What Counts as a Greenhouse Gas Emission?

A greenhouse gas emission is a release of a heat-trapping gas into the atmosphere. Carbon dioxide from burning fuel counts. So do methane leaking from energy systems or produced by agriculture, nitrous oxide associated with agricultural soils, and fluorinated gases escaping from refrigeration or electrical equipment. But an emissions total is not self-explanatory: its meaning depends on which gases, sources, geographic area, time period and accounting method it includes.

TL;DR

• An emission is a flow released during a period, such as tonnes of methane per year. It is not the same as the concentration already present in the atmosphere.
• Standard greenhouse gas inventories cover carbon dioxide, methane, nitrous oxide and fluorinated gases. Water vapor affects climate but is generally treated as a feedback rather than a persistent human forcing in emissions accounting.
• CO2-equivalent, or CO2e, converts different gases into a common accounting unit. The result depends on the selected metric and time horizon.
• A credible emissions number states its gases, units, time period, geographic or organizational boundary, treatment of land use, and whether the figure is gross or net.
• Emissions inventories are estimates built from activity data, emission factors, models and sometimes direct measurements. Atmospheric monitoring answers a related but different question.

Emission, concentration and warming are different quantities

Three concepts are often collapsed into one: emissions, atmospheric concentrations and climate effects. Keeping them separate resolves much of the confusion around greenhouse gas numbers.

Emissions are flows. A power plant may release a quantity of carbon dioxide during a year, or a gas system may release methane during an hour. An inventory adds such flows across a specified period and boundary. Concentration is the stock already present in the atmosphere, usually reported for carbon dioxide in parts per million and for some other gases in parts per billion.

Radiative forcing describes a change in Earth’s energy balance, expressed in watts per square metre. NOAA’s Annual Greenhouse Gas Index uses observed abundances of long-lived greenhouse gases to track their effective radiative forcing relative to pre-industrial conditions. It is not an emissions inventory and does not, by itself, forecast temperature.

The sequence is therefore: human activities release gases; emissions alter atmospheric concentrations; changed concentrations affect Earth’s energy balance; and the climate system responds. The links are physical, but the quantities are not interchangeable. A chart of annual emissions cannot be read as if it were a chart of atmospheric concentration or global temperature.

Which gases normally count?

Standard emissions reporting principally covers carbon dioxide, methane, nitrous oxide and fluorinated gases. Their sources, atmospheric behavior and accounting treatment differ, which is why the gas composition of a total matters.

Gas or group	Common human-related sources	What matters when reading a total
Carbon dioxide, CO2	Fossil-fuel use, cement and other industry, and land-use change	Long-lived CO2 accumulates, so cumulative net emissions are central to its warming effect.
Methane, CH4	Energy systems, livestock and other agriculture, and waste	Methane is powerful per unit mass but remains in the atmosphere for about a decade; the selected equivalence metric strongly affects its CO2e value.
Nitrous oxide, N2O	Agricultural soils and industrial activities	It persists for more than a century and has a high warming impact per unit mass.
Fluorinated gases	Refrigeration, industrial processes and electrical equipment	Different compounds have different lifetimes and warming effects, so inventories generally convert them to CO2e.

Carbon dioxide usually dominates global totals, but that does not make the other gases optional. The EDGAR database estimates that global emissions in 2024, excluding land use, land-use change and forestry, were 53.2 billion tonnes of CO2e. Its estimated composition was 74.5% fossil CO2, 17.9% methane, 4.8% nitrous oxide and 2.8% fluorinated gases. Those percentages describe that dataset and boundary, not a universal mixture for every country or sector. EDGAR documents the estimate and its coverage.

Why water vapor is usually handled differently

Water vapor is a greenhouse gas in the physical sense: it absorbs infrared radiation and contributes to the greenhouse effect. In the lower atmosphere, however, its abundance is governed mainly by temperature, and water rapidly condenses and precipitates. The IPCC therefore describes atmospheric water vapor primarily as a feedback that amplifies warming initiated by persistent forcings such as carbon dioxide, methane and nitrous oxide.

