Radiative Blindness in Satellite UV Products Over Urban Industrial Canopies: A Case Study in Sao Jose dos Campos, Brazil

Satellite-derived ultraviolet (UV) products are the primary source of solar radiation data for public health and environmental monitoring in the Global South. However, their coarse spatial resolution often fails to capture attenuation caused by localized urban aerosol layers, leading to significant exposure estimation errors. This study investigates the decoupling between ground-level air quality and NASA POWER UV products in São José dos Campos, Brazil (January–June 2025). By integrating ground-based PM10 and O3 observations (CETESB) with satellite estimates, we quantify radiative bias under varying cloud regimes and anthropogenic cycles. While photochemically driven variables (O3) exhibit robust correlation with satellite UV (r = 0.56), particulate matter (PM10) shows a statistically negligible association (r = 0.067), even after controlling for cloudiness. Crucially, we identify a distinct “Weekend Effect,” in which ground-based PM10 concentrations drop by 18% on Sundays, yet the satellite-derived UV product shows no statistically significant response (p > 0.5). Theoretical approximations suggest that this “satellite blindness” results in systematic overestimation of surface UV irradiance by up to 12% during clear-sky pollution events. These findings reveal a critical bias in remote sensing products over urban microclimates and underscore the need for ground-based IoT radiometer networks to correct satellite models in real time.