Coal mine Methane Measurement Estimation Tool Methodology

Background

Using reported measurement-based methane data

GEM’s coal mine methane measurement estimation tool (CoMMET) produces mine-scale methane emissions estimates based on measurement-based methane data collected at individual mines. This mine-scale data is important for both scientific and management purposes. Evidence suggests that select mines bear disproportionate responsibility for the total emissions from this sector. Accurate mine-scale data can help identify these “super-emitting” facilities. Mine-scale data is also essential for connecting emissions to responsible  mine operators, owners, and financers, as well as for informing management decisions around mitigation technology.

CoMMET combines GEM’s current and historic coal production data with publicly available, directly measured methane data to estimate emissions at select coal mines across the globe. The CoMMET output is an additional data product which complements two other methane data products available in the Global Coal Mine Tracker: modeled methane emissions estimates based on the Model for Calculating Coal Mine Methane (MC2M) approach, and reported methane emissions. Because CoMMET is built solely on directly measured methane emissions data, it provides both a unique snapshot of methane emissions at specific facilities and also highlights the major measurement data gaps which characterize this sector.

CoMMET uses directly measured data because measurement-based estimates offer the most direct reflection of coal mine methane emissions at the facility scale. Estimates based on facility-scale measurements are ranked the highest quality by the IPCC, Global Fossil Exploitation Inventory, and the International Methane Emissions Observatory (IMEO) when compared to emissions factors determined from basin-level or national data. Direct measurement of emissions also shows the effects of abatement activities a facility may have done to lower their emissions. Estimates built on direct measurement can also be subject to bias based on instrumentation or the process used to annualize point-in-time measurements. However, they also represent a distinct improvement over those built on less granular emissions factors, and can more immediately reflect the management activities at a particular mine.

Despite the importance of publicly available, mine-scale methane emissions measurement data, very few coal mines around the world directly measure and publicly report their methane emissions data in a manner which enables annual estimation. The model specifications and outputs for CoMMET are enumerated in detail below. However as a general rule, CoMMET integrates only “directly measured methane data”; this is broadly defined as data which fall under any of the following categories:

• Company-reported measurements conducted under direct regulatory oversight (i.e. ventilation shaft and drainage station emissions data for underground mines in the United States, Poland, and Australia)
• Published, peer-reviewed scientific studies if they integrated either direct bottom-up or top-down measurement approaches, and reported the data publicly at the individual mine scale on an annual basis:
  • Bottom-up data were only considered if the author included measurements at both drainage stations and ventilation shafts, or indicated that drainage station measurements were not necessary. Data which could not be annualized due to missing mine-scale production data (i.e. emissions factors drawn from Chinese gassiness level data) were not included.
  • Satellite and plane flight data were only included if they reported their results at the mine scale. Increasing numbers of authors are using high-resolution and area mappers to produce estimates at the mine scale, and many of these use GEM location data to anchor their modeling. However, very few authors chose to publicly disclose either the name or GEM mine identifiers associated with these mines.
• Select mine-scale results are from the International Methane Emissions Observatory’s Steel Methane Programme (SMP) Coal Methane Database. Given CoMMET’s focus on measurement data, GEM only included the entries from the IMEO Coal Methane Database where IMEO used directly measured, mine-scale data as inputs. Because IMEO integrated satellite data which could be highly variable in terms of number of observations per mine, GEM analysts faced the question “How many mine-scale satellite plume observations should be enough to qualify the mine for CoMMET inclusion?” To address this, GEM filtered the IMEO results based on the provided “certainty threshold” such that all mines had emissions values with certainties which matched or surpassed those created through company-reported, regulatory monitoring data. This allowed GEM  to integrate data from mines where IMEO had combined sporadic mine-scale satellite observations with an underlying inventory based on regional measurements.

Enhancing the Global Coal Mine Tracker database

CoMMET’s results complement the existing estimates in GEM’s Global Coal Mine Tracker. Historically, GEM has used a model called MC2M which estimates potential coal mine methane emissions based on mine characteristics including depth, coal rank, mine geometry, and coal production. This model, based on peer-reviewed research, produces results which conform with evidence of underreporting in some regions, though there is mixed evidence it may overestimate emissions in other parts of the globe (see the direct discussion of GCMT data in Bai et al. 2024 and Han et al. 2024 in China). As of August 2026, emissions estimates from CoMMET are often lower than those created by MC2M.

