First satellite-based regional hourly NO2 estimations using a space-time ensemble learning model: A case study for Beijing-Tianjin-Hebei Region, China

Surface Nitrogen dioxide (NO2) concentrations have been generated with satellite retrievals using multiple statistical algorithms. However, they are often given at coarse frequencies ("snapshot", daily or even longer), limiting their applications in epidemiological studies and assessing the evolution of NO2 pollution. This study investigated the potential applicability of Himawari-8 derived hourly fine particulate matter concentrations in producing hourly NO2 concentrations by constructing a space-time ensemble model. The Beijing-Tianjin-Hebei (BTH) region, one of the serious pollution regions in China, is the study region chosen. The proposed model performs well in estimating hourly NO2 concentration with a high cross-validation (CV) coefficient of determination (R-2 = 0.81) and low CV root-mean square (RMSE = 9.71 mu g/m(3)), mean prediction errors (MPE = 6.33 mu g/m(3)), and relative prediction errors (RPE = 22.5%). On daily, monthly, seasonal, and annual time scales, CV R-2 increases to 0.89, 0.93, 0.97, and 0.99, respectively. The annual mean model estimated NO2 concentration over BTH region is 28.2 +/- 6.5 mu g/m(3), with relatively higher NO2 concentrations are seen in southern and southeastern BTH. Winter experiences the most severe NO2 concentrations, followed by autumn, spring, and summer. Surface NO2 concentrations are higher (lower) in the morning (afternoon) and tend to decrease gradually with time. The model generally captures the hourly evolution of NO2 concentrations for the severe pollution episode but shows some underestimations. The annual mean NO2 concentrations were 2.8% lower on the weekend than on weekdays. In addition, the weekend effects of NO2 concentrations are larger at rush hour and lower in the noon. The hourly NO2 products derived from proposed approach are potentially useful for improving our understanding of the source, evolution, and transportation behavior of NO2 pollution episodes and for exposure-and health-related research. The proposed approach also enriches the potential applications of geostationary satellites (e.g., Himawari-8).