High resolution mapping of methane emissions from marine and terrestrial sources using a Cluster-Tuned Matched Filter technique and imaging spectrometry

• Cluster-Tuned Matched Filter technique permits mapping local CH4 anomalies.
• Distinct anomalies consistent with known CH4 sources and local wind direction.
• Sources include Coal Oil Point seep field, La Brea Tar Pits, and hydrocarbon tanks.
• True anomalies and false positives identified with segmentation and classification.
• Technique could be applied to additional trace gasses with distinct absorptions.

In this study, a Cluster-Tuned Matched Filter (CTMF) technique was applied to data acquired by the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) over marine and terrestrial locations known to emit methane (CH4). At the Coal Oil Point marine seep field, prominent CH4 anomalies were consistent with advection from known areas of active seepage. For a region with natural CH4 and oil seepage located west of downtown Los Angeles, significant CH4 anomalies were identified for known sources at the La Brea Tar Pits and in close proximity to probable sources, including an office complex documented as venting CH4 continuously and hydrocarbon storage tanks on the Inglewood Oil Field. However, interpretation of anomalies was complicated by noise and false positives for surfaces with strong absorptions at the same wavelengths as CH4 absorption features. Segmentation of results identified 16 distinct locations of contiguous pixels with high CTMF scores and segments were classified into probable CH4 anomalies and confusers based on the spectral properties of the underlying surface over the full spectral range measured by AVIRIS. This technique is particularly well suited for application over large areas to detect CH4 emissions from concentrated point sources and should permit detection of additional trace gasses with distinct absorption features, including carbon dioxide (CO2) and nitrous oxide (N2O). Thus, imaging spectrometry by an AVIRIS-like sensor has the potential to improve high resolution greenhouse gas mapping, better constraining local sources.