High-Resolution Mapping of Ground-Level PM 2.5 in China Using FY-4A and FY-4B TOA Reflectance With Seamless Coverage

Fine particulate matter (PM2.5) is a key air pollutant that significantly impacts ambient air quality and public health. High-resolution, full-coverage PM2.5datasets are essential for effective air quality monitoring and management. In this study, we develop an innovative framework to estimate ground-level PM2.5concentrations across China with high spatiotemporal resolution by synergistically integrating top-of-atmosphere (TOA) reflectance data from FengYun-4A (FY-4A) and FengYun-4B (FY-4B) geostationary satellites. By exploiting the complementary orbital coverage and observation geometry of both satellites, we established a robust random forest model to derive hourly daytime PM2.5concentrations at 1 km resolution throughout 2023. This dual-satellite fusion strategy significantly enhances data availability by up to 63% while achieving higher estimation accuracy compared to single-satellite approaches. To address the remaining data gaps, we proposed an advanced spatiotemporal interpolation method (ST-IDW) with adaptive weighting schemes, generating a daily seamless dataset-dubbed FY-4 and 4B Geostationary PM2.5(FYGPM). Comprehensive validation demonstrates that the FYGPM dataset holds unparalleled advantages in data coverage while achieving accuracy comparable to the leading existing product. It successfully captures fine-scale pollution patterns, intra-urban variability, and urban-rural gradients across China, with spatial distributions showing strong alignment with impervious surface patterns and successfully identifying low PM2.5concentrations over water bodies. This study represents the first operational high-resolution PM2.5product for China derived from dual geostationary satellite observations. The FYGPM dataset serves as a valuable resource for air quality research, urban pollution analysis, and environmental policy-making at both national and local scales.