Fugitive methane detection at municipal landfills
Municipal landfills are among the most variable methane sources on Earth, with emissions spiking episodically and mixing point-source leaks with diffuse surface flux. Satellite retrievals must overcome heterogeneous waste-surface emissivity before a plume rate means anything.
Sensors
- GHGSat-C series: Shortwave-infrared imaging spectrometer at roughly 25 m pixel resolution and sub-kilometre swath. Designed for point-source attribution; GHGSat has publicly documented landfill plumes exceeding 1,000 kg CH₄/hour at individual sites. Tasked on demand, so revisit is driven by contract rather than orbital repeat.
- Sentinel-5P TROPOMI: Daily global coverage at 5.5 × 3.5 km pixel resolution (post-2019 update). Detects column-averaged XCH₄ anomalies; landfill signals are typically too small to isolate at a single site but become visible when aggregated over urban waste clusters or when a very large facility dominates a pixel. Free, open archive from 2018.
- EMIT (ISS-mounted imaging spectrometer): NASA EMIT covers the full VSWIR range at roughly 60 m resolution per pixel, with a 75 km swath. ISS orbital geometry gives irregular revisit (days to weeks at mid-latitudes). EMIT has published methane point-source detections at landfills; the wide spectral range helps constrain surface emissivity across heterogeneous waste materials.
- Carbon Mapper (Tanager-1 and Planet partnership): Airborne and, from 2024, satellite-borne imaging spectrometer at approximately 30 m resolution. Carbon Mapper's published detection threshold is roughly 10–25 kg CH₄/hour for strong point sources under good atmospheric conditions; diffuse landfill emissions near that floor remain ambiguous.
- Airborne AVIRIS-NG / GAO (context sensor): Not a satellite, but the published baseline for landfill plume science. Sub-metre to 3 m resolution at flight altitude; used to validate satellite retrievals and to map spatial heterogeneity of surface emissivity across waste cells. Relevant when a regulator needs court-ready flux estimates.
Why landfills are harder than oil wells
A gas well leaking methane presents a single point source against a relatively uniform background. A landfill presents dozens of competing signals at once: active tipping faces, capped cells under plastic sheeting, gas-collection headers, flare stacks that may or may not be lit, leachate ponds, and decomposing organic layers at varying depths. Each surface has a different reflectance in the shortwave infrared, which is exactly the spectral region satellite sensors use to detect CH₄ absorption at 1.65–2.3 µm.
The consequence is a difficult background-fit problem. Satellite methane retrievals work by comparing observed SWIR radiance against a modelled clear-sky spectrum; the methane column is inferred from the residual absorption. When the surface beneath a plume is spectrally heterogeneous, as it is over mixed waste, the background model can either over- or under-estimate the continuum, introducing flux errors that can exceed 30 percent in published sensitivity studies. High spatial resolution sensors such as GHGSat-C and EMIT reduce this problem by keeping individual pixels more spectrally homogeneous, but they do not eliminate it.
Episodic emissions and the pressure-drop problem
Landfill methane flux is not constant. Barometric pressure drops reduce the overburden on waste cells, allowing gas to migrate laterally and vent at the surface. Temperature and precipitation also modulate microbial methanogenesis rates. The practical implication is that a satellite overpass on a high-pressure, cold morning may record a fraction of the flux that would be measured 48 hours later during a frontal passage.
This makes revisit frequency a genuine scientific constraint, not just a commercial convenience. TROPOMI's daily global pass is useful for tracking persistent anomalies over very large sites, but its 5.5 km pixel resolution means it integrates signal from surrounding urban areas and agriculture, making attribution to a specific landfill difficult without ancillary wind data and dispersion modelling. GHGSat and Carbon Mapper provide the spatial resolution needed for attribution but are tasked instruments, so catching an episodic emission event requires either scheduled repeat visits during meteorologically favourable windows or an alert-driven tasking protocol tied to barometric pressure forecasts.
What the retrieval actually measures, and what it does not
Satellite sensors measure column-averaged dry-air mole fractions of CH₄ (XCH₄) above the surface. Converting that to a mass emission rate requires a wind field, a plume length, and assumptions about atmospheric mixing height. Published methods include the cross-sectional flux integration approach, in which the plume is treated as a curtain of enhanced concentration perpendicular to the wind, and Gaussian plume inversion. Both methods carry uncertainty that grows when wind speed is below roughly 2 m/s (plume stagnates and spreads unpredictably) or above roughly 10 m/s (plume disperses faster than the sensor can resolve it).
Diffuse surface emissions, which are the dominant pathway at many well-managed landfills with functioning gas-collection systems, are particularly hard to quantify from orbit. A distributed flux of 50 kg/hour spread across a 20-hectare site may produce no detectable plume signature at all at GHGSat resolution, even though the cumulative annual emission is significant. Ground-based flux chambers or eddy-covariance towers remain the only reliable method for characterising that diffuse component. Satellite data is most defensible as a tool for detecting and quantifying the large, discrete leaks from gas-collection infrastructure, not for producing a whole-site mass balance.
