Aiming at improving the air quality and protecting public health, policies such as restricting factories, motor vehicles, and fireworks have been widely implemented. However, fine-grained spatiotemporal analysis of these policies’ effectiveness is lacking. This paper collected the hourly meteorological and PM2.5 data for three typical emission scenarios in Hubei, Beijing–Tianjin–Hebei (BTH), and Yangtze River Delta (YRD). Then, this study simulated the PM2.5 concentration under the same meteorological conditions and different emission scenarios based on a reliable hourly spatiotemporal random forest model ( R 2 exceeded 0.84). Finally, we investigated the fine-grained spatiotemporal impact of restricting factories, vehicles, and fireworks on PM2.5 concentrations from the perspective of hours, days, regions, and land uses, excluding meteorological interference. On average, restricting factories and vehicles reduced the PM2.5 concentration at 02:00, 08:00, 14:00, and 20:00 by 18.57, 16.22, 25.00, and 19.07 μ g / m 3 , respectively. Spatially, it had the highest and quickest impact on Hubei, with a 27.05 μ g / m 3 decrease of PM2.5 concentration and 17 day lag to begin to show significant decline. This was followed by YRD, which experienced a 23.52 μ g / m 3 decrease on average and a 23 day lag. BTH was the least susceptible; the PM2.5 concentration decreased by only 8.2 μ g / m 3 . In addition, influenced by intensive human activities, the cultivated, urban, and rural lands experienced a larger decrease in PM2.5 concentration. These empirical results revealed that restricting factories, vehicles, and fireworks is effective in alleviating air pollution and the effect showed significant spatiotemporal heterogeneity. The policymakers should further investigate influential factors of hourly PM2.5 concentrations, combining with local geographical and social environment, and implement more effective and targeted policies to improve local air quality, especially for BTH and the air quality at morning and night.
Super Resolution Deduction: Inferring Fine-Grained Capacity for Urban Signal Station Deployment
