Global Methane Emitters Tracker (GMET)

Global Methane Emitters Tracker (GMET) is Global Energy Monitor's attribution layer between the sky and the supply chain: a global census of fossil fuel facilities that emit methane, with emission estimates pinned to each one and satellite-detected plumes attached where partners observed them. Version 3, released 12 December 2025, covers 18,023 methane-emitting assets across 166 countries - more than 5,000 coal mines alongside oil and gas extraction areas, gas pipelines and LNG terminals - linked to 3,474 analysed plume observations carrying roughly 93 million tonnes of associated methane. GEM positions it as the only public database attaching emission estimates to gas pipelines and LNG terminals worldwide.

The release keeps an unusually clean shape: one eight-sheet package fronted by a Data Dictionary, then separate sheets for Pipelines, LNG Terminals, Coal Mines (non-closed), Plumes, Oil and Gas Extraction Areas and Oil and Gas Reserves - about 7,400 extraction-area rows, roughly 3,475 plume rows and some 2,300 reserve rows. Ten country/region summary boards travel alongside, aggregating emissions for pipelines, extraction areas and coal mines, 2P-reserve emissions potential, and plume counts. It sits in the oil gas exploration production data hub as the emissions read on assets its sibling trackers catalog. Get a sample of this dataset and we will return the sheets and countries you name, shaped like the rows below.

One row per emitting thing - an asset on the ground, a plume in the air, a country on a summary board:

-- sheet: Oil and Gas Extraction Areas -------------------------
gem_goget_id  unit_name  status     operator  country_area  latitude  longitude
OG0000001     Matzen     operating  OMV       Austria       48.41667  16.71667

-- sheet: Plumes ------------------------------------------------
gem_plume_id  provider      instrument                   observation_date         emissions_kg_hr  infrastructure_type
CH4T1         CarbonMapper  Global Airborne Observatory  2020-11-09T18:47:16+00  37.54922973      wellpad

-- country summary boards ---------------------------------------
board                                                   country      figures
Gas Pipeline Methane Emissions by Country/Area          Afghanistan  operating 524 t; proposed/construction 4,101 t
Oil and Gas Methane Potential at 2P Reserves by Country Algeria      12 assets with 2P estimates; 36.12 Mt total
GMET Methane Plumes by Country/Area                     Algeria      imager NASA-JPL EMIT; 25 plumes reviewed; 1 attributed

Every figure above is quoted from the release's own tables, not paraphrased. The asset row is OMV's Matzen unit in Austria - operating, keyed OG0000001, plotted at 48.41667 N - the exact record an analyst would join to an operator universe. The plume row is CH4T1, read by Carbon Mapper's Global Airborne Observatory at 37.54922973 kg/hr over a wellpad, timezone-stamped to the minute of capture. The three boards show how the summaries roll up: Afghanistan's pipeline methane split between operating stock and proposed construction, Algeria's 2P reserves carrying 36.12 Mt of emissions potential across 12 assessed assets, and the attribution funnel in the open - 25 Algerian plumes reviewed against NASA-JPL EMIT imagery, one attached to a catalogued asset. Your sample arrives carrying the sheets and geographies you actually need rather than the full 166-country span.

The dictionary is the release's own vocabulary, and it runs on two spines joined by GEM identifiers: asset rows that say who operates what and how much the surrounding field emits, and plume rows that say what a satellite actually saw, when, and how confidently it was attributed. All fourteen verified entries:

• Geo: Global - 166 countries across 18,023 emitting assets: 5,000-plus coal mines, oil and gas extraction areas, gas pipelines and LNG terminals, plus plume observations wherever partner sensors flew
• Temporal: December 2025 release (Version 3); plume imagery spans roughly August 2022 to March 2024 with mid-2025 coal mine coverage; extraction-area records carry a February 2025 vintage and pipeline records December 2024
• Granularity: Asset-level - mine, extraction area, pipeline, terminal - and individual plume observation, rolling up to country/region summary boards

That granularity stack is the product. One release answers who operates the emitter (asset rows), what did the sensor measure (plume rows), and how does my country compare (summary boards) without a second integration. The honest caveat is printed on the numbers themselves: emission figures are modelled values derived from partner data - Climate TRACE field estimates, Carbon Mapper and OCI+ plume readings - not stack measurements, and each plume carries its own uncertainty range and attribution certainty. Treat the tracker as the most complete attribution layer available, not a regulatory monitoring report.

API, files, or your warehouse. Daily, weekly, or hourly.

Pick the asset classes - pipelines, LNG terminals, extraction areas, coal mines - pick the countries, pick the cadence, pick the landing zone. Because the sheets share GEM identifiers, an extract arrives pre-joined rather than as loose tables needing reconciliation: plume observations ride on the assets they were attributed to, and country boards reconcile to the rows beneath them. The sample comes first - name the scope and it lands shaped to it, dictionary attached, before any commitment. Builders wiring this into monitoring pipelines can see the pattern in developers builders use cases.

Ranked by how directly a row settles the day job:

1. ESG and emissions accounting teams - asset-level methane baselines for portfolios and supply chains, with named operators instead of national aggregates that blur responsibility
2. Investors and quant researchers - emissions exposure screens and transition-risk factors built on asset identifiers stable enough to join against financial data; investors quants use cases shows where factor inputs slot in
3. LNG and midstream analysts - emission estimates for pipelines and terminals, the two asset classes almost every other emissions source skips
4. Satellite and geospatial analysts - 3,474 attributed plumes with provider, instrument, timestamp and emission rate, ready to benchmark against independent detections
5. Journalists and academics - facility-level stories and peer-reviewed designs that need citable numbers attached to named assets rather than country totals
6. Data engineers - one keyed release replacing a patchwork of per-country disclosures; data scientists use cases covers the modeling side

For contrast inside the same industry: the Global Registry of Fossil Fuels scores reserves and their embedded carbon, while GMET scores what those assets emit today - potential versus performance, stacked.

Quants get uniform numeric columns - emission rates, tonnage estimates, decimal-degree coordinates - that survive feature assembly without unit strings to strip. ESG and market researchers get the named-operator layer that turns a corporate disclosure into a checkable claim, facility by facility; the market researchers use cases page maps the workflow. Developers get a single keyed schema across four asset families, so adding LNG terminals to an extraction-only pipeline is a filter change, not a migration. Across all of them the constant holds: one dictionary, two spines, ten roll-up boards - scoping a new country is a request, not a research project. Persona-by-persona detail lives on the oil gas exploration production data hub.

The neighbors divide by which half of the problem they solve. Global Oil and Gas Extraction Tracker (GOGET) is the ownership counterpart - GMET rows carry GOGET IDs precisely so emissions snap onto operators, status and location already cataloged there. Global Registry of Fossil Fuels quantifies the carbon embedded in reserves yet to be burned; GMET measures what is escaping now. JODI-Gas World Database supplies officially reported monthly gas flows at economy level - the head-to-head comparison scores attribution depth against official breadth. For geometry to overlay, OGIM v2.5.1 maps the infrastructure footprint the coordinates land on.

Three glossary notes sharpen the vocabulary before you request anything: what asset-level tracking commits a model to, why an emissions inventory model differs from a measurement campaign, and how a satellite footprint bounds what any sensor can attribute. The publishing organization's profile lives at Global Energy Monitor, and the best oil gas exploration production datasets ranking shows where this record lands in the slice. Start at Get a sample of this dataset.