MethanE DART: The Methane Emissions-Detection, Analysis, and Resource management Tool

Increases in methane emissions over the past decade have outpaced other greenhouse gases. Methane, a potent greenhouse gas, is vital to understanding climate change and mitigation strategies. Fugitive emissions from the natural gas industry represent the most preventable of anthropogenic emissions and will be the target of a direct tax penalty under the Inflation Reduction Act. Newly available satellite data from several missions provide an opportunity to monitor methane emissions from space at finer spatial resolutions. We propose a new geospatial analysis platform, the Methane Emissions-Detection, Analysis, and Resource Management Tool (MethanE DART), which will empower researchers and policymakers to effectively utilize the emerging satellite datasets, enabling the rapid identification, analysis, and categorization of point-source methane emissions. The software will incorporate three separate machine learning (ML) processes consisting of (1) an object-detection ML analysis to identify distinct point-source emitters, (2) a supervised neural network to estimate the flux of methane attributable to the point-source, and (3) a classification ML analysis to identify the industry of the identified emitters. The proposed analysis platform will provide an unparalleled view of global-scale industry-wide methane emissions, empowering natural gas operators, federal oversight entities, and researchers to make informed decisions and develop effective mitigation strategies.