Fully Automatic Trace Gas Plume Detection
For remote sensing and environmental monitoring, this work automates plume detection from large-volume spectrometer data, reducing human oversight and enabling detection of understudied gases.
The paper presents a fully automated framework for detecting trace gas plumes in imaging spectrometer data, combining machine learning and physics-based fitting. Applied to EMIT data, it achieves automatic detection with negligible false positives and reveals that at least 25% of plumes may have been missed by human review, while also detecting NH3, NO2, and the first CO plume in EMIT imagery.
Future imaging spectrometers will increase data volumes by orders of magnitude, requiring automated detection of trace gas point sources. We present a fully automated framework that combines machine learning-based morphological analysis with physics-based spectroscopic fitting to detect plumes without human participation. Applied to EMIT imaging spectrometer data, the system operates in two modes: "daily digest" that runs automatically on all downlinked data, flagging the largest events for immediate response, and a retrospective analysis that identifies plumes missed by prior human review. The daily digest demonstrates that a significant fraction of the largest plumes can be detected automatically with negligible false positives, while retrospective analysis suggests at least 25% of plumes may have been overlooked. In addition to the previously observed methane point sources, we extend detection to three understudied trace gases: NH3, NO2 and the first observations of carbon monoxide (CO) plume in EMIT imagery.