Spatio-temporal correlation between energy consumption and PM_2.5 concentration based on nighttime light images

PM_2.5 is the primary pollutant in urban air in China, causing serious harm to human physical and mental health, arousing widespread concern．Study on the spatial and temporal relationship between PM_2.5 and energy consumption will provide some theoretical basis to formulate effective atmospheric environmental protection policies and to promote urbanization．Energy consumption statistical data and nighttime light images were used to define spatial patterns in energy consumption in Shaanxi Province. Time-space relationship between energy consumption and PM_2.5 concentration with PM_2.5 remote sensing data were studied by spatial correlation analysis．Random forest regression was used to dissect energy consumption factors affecting changes in PM_2.5 concentration．It was found that from 2001 to 2013, PM_2.5 concentrations in Shaanxi Province initially increased and then declined, with the highest value at 28.5 μg·m⁻³．The spatial heterogeneity in PM_2.5 distribution in the province was marked, with the Guanzhong region showing the highest PM_2.5 concentration．Energy consumption in Shaanxi Province was found to increase year by year, with a spatial distribution similar to that of PM_2.5 concentration．Energy consumption in the Guanzhong region was the largest．The Moran’s index of energy consumption and PM_2.5 concentration in Shaanxi Province reached 0.289, indicating an obvious positive spatial correlation-areas with high energy consumption had high concentrations of PM_2.5．Population density, road network density and total energy consumption were found to be important driving factors for changes in PM_2.5 concentration in Shaanxi Province．