Abstract The Tropospheric Emissions: Monitoring of Pollution (TEMPO) mission, launched in April 2023 as the first geostationary sensor dedicated to monitoring air quality over North America, provides both traditional aerosol optical depth (AOD) retrievals and novel aerosol layer height (ALH) products. In this study, we report on the use of AOD and ALH for estimating surface‐level particulate matter with a diameter less than 2.5 μm (PM 2.5 ) at hourly resolution using a geographically weighted regression (GWR) approach. We assess three methods: (a) using AOD alone, (b) applying multivariate linear regression with both AOD and ALH as inputs, and (c) using boundary layer AOD derived from column AOD and ALH as inputs to the GWR algorithm. Results show that the method using boundary layer AOD outperforms AOD‐only retrievals in regions where ALH exceeds 3 km and performs comparably to the AOD‐only method elsewhere. Ten‐fold cross‐validation yields an overall R 2 of 0.47, a mean bias of 0.10 μg/m 3 , and an RMSE of 6.08 μg/m 3 using the boundary‐layer AOD method. By contrast, the multivariate regression method yields the weakest performance. To enhance PM 2.5 estimation further, we develop a strategy that integrates TEMPO and Advanced Baseline Imager retrievals. In this approach, sensor averaged PM 2.5 estimates are used, except in two cases where TEMPO alone is favored due to its superior skill: when ALH exceeds 3 km or when TEMPO aerosol detection identifies blowing dust.
Zhang et al. (Fri,) studied this question.