CoMMET produces lower methane emission estimates than M2CM for 76% of non-closed mines. CoMMET estimates are typically around half of M2CM values, with a median ratio of 53% and an average ratio of 72% - the higher average driven by a subset of mines where CoMMET estimates exceed M2CM. While the methods agree reasonably well on average, CoMMET estimates for individual mines can range from 6% to 384% of the corresponding M2CM value.

We caution users that, while we believe  measurement-based emissions estimates are generally more accurate, monitored data are still missing from the vast majority of mines, and scientists believe that fossil fuel methane estimates in general are “under-constrained,” meaning that they could be responsible for a greater share of total methane emissions than previously understood. For this reason, we have retained both columns of data in the Global Coal Mine Tracker.

CoMMET’s multi-tiered estimation process

Reported coal mine methane emissions are integrated into GEM's GCMT database using a hierarchical tiered approach, with calculations from higher tiers taking priority over those from lower tiers.

• Tier 1 - Using IMEO SMP Coal Methane Database emissions factors
• Tier 2 - Using emissions factors from scientific studies
• Tier 3 - Using reported measured emissions
• Tier 4 - Using calculated emissions factors

Emissions from the IMEO’s SMP Coal Methane Database are given top priority because they integrate multiple data sources for each mine and provide a quantification of the overall uncertainty for each mine estimate. Data from peer-reviewed scientific literature are rated above reported values due to greater independence and higher reproducibility as compared to company-reported values. Reported emissions are rated higher than those calculated using emissions factors because they are a more direct reflection of activity at the mine during that particular year.

Associating mine emissions data with GCMT mines -  'crosswalks', or linkages

Mine data used in each tier needs to be associated with mines in GCMT. GEM mine identifiers are included in the SMP Coal Methane Database. For all other tiers, GEM creates linkages between reported mine facility names and GEM mine identifiers.

Emissions results integration with GCMT

Results are delivered through the following new columns which have been added to both the Non-closed mines and Closed mines worksheets of GEM's GCMT database spreadsheet.

• Measurement based emissions estimate: The calculated or reported CMM estimate
• Emissions Calculation Detail: An amalgam of the year and source from which the measurement was calculated or reported
• Measurements methodology: A brief description of the measurements methodology behind the emissions estimate

Tier 1 - Using IMEO SMP Coal Methane Database emissions factors

Tier 1 coal mine methane estimates are calculated using emissions factors from the IMEO's Steel Methane Programme's Coal Methane Database spreadsheet, acquired through the download link on IMEO's Eye on Methane map website.

Data source

According to the IMEO, the Coal Methane Database spreadsheet is scheduled to be updated once every six months. Each row of the Mine-level data worksheet is associated with one unique coal mine. The data includes GEM mine identifier linkages and other data brought over from an earlier version of GCMT. The Mine-level data 'IMEO methane intensity (Mg CH4/Gg Coal)' column, described in the data source's data dictionary as 'Methane emissions per unit of coal produced', is used as a methane emissions factor, and the Mine-level data 'IMEO methane emission uncertainty (percent)' column is described in the dictionary as a 'relative uncertainty metric for IMEO methane estimate'. At the time of writing:

• the data source contains surface, underground, and surface & underground mine types
• all mines were in an operating status when the data source was created. Note some of them are no longer operating as per GCMT May 2026. Coal grades across these mines include metallurgical coal and/or thermal coal
• the reporting years span 2022-2024
• some rows include references to science studies

Historical data

The data contains an emissions factor for each mine for one given year only, i.e., the data contains no historical information. CoMMET does not create a historical summary from this data.

How results are calculated

Mine-level data rows are first filtered by removing those with 'IMEO methane emission uncertainty (percent)' values above configured thresholds. These are thresholds at which mines had some form of mine-scale measurement data included, whether a satellite study, science study, or company-reported measurement. Thresholds on a country basis were chosen by first sorting the uncertainty values in descending order, and finding the first mine which includes only non-measurement based data as its inputs. For all other countries, a general threshold was determined using the columns ‘Data approach’ and ‘Data source category” from the Methane emissions data worksheet. The threshold above this mine was then chosen. In general, IMEO assigns an uncertainty value of 30% for all company-reported measurement data inputs. We chose thresholds below this because on occasion the inclusion of multiple types of inventories, even if they weren’t measurement based, resulted in uncertainty values below 30% for some mines. As of the time of writing, we used a general threshold of 25%, and 15% for Australia.

