Methane leak detection along gas pipeline corridors
Shortwave infrared spectroscopy from orbit can detect methane column anomalies above gas pipeline corridors, pinpointing leaks from individual flanges to diffuse corridor-wide seepage. Three complementary satellite systems now cover detection thresholds from sub-100 kg/hr point sources to basin-scale plumes.
Sensors
- GHGSat-C series (Claire, Iris, Luca, Oliver, Hugo and others): Dedicated wide-angle Fabry-Pérot imaging spectrometers targeting the 1.65 µm and 2.3 µm methane absorption bands. Ground sampling distance approximately 25 m; scene footprint roughly 12 × 12 km. Minimum detectable emission rate quoted by GHGSat at around 100 kg/hr for a point source under good atmospheric conditions. Tasked on demand; revisit over a specific site depends on constellation size and scheduling, typically one to several days with the current fleet of more than ten satellites.
- Sentinel-5P TROPOMI: ESA/Copernicus UV-SWIR imaging spectrometer covering the 2305–2385 nm methane band at 5.5 × 7 km (reprocessed to 5.5 × 3.5 km since August 2019). Daily global coverage. Detection threshold is roughly 1–2 ppb column enhancement above background, which in practice means large plumes or persistent leaks aggregating tens of thousands of kg/hr are reliably detected. Individual pipeline leaks below that scale are generally below detection unless multiple adjacent sources combine.
- EMIT (Earth Surface Mineral Dust Source Investigation, ISS): NASA JPL imaging spectrometer covering 380–2500 nm at 60 m ground sampling distance. Mounted on the International Space Station; coverage is non-systematic and latitude-limited to roughly 52° N/S, with revisit governed by ISS orbital precession. Published studies have confirmed detection of individual super-emitter plumes exceeding approximately 500 kg/hr. Provides hyperspectral context useful for distinguishing methane from interfering surface mineralogy.
- MethaneSAT: Environmental Defense Fund satellite launched March 2024. Wide-field spectrometer covering the 1.65 µm methane band; 200 km swath at approximately 100–400 m effective resolution for area-flux mapping. Designed specifically for oil and gas basin-scale surveys, producing emission-rate maps rather than single-pixel column retrievals. Data access policy is open for research; commercial pipeline use is via licensing arrangements.
Why 2.3 micrometres, and what it actually tells you
Methane has strong vibrational absorption features centred near 1.65 µm and 2.3 µm. At these wavelengths, sunlight reflected from the ground passes twice through the atmospheric column, so any excess methane between the surface and the sensor attenuates the signal in a characteristic pattern that retrieval algorithms can separate from water vapour and carbon dioxide interference. The output is a column-averaged dry-air mole fraction, XCH4, measured in parts per billion. A leak shows up as a localised enhancement above the regional background, which itself varies by season and latitude.
The physics sets hard limits. Retrievals require reflected sunlight, so night-time acquisition is impossible and high solar zenith angles degrade precision. Thick cloud completely blocks the signal; thin cirrus introduces retrievable but uncertain bias. Aerosol loading over dust-prone corridors, common in the Middle East and Central Asia, raises retrieval uncertainty. Any serious pipeline monitoring programme needs to account for the fraction of acquisition attempts that will be clouded out, which in humid tropical regions can exceed 60% of days in wet season.
Three instruments, three different problems they solve
TROPOMI is the workhorse for regional situational awareness. Its daily global pass means no pipeline corridor goes unobserved for more than 24 hours, and its long archive (October 2017 onwards) allows baseline characterisation of background methane and identification of persistent anomalies. The limitation is coarse resolution: a 5.5 × 3.5 km pixel integrates everything beneath it. A single small compressor-station leak is invisible; a cluster of leaks along a 50 km corridor segment may just cross the detection threshold if the wind is favourable.
GHGSat's constellation bridges the gap to point-source attribution. At 25 m sampling, a plume from a single flange, valve or pig-trap vent is spatially resolved and can be attributed to a specific facility. The trade-off is that each scene covers only about 144 km², so systematic corridor coverage requires multiple tasking passes and costs accordingly. The workflow in practice is to use TROPOMI to flag anomalous segments, then task GHGSat over the flagged area for attribution.
EMIT adds an independent hyperspectral check for large events. Its 60 m pixels and 1 km swath width sit between the two in spatial terms, and because it covers the full SWIR spectrum rather than narrow methane bands alone, it can identify co-emitted species and rule out spectral confusion with certain surface minerals. It is not a systematic monitoring tool; it is a confirmation layer.
Turning column concentrations into emission rates
A column-concentration map is not an emission rate. Converting one to the other requires a wind field. The standard published approach, the integrated mass enhancement (IME) method, integrates the excess methane mass across a plume and divides by an effective transport time derived from reanalysis wind data such as ERA5. Published validation studies put uncertainty on individual retrievals at roughly a factor of two under good conditions, narrowing with multiple passes or with in-situ wind measurements co-located in time.
