Sudhanshu Pandey , John Worden, Daniel H Cusworth, Daniel J. Varon , Matthew D Thill, Daniel J. Jacob, Kevin W. Bowman

Methodologies for inferring surface emissions of atmospheric trace gases can be categorized into plume detection and area-scale estimation. Plume detections are observations of emissions from either individual or clustered point sources. Area estimates are derived from top-down atmospheric flux inversion models or bottom-up inventories, which infer mean emissions typically over spatial scales greater than 10 km and temporal scales greater than a week. Integrating information from these distinct methodologies can enhance our understanding of emission sources and improve the accuracy of emission estimates. However, such integration is challenging because plume-detecting instruments exhibit irregular and infrequent sampling, as well as varying detection sensitivities and spatial footprint sizes. In this study, we present a theoretical framework to relate plume and area estimates of dense point source methane emission fields. We show that the spatial footprint size of plume-detecting instruments impacts the emission rate distribution of plumes. In empirical tests, we find a robust linear relationship between the sums of gridded plume emission rates and area estimates for the Permian Basin’s oil and gas emissions. After accounting for the plume detectors' sampling of the Permian emission field, the weekly plume sums demonstrate a strong correlation with TROPOMI top-down area estimates (R2 > 0.94, P < 0.005). We assess the feasibility of using plume data to inform area estimates within a Bayesian assimilation framework. We perform two plume inversions using prior area estimates from (1) constant EDF bottom-up inventory and (2) weekly-updated TROPOMI inversion estimates. We find that the posterior estimate of the EDF plume inversion improves, bringing it in good agreement with independent TROPOMI estimates. In the TROPOMI plume inversion, the fine spatial resolution features of area estimates improve. Our analysis underscores that plume datasets obtained from aircraft, satellites, and in situ instruments can evaluate and improve area estimates of dense point source emission fields.