Coal mine methane detection from ventilation shaft emissions

Underground coal mines emit methane through surface ventilation shafts around the clock, making them tractable point sources for spaceborne SWIR spectrometers. Detection is real but constrained by shaft arrays that dilute flux below single-pass thresholds.

Sensors

GHGSat-C series: Shortwave-infrared Fabry-Pérot spectrometer operating around 1.65 µm. Pixel footprint approximately 25 m, scene width roughly 12 km. Designed for point-source quantification; published minimum detectable emission rate around 100–400 kg CH4/hr for a single overpass under good atmospheric conditions, depending on wind speed and scene contrast.
Sentinel-5P TROPOMI: UV-SWIR push-broom spectrometer with a 7 × 5.5 km pixel (resampled to 5.5 × 3.5 km since August 2019 upgrade). Global daily coverage. Detects regional methane column enhancements; individual mine shafts are below its resolution, but clusters of mines in coal basins (Shanxi, Silesia, Bowen) produce statistically significant XCH4 anomalies in multi-week composites.
EMIT (ISS-mounted): JPL imaging spectrometer covering 380–2500 nm at roughly 60 m ground sampling distance. Opportunistic ISS orbit gives irregular revisit (not sun-synchronous). Published detections include coal mine methane plumes; the non-polar ISS inclination of 51.6° limits coverage of high-latitude coalfields.
Carbon Mapper (Planet/JPL): Tanager-1 launched 2024; SWIR imaging spectrometer at approximately 30 m resolution. Designed explicitly for point-source methane and CO2 quantification, with systematic campaign tasking. Published detection floor targets around 100 kg CH4/hr. Revisit depends on tasking priority, not a fixed cadence.

Why ventilation shafts are a tractable target

Coal seams contain adsorbed methane that desorbs as mining removes confining pressure. Regulations in most jurisdictions require active ventilation to keep shaft-air methane concentrations below explosive thresholds, typically 1–1.5% by volume. The consequence is continuous, high-volume air flow carrying diluted methane to surface. A single large ventilation shaft at a gassy longwall operation can exhaust millions of cubic metres of air per hour, with methane concentrations of 0.3–0.8% by volume, translating to emission rates that published studies place between a few hundred and several thousand kilograms of CH4 per hour for individual shafts.

That makes them meaningfully different from diffuse agricultural or wetland sources. The emission is spatially concentrated at a known surface structure, temporally continuous rather than episodic, and in principle verifiable against ventilation-flow records that regulators already hold. The physics is cooperative. The politics sometimes is not.

What the satellites can actually see, and what they miss

GHGSat's published work, including peer-reviewed results from its Claire and C-series satellites, demonstrates detection of individual coal mine ventilation plumes in China, Australia and the United States. The company has reported quantified emission rates from single shafts in the range of hundreds to low thousands of kg CH4/hr, consistent with independent bottom-up estimates. EMIT, operated by JPL on the International Space Station, has similarly published detections of coal mine methane plumes at roughly 60 m resolution, with some Chinese coalfield detections appearing in the 2022 Science paper by Thorpe et al. that accompanied the instrument's public data release.

The hard limit is the diffuse-shaft problem. Many older mines, particularly in China, Poland and Ukraine, vent through numerous small shafts spread across a wide surface footprint. Each shaft individually may emit below GHGSat's single-overpass detection floor of roughly 100–400 kg CH4/hr. The aggregate basin-level signal appears in TROPOMI composites, but attributing that signal to specific facilities, rather than the coal region as a whole, requires combining the coarse atmospheric column data with shaft location databases and wind-field modelling. Without shaft coordinates, TROPOMI anomalies are suggestive, not actionable.

Cloud cover is an unignorable constraint. SWIR spectrometers require clear-sky columns. In humid coalfield regions, persistent cloud can reduce usable overpass frequency to a fraction of nominal revisit rates. Wind speed matters equally: plumes must be slow enough to accumulate a detectable column above the shaft but fast enough to form a directional signature that aids source attribution. The published sweet spot for GHGSat-class instruments is winds of roughly 1–5 m/s.

Shaft location databases as the forcing function

Detection probability is not uniform across a coalfield. It is a function of where the sensor is pointed and whether the target coordinates are known in advance. Systematic detection campaigns, rather than serendipitous passes, require a prior shaft location database. Several sources exist in the public domain: national mining permit registers, the Global Coal Mine Tracker maintained by Global Energy Monitor, and historical topographic maps. These are incomplete and often years out of date, but they provide enough structure to prioritise tasking.

