北京地区霾/颗粒物污染与土地利用/覆盖的时空关联分析

土地利用/覆盖变化(LUCC)直接或间接影响颗粒物污染,为定量评价颗粒物污染与土地利用在时序变化和空间分布上的关系,使用北京市长期气象资料、土地利用数据、实测颗粒物浓度数据和遥感影像,结合GIS空间分析和典型相关分析方法,对北京市霾日数与土地利用变化的响应关系、以及局地土地利用类型对颗粒物污染浓度空间分布的影响进行了定量评价.结果表明:从时序变化的角度看,北京市(行政区内)1996~2008年间霾日数与生态用地和耕地表现为显著负相关,相关系数分别为-0.574和-0.592,与建设用地、居民点及工矿用地和交通用地则呈显著正相关,相关系数分别为0.595、0.609和0.590;北京城区(五环内)1989~2012年间霾日数与生态用地、耕地和建设用地变化趋势具有良好的响应关系,其中城市区域扩张对霾日数影响显著,建成区面积与霾日数显著正相关,相关系数为0.876.从空间分布的角度看,建筑工地、道路和裸地3种下垫面类型的颗粒物浓度要明显高于其他类型;在0.5km和1km缓冲区内,绿地面积与PM1.0浓度呈显著负相关,相关系数分别为-0.542和-0.507,建筑面积与PM1.0浓度呈显著正相关,相关系数分别为0.469和0.537.总体来看,北京地区颗粒物污染水平、空间分布格局与土地利用/覆盖状况具有良好的时空关联性.

Response analysis of haze/particulate matter pollution to Land Use/Cover in Beijing

Land-Use/Cover Change (LUCC) has direct or indirect effects on particulate matter pollution. To quantitatively evaluate the correlation between particulate matter pollution and land use in terms of temporal variation and spatial distribution, long-term meteorological data, land use data, measured particulate concentration data and remote sensing images, as well as the GIS spatial analysis methods and canonical correlation analysis methods were used to perform quantitative evaluation of the correlation between the number of haze days and land use change in Beijing, as well as the effects of local land use types on spatial distribution of particle pollution. The results showed that: (1) In respect of temporal variation, from 1996 to 2008, the number of haze days in Beijing presented a significant negative correlation with ecological land and agricultural land, with the correlation coefficients being -0.574 and -0.592 respectively, but presented a significant positive correlation with construction land, residential settlements and industry and mining, and transportation land, with the correlation coefficients being 0.595, 0.609, and 0.590 respectively; from 1989 to 2012, the number of haze days in urban areas of Beijing presented good responding relationship with the variation trend of ecological land, agricultural land, and construction land, in which the expansion of urban areas has significant effect on the number of haze days, and the coverage of built-up areas presented a significant positive correlation with the number of haze days, with the correlation coefficient being 0.876. (2) In terms of spatial distribution, the concentration of particulates at underlying surface including construction sites, roads, and bare land was obviously higher than that at land of other types; within the 0.5km and 1km buffer area, the area of green land presented a significant negative correlation with the PM1.0concentration, with the correlation coefficients being -0.542 and -0.507 respectively, and the area of construction land presented a significant positive correlation with the PM1.0 concentration, with the correlation coefficients being 0.469 and 0.537 respectively. As a whole, the pollution level and spatial distribution pattern of particulate matter were closely related to the Land Use/Cover status.