That is why a visible plume of condensed water from a cooling tower is not normally treated like a CO2 emission in a national greenhouse gas inventory. The relevant question is not merely whether a substance can absorb infrared radiation. It is whether the release produces the kind of persistent atmospheric change the inventory is designed to track.

What CO2-equivalent means—and what it hides

A multi-gas inventory needs a way to add unlike gases. CO2-equivalent, written CO2e, does this by applying an equivalence metric to each gas. The common choice is 100-year global warming potential, or GWP100, which compares the integrated warming influence of a pulse of a gas with the influence of the same mass of CO2 over 100 years.

Using values summarized by the US Environmental Protection Agency, one tonne of methane corresponds to roughly 27 to 30 tonnes of CO2e under GWP100, while one tonne of nitrous oxide corresponds to 273 tonnes of CO2e. The range for methane reflects distinctions in how sources and climate interactions are treated.

CO2e is an accounting aggregate, not a gas that an instrument samples from the air. It is useful for inventories, targets and comparisons, but it does not make the gases physically interchangeable. Methane lasts about a decade in the atmosphere, whereas nitrous oxide persists for more than a century; CO2 has a substantial cumulative effect that cannot be represented by one simple atmospheric lifetime. A single CO2e total compresses those different time profiles into one number.

Whenever CO2e appears, look for the metric and time horizon. A total based on GWP100 is not automatically comparable with one using another time horizon or equivalence method. This also matters for net-zero greenhouse gas claims, which need to identify the aggregation metric being used.

How emissions are estimated

Most emissions are not measured continuously at every exhaust pipe, field, landfill or leaking component. Inventories generally combine activity data with emission factors. Fuel use can be multiplied by a factor representing emissions per unit of fuel; livestock numbers can be combined with factors or models appropriate to the animals and management system. Direct measurements may be used where they are available.

IPCC inventory guidance describes a tiered approach. Tier 1 uses default emission factors and broadly available activity data. Tiers 2 and 3 use more country-specific factors, detailed models or more sophisticated measurement systems. A higher tier can better represent local conditions, although the quality of any estimate still depends on its underlying data, assumptions and implementation.

Atmospheric monitoring takes another route. Instruments collect calibrated air measurements that reveal concentrations above monitoring sites and across networks. With meteorology and atmospheric models, observations can help test whether inventory estimates are consistent with what appears in the atmosphere. But a concentration measurement does not automatically identify every source responsible for it.

This is a useful example of converging evidence rather than competing definitions. Inventories work from activities toward emissions; atmospheric observations work from measured air back toward possible sources and sinks. Differences can reveal missing sources, uncertain emission factors, transport-model limitations or timing and geographic mismatches.

The boundary can change the headline number

Two reputable global totals can differ without either being erroneous. The first step is to compare their boundaries. Does the total include only CO2 or all major gases? Does it include land-use change? Is it expressed as tonnes of gas or tonnes of CO2e? Is the latest year an observation, an inventory estimate or a projection?

For example, EDGAR’s 2024 headline estimate covers multiple gases in CO2e but excludes land use, land-use change and forestry. By contrast, the 2025 Global Carbon Budget projected 38.1 billion tonnes of CO2 from fossil fuels and cement and 4.1 billion tonnes from land-use change. Those Global Carbon Budget figures are CO2-only estimates, with fossil and land-use emissions reported separately. Adding or comparing the headlines without reconciling their gases, years and land-use treatment would mix unlike quantities. The Global Carbon Budget explains its 2025 estimates and definitions.

The same boundary problem appears at smaller scales. A city total may count emissions occurring within city limits, while a consumption-based estimate may also assign emissions embodied in imported electricity, food or manufactured goods. An organization may report releases from operations separately from those in its supply chain. Before interpreting a trend, verify that the boundary has not changed.

Gross emissions, removals and avoided emissions

Gross emissions are releases before subtracting removals. Net emissions combine emissions and removals over a stated period and boundary. Net-zero CO2 means anthropogenic CO2 emissions and anthropogenic CO2 removals balance. It does not mean that no facility, vehicle or land-use activity releases CO2.