Methane estimates are calculated for remaining mines by multiplying GCMT production values by IMEO methane intensity values. No filtering is done based on GCMT Mine Type. The IMEO reporting year is not required to match the GCMT production year for this calculation, which is applied to mines in both the GCMT Non-closed mines and Closed mines worksheets. New CoMMET columns values are calculated as follows:

GCMT column name	Contents
Measurement based emissions estimate (kt/year)	The calculated coal mine methane emissions estimates
Emissions calculation detail	Information about emissions factor origin
Measurements methodology	A description of the measurements methodology provided by GEM

Tier 2 - Using emissions factors from scientific studies

Tier 2 coal mine methane estimates are calculated using emissions factors from specific science studies. Earlier-tier results are not overwritten, preserving their priority, and  no filtering is done based on the GCMT Mine Type.

Data source

Tier 2 uses an internal GEM dataset of science studies compiled by GEM researchers. Each row of the dataset contains data for one mine, and includes the GEM mine identifier linkage, a coal mine methane emissions factor in million tonnes of methane per tonne of coal (Mt CH4/Mt), a link to the study, the year of the study and methodology information for each study.

Ultimately, only data from Borchardt et al. 2025, Palmer et al. 2021 and Sadavarte et al. 2021 met the criteria for inclusion as described above, and these studies all covered large Australian mines in the Bowen Basin. All of these mines were already covered in Tier 1 as they were included in the IMEO SMP Coal Methane Database, except for Coppabella Coal Mine which was filtered out due to uncertainty. For this mine, the emission factor from Palmer et al. 2021 is used.  In summary, science studies provided input for just this one mine this release. For context, the scientific papers reviewed for the August 2026 release of CoMMET include:

• Twenty three satellite studies of coal mine methane reviewed in Remote Sensing by Chauhan and Raval (2025)
• Scientific studies listed in the IMEO’s Steel Methane Program database using direct measurement data:
  • Borchardt et al. 2023
  • Singh et al. 2022
• Bottom-up data from China organized by Zhang et al. (2025)
• Additional studies drawn from the references of the Global Methane Status Report including:
  • Swolkień et al., 2022
  • Fiehn et al., 2020
• Karacan et al., 2025

In addition to these, two recent studies providing mine-scale quantification using plane flight and satellite data were not included because they were published after the first iteration of results had gone through final CoMMET QA/QC:

• Penn et al. 2026. These authors quantified emissions for thirteen underground US coal mines based on remote sensing data from plane flights. These measurements matched closely with in situ measurement data from the US Greenhouse Gas Reporting Program included in this study.
• Harris et al. 2026. These authors used data from plane flights in the Bowen basin in Australia to quantify emissions rates for seventeen mines. The results from the eight underground mines showed good agreement with the measured bottom-up reported data (also used in CoMMET). The surface mines were shown to be on average 3.6 times reported values which were based on estimates of gas content, rather than direct measurement. The only mine publicly identified by name by the authors was Hail Creek, which had been studied by Borchardt et al., 2025. Because this study was conducted by the IMEO, it is likely that it will be integrated into a future version of their Coal Methane Database.

Historical data

The emissions factor for each mine row is for just one given year, i.e. contains no historical information and CoMMET does not create an historical summary from this data.

How results are calculated

Methane estimates are calculated for mines by multiplying the emission factor by GCMT production values. The study year does not need to match the GCMT production year for this calculation, which is applied to mines in both the GCMT Non-closed and Closed mines worksheets. No filtering is done based on GCMT Mine Type. New CoMMET columns values are calculated as follows:

GCMT column name	Contents
Measurement based emission estimate (kt/year)	The calculated coal mine methane emissions estimates
Emissions calculation detail	Information about emissions factor origin
Measurements methodology	Methodology information in from science studies data

Tier 3 and 4 preparation

In preparation for Tier 3 and Tier 4 processing, the following are generated for mines with GEM linkages across countries defined in Tier 4:

• GCMT database annual production history, based on data in Non-closed mines, Closed mines and Historic Production (2018-2025) worksheets
• a history table of annual emissions from the reported data
• a table of the most recent (i.e. 'latest') year emissions from the reported data
• a GCMT production and reported emissions table, created through an outer join on GEM mine identifier and year of production or emissions
• a table of calculated historical emissions factors, based on the production and reported emissions table, where both production and emissions data is available
• a table of most recent (i.e. 'latest') emissions factors

A final preparation step filters the history table of annual emissions and the latest emissions factors table to just mines of specific types (e.g. 'Underground'.)

