Enhancing and Localizing Methane Point Emissions using Wind-Rotation

Data provided by the TROPOMI sensor, of Copernicus Sentinel-5, has established itself as an important tool in the remote sensing of anthropogenic methane. With its repeat time of 1 day, TROPOMI data is excellent for detecting large emission events. Its spatial resolution, at 7x5.5km2, has previously been insufficient to pinpoint methane point sources. However, a processing technique can enhance TROPOMI data, and allow point sources to be localized with more confidence.

First reported by Maasakkers et al. [1], wind rotation is a simple, but effective, technique for enhancing methane plumes in TROPOMI imagery. In so-doing, it can be used to roughly localize methane emissions, and/or inform the targeting of more fine grained imagery, such as that produced by GHG-SAT [2].

The technique has 3 stages, for a given AOI and time window: firstly, the spatial resolution of TROPOMI data is enhanced via spatial binning; secondly, each TROPOMI product is rotated to true north, around a fixed point, based on historical wind data; lastly, the stack of TROPOMI products are reduced into a single, wind-rotated mean product. This process is then repeated over a grid of rotation centers. Candidate plumes are then evaluated, and the best enhancement selected. The point of rotation of the optimal plume corresponds to its point of emission.

The following example shows this part of this process, as applied to a coal mine methane emission in Russia from June 2022 [3]. As the rotation center gets closer to the point of emission, the plumes become more coherent.

Fig.1. grid of wind rotated methane rotations, for a Russian coal mine (June 2022). Plumes become more distinct, as we get closer to the suspected point of emission.

Maasakkers et al. [1] gave the example of a landfill in Buenos Aires (analysis here repeated). The left-most figure shows spatially binned, but un-rotated, mean TROPOMI data. The right-most figure shows the effect of wind rotation on plume enhancement.

Fig.2. The left figure shows mean TROPOMI data for the Buenos Aires AOI. The right figure shows wind-rotated mean TROPOMI data, with an optimal rotation center, resulting in a clear enhancement of a methane plume.

At GeoSynergy, we have been exploring how this technique can be developed, and applied to the detection of methane emissions from Australian coal mines. In some cases, where TROPOMI data is sufficiently rich, the technique can be applied to

monthly TROPOMI data, such as in the following examples (the point of rotation is marked with a white pixel, rotation center not fully optimized). Where wind-rotation is sufficiently high quality, it may even be possible to localize emission point sources to approximate mine infrastructure.

Fig.3. wind-rotated methane plume, coal mine in QLD, August 2018.

Interestingly, inactive mines are substantial methane emitters [4]. The following example shows a wind-rotated TROPOMI CH4 emission from an inactive mine in NSW (Fig.4).

Fig.4. mean, vs wind-rotated, methane for a closed mine NSW, November 2020.

We are currently undertaking research and development, to explore whether this technique can be applied to remote sensing of methane emissions from smaller targets, such as farm dams, which have been identified as significant [5]. Because farms dams are relatively small, it may be difficult to distinguish their emissions from atmospheric methane, using the wind rotation approach. An alternative approach, regional, may be better suited.

In a recent study by Huber et al. [6] the authors selected a test site, and 2 reference sites. They used "forested regions upwind of the cropland domain as reference sites, to estimate daily average background TROPOMI NO2" (figure 5 - dashed boxes). Reference sites provide "nearby, clean upwind domains containing few major NOx sources, which are ideal for background quantification". Which reference site they used varied, according to the predominate daily wind direction (historical ERA5 data). Days with predominantly N wind were excluded (due to potential urban emission contamination). "The reference sites facilitate the calculation of the NO2 column enhancement over the cropland domain by subtracting the average inflow background NO2 VCD from the average cropland NO2 VCD." We are currently in the process of testing this approach, using spatially binned TROPOMI data.

Fig.5. (a) Satellite image of the Mississippi Delta study region showing cropland (solid white box) and upwind (dashed white boxes) domains used to calculate daily NO2 enhancements.Light green regions are primarily cropland; dark green regions are primarily forest.(b) TROPOMI tropospheric NO2 VCDs on 14 May 2019, 5 days before a precipitation event.(c) TROPOMI tropospheric NO2 VCDs on 24 May 2019, 5 days after a precipitation event, with enhanced NO2 VCDs over cropland indicative of a drydown soil NOx emissions pulse [6].

Because of the spatial and temporal resolution of TROPOMI, screening of large areas is possible. The fact that candidate plumes need to be evaluated by eye, and the best

enhancement selected, is potentially limiting. However, with a sufficient dataset, it is plausible that a machine learning and/or deep learning model could be trained to predict “plume enhancement” from input plume candidates. This would enable almost real time wind-rotated TROPOMI screening, simultaneously, for multiple locations.

We will continue to discuss new satellite products, and new analysis modalities, in future blog posts.

References

[1] Maasakkers, J.D., Varon, D.J., Elfarsdóttir, A., McKeever, J., Jervis, D., Mahapatra, G., Pandey, S., Lorente, A., Borsdorff, T., Foorthuis, L.R. and Schuit, B.J., 2022. Using satellites to uncover large methane emissions from landfills. Science advances, 8(31), p.eabn9683.

[2] GHGSat, Methane emissions from opencast coal mines measured from space. https://www.ghgsat.com/en/newsroom/world-first-methane-emissions-from-opencast-coal-mines-measured-from-space/

[3] GHGSAT report of the Raspadskaya coal mine methane emission, Kemerovo Oblast, June 15 2022. https://www.ghgsat.com/en/newsroom/russian-mine-produces-biggest-methane-leak-ever-seen-by-ghgsat/

[4] Ember, Abandoned Mine Methane. https://ember-climate.org/insights/research/tackling-australias-coal-mine-methane-problem/

[5] Blue Carbon Labs, Australian Farm Dam methane. https://www.bluecarbonlab.org/farm-dams/

[6] Huber, D.E., Steiner, A.L. and Kort, E.A., 2020. Daily cropland soil NOx emissions identified by TROPOMI and SMAP. Geophysical research letters, 47(22), p.e2020GL089949.