【摘要】 为了解贵阳市冬季大气污染现状,以贵阳市污染相对严重的白云区为研究对象,连续采集PM_2.5、PM₁₀浓度数据,利用普通克里金法进行空间插值获取PM_2.5、PM₁₀分布特征。通过留一法交叉验证,比较6种半变异函数模型（三角函数、高斯函数、球面函数、指数函数、J-Bessel函数和K-Bessel函数）的空间插值精度,选出最适的函数模型;采用分区统计和格网统计的方法,对不同土地利用类型、植被覆盖度下的PM_2.5、PM₁₀平均浓度进行比较分析。结果表明,三角函数是PM_2.5空间插值的最适模型,指数函数是PM₁₀空间插值的最适模型;贵阳市白云区冬季大气PM_2.5、PM₁₀浓度总体表现出城区浓度高,郊区浓度低的分布特征;土地利用类型和植被覆盖度对PM_2.5和PM₁₀浓度有着较强的影响。更多还原

【Abstract】 In order to understand the current atmospheric pollution situation of Guiyang during winter,the Baiyun District was selected as the research area to collect PM_2.5and PM₁₀concentration data continuously,and ordinary Kriging method was used for spatial interpolation to analyze characteristics of PM_2.5and PM₁₀distribution.Through the"leave-one-out"cross validation method,the spatial interpolation accuracy of different semi-variable function,such as the trigonometric function,Gaussian function,spherical functions,exponential functions,J-Bessel function and K-Bessel function was compared,and the most suitable function model was selected.The average PM_2.5,PM₁₀concentration of different land use type and vegetation coverage was analyzed respectively by means of zoning statistics and grid statistics.The results showed that the triangular function model was the most suitable for the spatial interpolation of PM_2.5and the exponential function model was the most suitable for the spatial interpolation of PM₁₀.In the Baiyun District,the concentration of PM_2.5and PM₁₀presented the distribution characteristics as higher in urban areas and lower in suburban areas.Land use type and vegetation coverage had a strong impact on PM_2.5and PM₁₀concentrations.更多还原