Identification of false methane plumes for orbital imaging spectrometers: A case study with EMIT

Orbital imaging spectroscopy in the shortwave infrared is a powerful method for detecting and monitoring both anthropogenic and natural methane point sources. Analysts map the absorption of methane by applying spectroscopic detection methods like the matched filter or Differential Optical Absorption Spectroscopy (DOAS) independently at each pixel in a scene. Unfortunately, false detections due to instrument noise and surface spectroscopy features that mimic gas methane absorption features complicate detections of subtle methane enhancements at low signal levels. Current systems rely on manual review to vet these detections. This review process is time consuming and can be subjective for plumes close to the instrument detection limit. As observations scale up from proof of concept studies with hundreds of plumes to global monitoring systems having thousands, improved automation will be necessary to reduce the cost and subjectivity of manual review. This paper presents a new purely spectroscopic approach for rejecting false positives based on the aggregate plume transmittance. Specifically, we use the total plume transmittance as a discriminator with improved model properties and noise statistics that can dramatically improve false positive rejection rates. A sampling approach provides rigorous detection p-values that account for spatial correlations in background interference. We demonstrate the approach on a catalog of thousands of plumes from the EMIT (Earth Surface Mineral Dust Source Investigation) imaging spectrometer, and illustrate its use in the context of a methane plume mapping workflow. This process identifies 102 potential errors in the existing review process: 59 likely false positives and 43 false negatives. Including spectral information during the plume review process can disambiguate marginal detections, making methane point source observing campaigns more robust and sensitive than a review process based primarily on plume morphology.