 计算在不同格网尺度下环首都地区空气中的可吸入颗粒物及高程因子的空间相关效应的影响变程, 建立二者间的空间相关性优化模型, 从而确定可吸入颗粒物浓度随着高程变化的整体趋势。 结果表明: 高程DEM与空气污染指数API的最大相关影响距离为14.74 km, 且随着样本点间的距离增大, DEM的空间自相关性呈现逐渐减弱的规律, 即可吸入颗粒物浓度随着高程的增加而减小的整体趋势。 同时, 建立了高程DEM与环境间的空间相关性模型, 该模型符合地统计学的高斯球状模型, 相关系数r均高于90%, 模型拟合度较高。 试验为日后相关部门控制空气污染指数随着高度的变化选择不同树种进行绿化提供了一定的理论和实践指导依据。

Increased concentration of air respirable particulate matter associated with a number of combination factors. Spatial dispersion is also correlated with elevation DEM. In order to study the fog haze pollution associated with digital elevation model of spatial relations, this paper used the capital area ring within 100 km as the research scope, partitioning different length scale grid according to the rectangular grid method in the study area, obtaining visible light image data and hyperspectral image data by using unmanned aerial vehicle (UAV) for the extraction and integration air pollution factor and elevation factor within the scope of this study. GS+ software of kriging interpolation method was used to research the spatial correlation of variable data extraction; the MODIS remote sensing image data combined with field survey were used to analyze nonlinear regression of the terrain and environmental data.With the Calculation of variation effects of the particulate matter in the air and the spatial of the elevation factor under different grid scale ring of capital region, an optimization model of spatial correlation between them was established. Then the relation between the concentration of PM10 and height was determined. The biggest influence distance of elevation DEM associated with particulate matter API is 14.74 km. DEM space since the correlation of waning with the increase of the distance between sample points, which is also an important innovation of this paper. This result shows that the spatial correlation between the elevation DEM and environment conforms to the statistical spherical Gaussian model, correlation coefficient R2 were over 90%, which model fittings good. This study provides a certain theoretical and practical guidance for the control of air pollution index in the future as the change of height to select different tree species for afforestation.