High-frequency monitoring of anomalous methane point sources with multispectral Sentinel-2 satellite observations

Abstract

Abstract. We demonstrate the capability of the Sentinel-2 MultiSpectral Instrument (MSI) to detect and quantify anomalously large methane point sources with fine pixel resolution (20 m) and rapid revisit rates (2–5 d). We present three methane column retrieval 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 is the multi-band–multi-pass (MBMP) method, which uses a combination of the two bands and a non-plume reference observation to retrieve methane columns. The MBMP method can quantify point sources down to about 3 t h−1 with a precision of ∼ 30 %–90 % (1σ) over favorable (quasi-homogeneous) surfaces. We applied our methods to perform high-frequency monitoring of strong methane point source plumes from a well-pad device in the Hassi Messaoud oil field of Algeria (October 2019 to August 2020, observed every 2.5 d) and from a compressor station in the Korpezhe oil and gas field of Turkmenistan (August 2015 to November 2020, observed every 5 d). The Algerian source was detected in 93 % of cloud-free scenes, with source rates ranging from 2.6 to 51.9 t h−1 (averaging 9.3 t h−1) until it was shut down by a flare lit in August 2020. The Turkmen source was detected in 40 % of cloud-free scenes, with variable intermittency and a 9-month shutdown period in March–December 2019 before it resumed; source rates ranged from 3.5 to 92.9 t h−1 (averaging 20.5 t h−1). Our source-rate retrievals for the Korpezhe point source are in close agreement with GHGSat-D satellite observations for February 2018 to January 2019, but provide much higher observation density. Our methods can be readily applied to other satellite instruments with coarse SWIR spectral bands, such as Landsat-7 and Landsat-8. High-frequency satellite-based detection of anomalous methane point sources as demonstrated here could enable prompt corrective action to help reduce global methane emissions.

Automated detection of atmospheric NO<sub>2</sub> plumes from satellite data: a tool to help infer anthropogenic combustion emissions

“…Our approach will also miss direct CO 2 and methane emissions, e.g., pipeline leaks, coal mines (Palmer et al, 2021). For these sources, we still have to rely on high spatially-resolved CO 2 and methane data (Varon et al, 2021). In contrast, we also find persistent NO 2 plumes from regions where ODIAC does not currently include CO 2 emissions.…”

Joint Use of in-Scene Background Radiance Estimation and Optimal Estimation Methods for Quantifying Methane Emissions Using PRISMA Hyperspectral Satellite Data: Application to the Korpezhe Industrial Site

“…These sources belong to the Korpezhe oil and gas field (38.52 • N, 54.21 • E). One of these emission points (D panel in Figure 2) is near a compressor station and has been identified as a strong CH 4 emitter [56]. On this industrial site, we were able to recover eleven images with cloudless sky or with few clouds between 2020/04/19 (YYYY/MM/DD) and 2021/06/22.…”

Mapping methane plumes at very high spatial resolution with the WorldView-3 satellite

“…The Algeria site covers a largely homogeneous area similarly to Turkmenistan although it contains some cirrus/cloud shadows at the right half of the selected subset that alter the radiance in SWIR B7. 235 Finally, it is important to remark that the sites selected for the simulations are realistic insofar as they are well-known areas of O&G and coal production facilities with known methane emissions (Irakulis-Loitxate et al, 2021a;2021b;Varon et al, 2021;Cusworth et al, 2021).…”

“…GHGSat and these hyperspectral missions share a relatively high sensitivity to methane concentration enhancements and a spatial sampling in the range of 25-50 m, but their acquisitions are sparse in time and space. The Sentinel-2 (S2) multispectral mission, originally developed for land applications, can help to overcome this limitation with frequent and spatially-continuous observations (Varon et al, 2021), albeit with a substantially lower sensitivity to methane. Landsat-7 and 8 missions, whose spectral channel configuration can provide a methane detection sensitivity comparable to S2 (see Fig.…”