Supporting materials

Downloads

Glossary

Emission factor (EF)

Emission Factor = Methane Emissions (t) / Coal Production (kt)
This metric represents the emissions efficiency of the coal industry in a specific year.

UNFCCC-derived data

Coal mine methane emissions calculated using the emission factors in years when national emission inventories have been reported to the UNFCCC.

Area Flux Mappers vs Point Source Imagers

Satellite observations rely on two complementary sensing methods.

Area flux mappers, such as the TROPOMI instrument, observe large areas but can detect only the highest emissions. TROPOMI provides daily global coverage for methane monitoring, offering a consistent picture of total emissions across large areas. Point source imagers, by contrast, provide very fine-resolution estimates of smaller emissions over a limited area.

Together, these satellites reveal both the overall scale of methane emissions (via area flux mappers) and the specific sources requiring monitoring (via point-source imagers), helping to link the broader issue to concrete mitigation actions.

Abandoned Mine Methane (AMM)

Emission changes can be complicated by AMM, which differs fundamentally from emissions at active mines. Once a mine closes, ventilation ceases and methane accumulates and migrates to the surface through old shafts, fractures and surrounding geology.

Ventilation Air Methane (VAM)

This is methane released from underground coal mines when large volumes of ambient air are pumped through the mine to keep the atmosphere safe for workers. The air contains only a very small fraction of methane – typically between 0.1 % and 1 % – but because flow rates can reach up to one million cubic metres per hour, VAM emissions generally make up the majority of emissions from underground coal mines.

Methodology

Global coal mine methane emissions

Ember creates a complete national time series of coal mine methane emissions from 1990 by gap-filling reported data. Emissions reported by Annex I and non-Annex I countries to the United Nations Framework Convention on Climate Change (UNFCCC) are merged into a single dataset and combined with historical coal production data from the United States Energy Information Administration (EIA). Where both production and reported emissions exist, a methane Emission Factor (EF) is calculated.

Missing intensity values are filled for each country using the last known valid intensity value for a country and carrying it forward through subsequent years. This handles gaps that occur after a period of reporting. After the forward fill, a backward fill is performed. This uses the next known intensity value to fill in any remaining gaps, which is crucial for countries that started reporting recently, leaving their early years blank. For each year, reported UNFCCC values are prioritised where available, with gap-filled values used otherwise and classified accordingly.

Future methane emissions are estimated by applying emissions intensity factors to coal production forecasts from the International Energy Agency (IEA). Ember produces three forecast series: one that applies each country’s final historical UNFCCC-based intensity to future production, one that applies a single intensity derived from 2024 IEA emissions and production, and one that applies an intensity derived from the latest Global Energy Monitor (GEM) emissions data and reported production.

This report’s national-level analysis excludes emissions from closed or abandoned mines. Reporting for this methane source covers only a subset of countries and lacks comprehensive data.

Comparison to independent estimates

GEM estimates methane emissions for individual mines across the globe using nuanced assumptions for coal extraction volumes, method, coal rank and depth. Their bottom-up method estimates active CMM emissions at 55.7 million tonnes in 2024.

However, GEM does not account for methane mitigation activity, meaning its estimates will be higher than actual emissions in countries where capture and utilisation is practised. For this reason, GEM figures in this report are only compared to UNFCCC-derived estimates for countries where there is little to no known CMM mitigation — specifically India, Indonesia and South Africa.

The IEA uses similar assumptions as GEM, with additional constraints provided by atmospheric inversions and satellite data. It estimates that emissions from active coal mines amounted to 35.7 million tonnes in 2024.

East et al. (2025) provide a further independent estimate using TROPOMI satellite observations inverted against UNFCCC prior estimates at up to 25 km grid resolution for 2023. Unlike purely bottom-up methods, this approach is directly constrained by atmospheric observations, though results in some regions — particularly Indonesia and Kazakhstan — are limited by sparse satellite coverage.

