Sentinel-2 and Methane Detection

New Study in Atmospheric Measurement Techniques Journal

Research team from GHGSat and Harvard University demonstrates Sentinel-2 capability for detecting large methane point sources and shares technique for quantifying emissions

Researchers from GHGSat and Harvard University have collaborated on a new study, published today in the journal Atmospheric Measurement Techniques, investigating the role the European Space Agency’s Sentinel-2 satellites can play in detecting methane from space.  Specifically, the paper documents methods for quantifying methane emissions from Sentinel-2 observations, a previously unreported capability.

Sentinel-2 is part of the European Union’s Copernicus Earth Observation program and comprises two nearly identical satellites, spaced 180° apart on the same orbital path. Launched in 2015 and 2017, each carries a wide-swath, high-resolution MultiSpectral Instrument (MSI) with 13 spectral bands. Together, the Sentinel-2 satellites image the Earth’s surface at 10-60 m resolution, with global coverage achieved every 2-5 days.

Although detecting methane was not one of their original tasks, the instruments can detect the gas using light measured from two neighbouring spectral bands: one in which light is much more strongly absorbed by methane than the other. By testing a series of methodologies, the team has shown that Sentinel-2 can see some of the world’s largest emission point sources, leaks above 3,000 kg/hr of methane.

“These techniques can be used to identify and monitor large point sources across the globe, which could lead to reductions in methane emissions,” explains Daniel Varon, a postdoctoral research fellow at Harvard University. “They can also be applied to similar multispectral instruments in orbit today, like Landsat, to improve our understanding of the biggest anthropogenic methane point sources.”

The GHGSat – Harvard team tested three methane quantification methods that use shortwave infrared (SWIR) measurements from MSI spectral bands 11 (~1560-1660 nm) and 12 (~2090-2290 nm) to detect atmospheric methane plumes. The most successful method used measurements from the two bands on two different dates: the date in which the plume of interest was present and a reference observation in which it was absent. This method can quantify point sources down to about 3,000 kg/hr with a methane column density precision of ~30%-90% of background over favourable terrain.

Figure 1: Demonstration of Sentinel-2 methane column retrieval for a plume emitted from a point source (31.6585°N, 5.9053°E) in the Hassi Messaoud oil field of Algeria, on 20 November 2019.
Left Sentinel-2 top of atmosphere reflectances for band-12 observed on 20 November 2019 when the plume is present in the middle of the image.
Right Multi-band/multi-pass (MBMP) retrieval for 20 November 2019.


The paper quantifies methane emissions for two locations observed between June 2015 and October 2020. The first location is a point source at a well pad in the Hassi Messaoud oil field of Algeria and the second is a point source at a compressor station in the Korpezhe oil/gas field of Turkmenistan.

A methane emission of 3,000 kg/hr represents the equivalent of 142,000 cars on the road for one year. To monitor the smaller, more numerous methane sources typical of normal operations at oil and gas facilities requires more sensitive instruments specifically tuned to detect methane, such as the spectrometers aboard GHGSat’s growing fleet of satellites in low Earth orbit. This research on Sentinel-2 is part of GHGSat’s commitment to sharing expertise to improve the understanding and detection of greenhouse gas emissions, continuing the work announced last October when the PULSE methane concentration map was released for free public use.

The paper can be read at Atmospheric Measurement Techniques , a journal of the European Geosciences Union