For diffuse corridor seepage rather than point sources, a flux-inversion approach is more appropriate: fitting an atmospheric transport model to the observed concentration field to back-calculate the distributed source term. MethaneSAT's wide-swath design was specifically engineered for this, allowing basin-level emission inventories to be constructed from a single pass. The method demands careful treatment of boundary conditions and is sensitive to the assumed atmospheric mixing height, which introduces additional uncertainty in stable nocturnal boundary layers.
What the data cannot do on its own
Satellite methane detection identifies where and approximately how much. It does not identify why. A positive detection over a pipeline corridor is consistent with a flange leak, a deliberate vent, a pig-trap operation, a nearby landfill, or a geological seep. Attribution to a specific cause requires ground-truth: inspection records, operational logs, or drone or vehicle-mounted sensors deployed to the flagged location. The satellite layer is most valuable as a prioritisation tool that directs limited ground-inspection resources to the highest-probability segments.
Spatial resolution also limits what can be attributed confidently. At GHGSat's 25 m, a plume origin can be localised to within a few hundred metres under favourable wind conditions. That is sufficient to implicate a compressor station but may not distinguish between two adjacent facilities on a crowded processing site. Regulatory or legal proceedings typically require follow-up measurement at much higher spatial and temporal resolution than any current orbital system provides.
Building a monitoring programme that regulators and operators will accept
Regulators in the EU (under the Methane Regulation 2024/1787) and the US EPA are increasingly accepting satellite data as a screening tool, but not yet as primary evidence for enforcement. The practical implication is that a satellite-based programme needs a defined escalation protocol: TROPOMI flags a regional anomaly, GHGSat confirms and localises, ground inspection measures and attributes, and the whole chain is documented with retrieval uncertainty estimates at each stage.
Archive depth matters here. TROPOMI's record back to late 2017 allows operators to demonstrate whether an anomaly is new or persistent, which has direct bearing on regulatory exposure. GHGSat's commercial archive is shorter but growing. Satellize runs TROPOMI-based corridor screening as part of its analytics stack, producing segment-level anomaly reports with uncertainty bounds, and can add commercial GHGSat tasking on client licence for attribution follow-up. The escalation logic, thresholds and reporting cadence are configured per corridor rather than applied as a generic product.
For any operator considering this seriously: start with a baseline quarter of TROPOMI analysis over your corridor before any incident occurs. Knowing the seasonal background and the typical retrieval success rate under your local cloud climatology is the prerequisite for interpreting any future anomaly with confidence.
Typical figures
| Spatial resolution (GHGSat-C) | ~25 m ground sampling distance; ~12 × 12 km scene footprint |
| Spatial resolution (TROPOMI) | 5.5 × 3.5 km per pixel (since August 2019 reprocessing) |
| Spatial resolution (EMIT) | 60 m ground sampling distance; ~75 km swath |
| Revisit (TROPOMI) | Daily global coverage; ~1-day repeat at mid-latitudes |
| Revisit (GHGSat constellation) | On-demand tasking; 1 to several days depending on scheduling and site latitude |
| Spectral bands used | 1.65 µm and 2.3 µm methane absorption windows (SWIR); TROPOMI covers UV through SWIR |
| Minimum detectable emission rate (GHGSat point source) | ~100 kg/hr under good atmospheric conditions (GHGSat published specification) |
| Minimum detectable enhancement (TROPOMI) | ~1–2 ppb XCH4 column anomaly; practical threshold for isolated leaks is large or aggregated sources |
| Archive depth | TROPOMI: October 2017 to present; GHGSat commercial: 2016 (Claire) to present; EMIT: 2022 to present |
| Key data limitation | Cloud cover blocks all SWIR retrievals; aerosol and high solar zenith angle degrade precision; night acquisition impossible |
Analytics Satellize can run
| TROPOMI corridor baseline and anomaly screening | XCH4 column retrieval differential against seasonal background; pixel-level z-score flagging over defined pipeline buffer zone | Monthly segment-level anomaly report with retrieval success rate, flagged dates and uncertainty bounds; GIS polygon layer |
| Point-source plume attribution (GHGSat) | Integrated mass enhancement (IME) method applied to GHGSat column maps with ERA5 wind co-location | Per-event emission-rate estimate with uncertainty range, plume origin coordinates, and facility attribution confidence score; PDF report |
| Escalation trigger and tasking recommendation | Rule-based threshold logic combining TROPOMI anomaly magnitude, persistence across passes, and wind-back-trajectory plausibility | Automated alert with recommended GHGSat tasking coordinates and priority tier; JSON feed or email notification |
| Cloud-climatology retrieval-success forecast | ERA5 cloud-fraction climatology over corridor bounding box, stratified by month and time of day | Site-specific retrieval-probability calendar used to set monitoring cadence expectations; tabular report |
| Multi-year trend and persistence analysis | Time-series decomposition of TROPOMI XCH4 over corridor segments; Mann-Kendall trend test for persistent elevation | Annual trend report with segment ranking by emission persistence; suitable for regulatory disclosure or ESG reporting |
| EMIT hyperspectral confirmation for large events | Matched-filter retrieval on EMIT Level 1B radiance; cross-validation against TROPOMI and GHGSat detections | Event-specific confirmation note with plume extent map and co-emitter flags; appended to incident report |
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.