The workflow is straightforward in principle. Shaft coordinates go into a tasking request. The commercial operator schedules an overpass when weather forecasts suggest clear sky and low wind. The spectrometer acquires. Retrieval algorithms fit the observed radiance spectrum against a methane absorption model to produce an integrated column enhancement map, from which a flux estimate is derived using wind speed at plume height. The result is a quantified emission rate with an uncertainty range, not a binary detection. Honest reporting of that uncertainty is part of what makes the product useful to a regulator rather than a headline.

TROPOMI's role: basin-level accounting, not facility attribution

Sentinel-5P TROPOMI has produced the most comprehensive public record of coal-basin methane anomalies. Studies using multi-year TROPOMI composites have identified statistically significant XCH4 enhancements over the Shanxi and Inner Mongolia coalfields in China, the Upper Silesian Coal Basin in Poland, and the Bowen Basin in Australia. These basin-level signals are consistent with national inventory estimates in some cases and significantly higher in others, which is precisely the policy-relevant finding.

The attribution limit is real. A TROPOMI pixel at 5.5 × 3.5 km integrates over dozens of potential sources: ventilation shafts, coal processing facilities, post-mining surface emissions, and any co-located agricultural or waste sources. Disentangling these requires either higher-resolution coincident observations or a statistical inversion that uses wind fields and source priors. TROPOMI is best understood as a screening tool that flags which basins warrant targeted high-resolution tasking, not as a facility-level enforcement instrument on its own.

Honest limits and the monitoring gap they create

No current spaceborne system provides daily, all-weather, facility-level methane quantification for coal mines. GHGSat and Carbon Mapper offer the necessary resolution but rely on clear-sky tasking, which in practice means a given mine might receive two to four cloud-free observations per month in favourable climates and fewer in monsoon-affected or high-latitude regions. EMIT is not a monitoring system; it is a research instrument with opportunistic coverage.

The resulting monitoring gap is largest for small mines with dispersed shaft arrays in persistently cloudy regions. These are also, not coincidentally, the mines least likely to report accurately to national inventories. Combining TROPOMI basin anomalies with shaft databases and periodic GHGSat or Carbon Mapper tasking is the current best practice for closing that gap without in-situ access. It is imperfect. A regulator should understand they are getting probabilistic evidence, not a continuous meter reading.

Satellize runs this multi-sensor fusion workflow on open and commercial data, with shaft-database integration as a standard preprocessing step.

Typical figures

GHGSat-C spatial resolution	~25 m pixel, ~12 km scene width
TROPOMI pixel size	5.5 × 3.5 km (post-August 2019 upgrade)
EMIT ground sampling distance	~60 m
Carbon Mapper (Tanager-1) resolution	~30 m
TROPOMI revisit	Daily global coverage (Sentinel-5P sun-synchronous orbit)
GHGSat / Carbon Mapper revisit	Tasking-dependent; not fixed cadence. Typically days to weeks between cloud-free acquisitions of a specific site
Spectral band for CH4 retrieval	SWIR ~1.65 µm (GHGSat, Carbon Mapper); SWIR 2.3 µm also used by TROPOMI and EMIT
Minimum detectable emission rate (GHGSat-class)	Published range ~100–400 kg CH4/hr per shaft under clear sky, winds 1–5 m/s
Cloud constraint	All SWIR instruments require clear-sky column; optical depth >0.1 typically degrades retrieval quality
Archive depth	TROPOMI: from May 2018 (operational). GHGSat commercial archive: from 2019 (Claire) and 2020 onwards (C-series). EMIT: from August 2022.

Analytics Satellize can run

Shaft-level emission rate estimate	GHGSat or Carbon Mapper column retrieval combined with meteorological wind-field data; integrated mass balance approach	Per-shaft quantified flux report (kg CH4/hr) with uncertainty bounds, delivered as PDF and GeoJSON point layer
Basin-scale XCH4 anomaly map	TROPOMI multi-week composite with background subtraction and wind-rotated plume averaging	Gridded NetCDF anomaly layer and summary report comparing observed enhancement against national inventory estimates
Shaft prioritisation ranking	Overlay of public shaft location database against TROPOMI anomaly hotspots; ranked by modelled contribution probability	Tasking priority list with shaft coordinates and recommended overpass windows, formatted for direct submission to commercial operator
Temporal trend analysis	Time-series extraction from TROPOMI archive (2018 to present) at basin or facility level; seasonal decomposition	Annual trend chart and statistical summary indicating whether emissions are increasing, stable or declining relative to reported production changes
Detection gap assessment	Cloud-frequency analysis using ERA5 or Sentinel-2 cloud masks over target coalfield; modelled clear-sky overpass probability	Monitoring feasibility report quantifying expected cloud-free acquisition rate by month and season
Regulatory evidence package	Collation of GHGSat or Carbon Mapper quantified detections with shaft permit records and national inventory submissions	Structured evidence dossier suitable for submission to environmental regulator, including data provenance and retrieval methodology documentation

Sources

Who does the work

We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.