The study researched the relationship between vegetation cover and PM2.5 pollution. The raster NDVI dataset from 1998 to 2016 were reclassified into low, medium, and high vegetation coverage area, and the corresponding PM2.5 concentration in eight economic regions in China were then calculated. On this basis, the temporal and spatial characteristics of PM2.5 pollution were analyzed and Pearson correlation coefficient was used to explore its correlation with NDVI landscape pattern indexes separately from landscape and class level NDVI. The preliminary results showed that:①The northern, eastern, southern coastal, middle reaches of the Yangtze River, and the northeast economic zones have relatively low vegetation coverage in areas with relatively serious PM2.5 pollution. However, the middle reaches of the Yellow River, the southwestern and the Northwestern Economic Zones in areas with relatively low vegetation coverage showed lighter PM2.5 pollution. ②PM2.5 increased in most areas between 1998 and 2016. ③A significant correlation between PM2.5 and NDVI landscape pattern indexes was not found for all areas. ④Therefore, the impacts of the landscape shape index(LSI), percent of landscape(PLAND), number of patches(NP), largest patch index(LPI), and aggregation index(AI) on PM2.5 are heterogeneous.
Keywords: PM2.5; landscape pattern indexes; normalized difference vegetation index(NDVI); panel data model; vegetation coverage.