Methane (CH₄) is the second most abundant greenhouse gas in the atmosphere after carbon dioxide. On a mass-to-mass basis, methane has a much higher global warming potential than CO₂—about 80 times greater over a 20-year period and 30 times greater over a 100-year period after its release. Tracking CH₄ leaks is essential for accurate emissions reporting, especially in the petroleum industry. Satellite hyperspectral sensors in the VNIR-SWIR range have demonstrated reliable performance in detecting CH₄ emissions.

However, although they provide information on total column gas in the atmosphere, their spatial resolution is insufficient for mapping localized CH₄ plumes. The Carbon Mapper mission, for example, aims to address this limitation with a constellation of satellites. Its first satellite, Tanager-1, launched in 2024, offers a spatial resolution of 30 m and a predicted minimum detection threshold of ~66–144 kg CH₄/h for point sources. GHGSat has a constellation of eleven satellites for detection and quantification of methane emissions (>≈ 100 kg/hr) from individual industrial sites.

Airborne hyperspectral imaging also enables high accurate methane concentration retrievals – the MethaneAIR proved able to detect point sources above 100-200 kg/h, with a conservative detection limit around 120 kg/h. AVIRIS-NG studies have shown a 90% Probability of Detection (POD) ranging from 16-33 kg/hr, with actual detection varying significantly with wind speed. Although airborne and spaceborne sensors are well established for methane plume detection, the potential of droneborne instruments for the same task remains rather underexplored.

To address this gap and improve the mapping of discrete methane sources, we evaluate a scientific-grade hyperspectral droneborne imaging sensor for detecting CH₄ in two distinct environments: a tropical soil area (red latosol) and a neighboring grass-covered area. In both fields, images were acquired during controlled methane release experiments with varying flow rates (0.5, 5, 10 and 25 m3/h) and flight altitudes (20, 60, 100 meters high). The notion is to assess the potential of unmanned aerial vehicles (UAVs) for mapping and characterizing very small CH₄ plumes using sources in arid versus vegetated environments.

Results showed that all point sources tested were mapped at all drone altitudes employed in the experiment and using both bare soil and green-dry vegetation backgrounds. This achievement is here considered a “proof of concept” and opens a window of opportunity for droneborne application in detection of CH₄ leaks along pipelines, refineries and other industrial installations, and its used could be extended for detection of seepages derived from subsurface reservoirs.