Afghanistan and Nigeria emission data

When constructing the global emission factors, we screened emissions per country for implausible outliers. Countries with implied emission factors greater than 100 kt emissions per Mt of coal produced or changing by more than an order of magnitude from one year to the next were flagged as inconsistent with the rest of the dataset and with independent estimates. Under this criterion, UNFCCC estimated emissions for Afghanistan prior to 2004 and Nigeria prior to 2017 were treated as outliers and excluded from the aggregation of global emission factors.

Getting the percentage of underground versus surface coal mines

When determining the percentage of coal production that is sourced underground for each country, we first look for this statistic in the National Communication documents submitted to the UNFCCC. If a value is not available in these reports, the methodology defaults to using the corresponding figure from the GEM dataset.

Calculation of the Ember confidence score

The confidence score ranges from 0 (low confidence in coal mine methane reporting) to 6 (high confidence in coal mine methane reporting) and is based on a combined assessment of three key categories:

1. Recency of reporting to the UNFCCC (United Nations Framework Convention on Climate Change)
2. Similarity to other independent estimates
3. Robustness of the methods used to estimate emissions (tier of estimation used when reporting to UNFCCC)

Each category is scored from 0-2, the highest number being the best. The confidence score is the sum total of the three sub-scores.

Methodology: Coal Mine Methane Data Tracker

Acquiring the number of satellite observations and their emission intensity

The number of satellite observations and emission data used in this analysis were based on satellite data collected via API calls from three main platforms (Kayrros, IMEO MARS and Carbon Mapper). The collected data was treated for duplicates and missing values.

Calculating the percentage of national coal production that has an attributable satellite plume

We estimate the proportion of coal mines with detectable satellite plumes by aggregating site-level data from GEM Global Coal Mine Tracker (GCMT) and Global Methane Emitters Tracker (GMET). The number of coal mines per country is first counted from the GCMT dataset, which provides comprehensive mine-level information. We then identify the number of unique mining infrastructures with at least one attributed plume using the GMET dataset. This approach provides an estimate of the prevalence of detectable emissions across coal mine operations on a national level globally.

The GMET includes 66 mines flagged under the “auto attribution for edge cases” category. These are mines located near a detected plume where adjacent mines were also present, making attribution to a specific mine unclear. These plumes are nonetheless likely of CMM origin and are included in our national-level counts.

227 of all coal mines are reasonably associated with a satellite methane plume. Plume data is up-to-date as of July/August 2025.

Calculation of coal mining profits

We estimate global annual coal mining profits by aggregating the net profit figures reported for the 131 most profitable coal‑mining companies in Bullfincher’s ranking, which lists each company’s market capitalisation and profit data. The total of these profit values is taken as a proxy for worldwide coal mining earnings, representing a lower‑bound estimate that captures the largest publicly listed operators.

Supplementary Materials

Mission extensions are filling critical methane blind spots

The Earth Surface Mineral Dust Source Investigation (EMIT) mission has proven highly effective in identifying methane super-emitters. This unexpected ability to spot large methane plumes demonstrates that targeted satellite observations can uncover significant emission sources that were previously invisible to routine monitoring.

Originally, EMIT was designed to study mineral dust and map global mineral dust sources between 2022 and 2024. However, its success in methane detection led to its extension and refocus for broader scientific applications. This highlights how adapting and expanding satellite missions can fill important blind spots in global methane tracking.

Acknowledgement

Contributors

Ember: Adomas Liepa, Dody Setiawan, Sougol Aghdasi, Rajasekhar Modadugu, Yu-Ting Chang, Nishant Bhardwaj, Rini Sucahyo, Izabela Urbańska, Hannah Broadben, Reynaldo Dizon and Ardhi Arsala Rahmani.

We thank our external reviewers: Małgorzata Kasprzak (IMEO), Saul Lerman-Sinkoff (GEM), Tiffany Means (GEM), Flora Champenois (GEM) and Felicia Ruiz (Clean Air Task Force).

Cover image

Aerial photograph of the Tagebau Garzweiler open‑pit lignite mine near Jüchen, Germany.

Credit: Überform / Unsplash