Building a detection programme that holds up to regulatory scrutiny
Regulators and operators increasingly want satellite methane data to support enforcement or permit decisions. That requires a clear chain from raw radiance to reported flux, with documented uncertainty at each step. The minimum credible workflow involves: a quality-filtered retrieval (cloud fraction below roughly 5 percent over the scene, aerosol optical depth below 0.3); a co-located wind product at the time of overpass, typically from ERA5 reanalysis or a numerical weather prediction model; plume delineation with a documented threshold above background; and a flux estimate with a stated confidence interval.
For sites where gas-collection systems are required by permit, satellite detections of large plumes can trigger ground-based inspection. That is a realistic and proportionate use of the technology. Attempting to use a single satellite overpass as a definitive mass-balance audit is not, and any programme that claims otherwise deserves scepticism. The most effective regulatory deployments combine scheduled satellite tasking at high-pressure-drop windows with ground-truth from periodic airborne campaigns.
Archive depth and what historical data can show
TROPOMI's open archive runs from late 2017, giving roughly seven years of daily XCH₄ columns globally. For a large landfill that dominates its surrounding landscape, trend analysis over that archive can reveal whether emissions have increased as the site has grown or decreased after a gas-collection upgrade. The analysis requires careful deseasonalisation and wind-normalisation, but the method is well-established in the peer-reviewed literature.
GHGSat's commercial archive is shorter and access is licensed, but the company has published retrospective analyses of specific sites. EMIT data is publicly available via NASA Earthdata from its 2022 launch onwards. Carbon Mapper's satellite data is newer still. For most municipal landfill operators, the practical archive for high-resolution point-source work begins in 2022 at best, which limits trend analysis but is sufficient for baseline characterisation and incident investigation.
Satellize runs retrieval and dispersion workflows on open constellations including TROPOMI and EMIT, and can add commercial GHGSat tasking on client licence for sites where point-source resolution is required. The analytics approach follows the same published methods used in peer-reviewed landfill studies, with uncertainty ranges reported explicitly rather than suppressed.
Typical figures
| Best available spatial resolution (satellite) | ~25 m (GHGSat-C); ~60 m (EMIT); 5.5 × 3.5 km (TROPOMI) |
| Revisit (open archive) | Daily global (TROPOMI); irregular, days to weeks at mid-latitudes (EMIT via ISS) |
| Revisit (commercial tasking) | On-demand, subject to cloud and solar geometry; typically 1–5 day response (GHGSat, Carbon Mapper) |
| Key spectral bands | SWIR 1.65 µm and 2.3 µm CH₄ absorption windows; TROPOMI uses 2305–2385 nm |
| Minimum detectable flux (point source, favourable conditions) | ~10–25 kg CH₄/hour (Carbon Mapper published threshold); ~50–100 kg/hour practical floor for GHGSat at landfill backgrounds |
| Diffuse emission detectability | Not reliably detectable from orbit; ground-based flux chambers required |
| Wind speed operating range | 2–10 m/s optimal for plume flux integration; outside this range, flux uncertainty increases substantially |
| Cloud fraction limit | Scene cloud fraction below ~5% required for valid SWIR retrieval |
| Open archive depth | TROPOMI from October 2017; EMIT from August 2022; Landsat context imagery from 1972 |
| Delivery formats | NetCDF (TROPOMI, EMIT standard products); GeoTIFF plume masks; GIS-ready shapefiles; PDF regulatory summary reports |
Analytics Satellize can run
| Large-leak alert | Threshold detection on TROPOMI XCH₄ anomaly above site-specific background, filtered by wind direction and cloud fraction | Automated alert with estimated plume bearing and indicative flux range, delivered within 24 hours of overpass |
| Point-source flux estimate | Cross-sectional integrated mass enhancement (IME) or Gaussian plume inversion applied to GHGSat-C or EMIT radiance retrievals, co-located with ERA5 wind | Single-event flux report with stated uncertainty interval, suitable for regulatory submission |
| Multi-year emission trend | Wind-normalised XCH₄ anomaly time series from TROPOMI archive; deseasonalised regression against site operational records | Annual trend chart with confidence bounds; comparison against permit baseline |
| Surface emissivity characterisation | EMIT full VSWIR spectral mapping of waste-cell surface types to inform background-fit quality flags on subsequent CH₄ retrievals | GIS layer of surface spectral classes across the landfill footprint, with per-class retrieval reliability score |
| Pressure-correlated tasking schedule | Barometric pressure forecast integration to identify high-emission-probability windows; triggers commercial satellite tasking requests | Rolling 10-day tasking calendar with meteorological justification; pass confirmation log |
| Gas-collection system performance indicator | Comparison of satellite-detected plume flux against permitted collection efficiency; flags sites where observed emissions are inconsistent with reported capture rates | Quarterly compliance summary report with site-by-site anomaly flags |
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.