Country-specific preparation notes

• Australia Safeguard Mechanism data
  • This data is filtered as follows:
    • Australian and New Zealand Standard Industrial Classification (ANZSIC) value is 'Coal mining (060)'
    • Available corresponding National Greenhouse and Energy Reporting (NGER) ‘Method of estimation’ is not only Method 1 and/or Method 2 and/or Method 3
  • GHG Methane figures are converted from tonnes of carbon dioxide equivalent (tCO2-e) to million tonnes of methane (Mt CH4) using a configurable global warming potential of 28

Tier 3 - Using reported measured emissions

Tier 3 estimates are not calculated, but instead directly use reported emissions. Earlier-tier results are not overwritten, preserving their priority.

Data sources

Data sources are specific to each Tier 4 country and are described in tier 4 below.

Historical data

Historical data for each Tier 4 country are described in Tier 4 below.

How results are calculated

If GCMT production amount or production year are not given for a linked mine with reported data, the most recent reported emissions are used (see Note). Otherwise, if the GCMT production year is given, and the latest reported emissions year is the same or newer than it, the most recent reported emissions are used (see Note). Otherwise, reported emissions are not used. Only mines in a configurable GCMT Mine Types set are processed, which at the time of writing contains only 'Underground' mines.

Note: The reported emissions are used only if their reported dates are recent enough, i.e. not greater than a 'lookback period', given in years. At the time of writing, the lookback period was set to five years.

New CoMMET columns values are calculated as follows:

GCMT column name	Content
Measurement based emissions estimate (kt/year)	The reported emissions
Emissions calculation detail	Information about the reporting source and the year of the reported emissions
Measurements methodology	Depending on the country, either reported methodology information, or a concise methodology description provided by GEM

Tier 4 - Using calculated emissions factors

Tier 4 coal mine methane estimates are calculated using emissions factors calculated from reported emissions and GCMT production amounts. GCMT production amounts are then multiplied by the calculated emissions factors, resulting in methane emissions figures. Earlier-tier results are not overwritten, preserving their priority.

Data sources

As of the time of writing, these are as follows:

Country	Data source
Australia	Australian Government Clean Energy Regulator (CER) Safeguard Mechanism
Poland	Instrat's lignite and hard coal mining database - Kopalnie Mines
USA	US EPA Envirofacts API - subpart FF underground coal mines summary source table

Historical data

Historical data for tier 4 countries can be summarized as follows:

Country	Data source	Historical data
Australia	Safeguard Mechanism reports	Individual annual reports for 2023/2024 and 2024/2025
Poland	Instrat database	A single report containing emissions history ranging between 2010 and 2024
USA	EPA Envirofacts API	A single report containing emissions history ranging between 2011 and 2023

How results are calculated

GCMT production amounts are multiplied by the calculated emissions factors, resulting in methane emissions figures. New CoMMET columns values are calculated as follows:

GCMT column name	Content
Measurement based emissions estimate (kt/year)	The calculated emissions
Emissions calculation detail	Details about how the estimate was calculated
Measurements methodology	Depending on the country, either reported methodology information, or a concise methodology description provided by GEM

Results

Complete data availability

Results are included in the three new columns in the GCMT release as of August 2026. Future updates can be expected with the annual release of new GCMT data. We plan to explore potential methodological improvements to directly integrate satellite observations for these future updates, which will also expand our mine coverage.

August 2026 release notes

In total, only 134 non-closed and 14 closed mines were assigned values by the model, or less than5% of the mines in the Global Coal Mine Tracker. The following details why coverage is so limited.

There are 7136 total closed and non-closed mines in the Global Coal Mine Tracker (inclusion details are available here under “methodology”)

• For 81 of these we used emissions factors from the IMEO (167  of IMEO Coal Mine Database mines were removed via the thresholding process)
• 1 mine was updated using science study data (3 mines had emissions factors provided by science studies. Only one of these was not in IMEO’s database)
• 66 mines not included in IMEO’s database were linked with measurement based reporting data newer than 2021 (for 35 of these, GEM production data was more recent than reported emissions data, and so for these mines emissions factors were calculated)

This left 6,988 mines without publicly disclosed and citable recent measurement data.

Many of these mines do have high resolution satellite plumes associated with them (see GEM’s Global Methane Emissions Tracker for details). However, they have not yet been ingested into a model which can convert their emissions from hourly to annual estimates. Additionally, many provinces in China do disclose emissions factors on a mine basis, as discussed by Zhang et al., 2025. However, because China does not disclose production on a mine basis – only capacity – they were not included as CoMMET inputs.