An avoided emission is different from a removal. Replacing a high-emitting activity with a lower-emitting one may prevent emissions relative to a defined baseline, but it does not extract an equivalent quantity already in the atmosphere. Carbon capture can also describe different outcomes: capturing fossil CO2 before release reduces an emissions stream, while removing atmospheric CO2 and storing it durably is a removal. These distinctions are central to evaluating what carbon capture can and cannot do.

Natural land and ocean sinks absorb a substantial share of human CO2 emissions, but they do not erase the remainder. The Global Carbon Budget estimates that during 2015–2024 the ocean absorbed about 29% and land about 21% of human CO2 emissions. Atmospheric CO2 continued to rise because total human emissions exceeded the combined uptake by these sinks.

Why cumulative CO2 matters

For CO2, annual emissions are important partly because they add to a cumulative total. The IPCC assesses a near-linear relationship between cumulative net CO2 emissions and global warming. Every 1,000 billion tonnes of cumulative CO2 emissions produces an estimated 0.45°C of warming, with a likely range of 0.27°C to 0.63°C. The IPCC summary for policymakers explains this relationship and its uncertainty.

This is why stable annual CO2 emissions are not the same as stable atmospheric CO2 or stable temperature. If net CO2 emissions remain positive, the cumulative total continues to grow. Stabilizing human-caused warming at a chosen level requires reaching at least net-zero anthropogenic CO2 emissions, alongside strong reductions in other greenhouse gases.

A checklist for reading any emissions claim

Before accepting, rejecting or comparing an emissions number, identify the following:

1. Quantity: Is the figure emissions, atmospheric concentration, radiative forcing or temperature change?
2. Gases: Is it CO2 only, or does it include methane, nitrous oxide and fluorinated gases?
3. Unit: Is it tonnes of a specific gas, tonnes of carbon, or tonnes of CO2e?
4. Metric: If it is CO2e, which equivalence metric and time horizon were used?
5. Time: Is the number annual, monthly, cumulative or projected?
6. Boundary: Which country, sector, organization or product system is included?
7. Land use: Are land-use emissions and removals included, excluded or reported separately?
8. Gross or net: Have removals been subtracted, and are avoided emissions being mislabeled as removals?
9. Method: Does the estimate use activity data, emission factors, direct measurements, models or a combination?
10. Comparison: Were both values calculated with consistent definitions and methods?

The same discipline used to evaluate a climate graph online applies here: read the unit, baseline, boundary and method before interpreting the shape or headline. A precise-looking number can still answer the wrong question if those definitions are hidden.

The practical takeaway

What counts as a greenhouse gas emission depends first on a physical event—a heat-trapping gas is released—and then on an accounting definition. Carbon dioxide, methane, nitrous oxide and fluorinated gases normally form the core of reported totals. But the headline becomes meaningful only when its scope is visible.

When you encounter an emissions claim, do not ask only whether the number is large or small. Ask what flowed, during which period, across what boundary, in which unit and under which metric. Then check whether the claim is actually about emissions rather than concentrations, forcing or warming. Those questions turn an abstract total into evidence that can be compared, tested and used.

References

1. Annual Greenhouse Gas Index (AGGI) – NOAA Global Monitoring Laboratory
2. Education and Outreach – Global Monitoring Laboratory
3. Global Greenhouse Gas Overview | US EPA
4. EDGAR – The Emissions Database for Global Atmospheric Research
5. Chapter 7: The Earth’s Energy Budget, Climate Feedbacks, and Climate Sensitivity | Climate Change 2021: The Physical Science Basis
6. Understanding Global Warming Potentials | US EPA
7. APPROVED  Summary for Policymakers  IPCC AR6 WG III
8. Chapter 1: Framing, Context and Methods | Climate Change 2021: The Physical Science Basis
9. THE_2006_IPCC-final0507_km.indd
10. Global Carbon Budget | The Global Carbon Budget: FAQs 2025
11. IPCC AR6 Working Group 1: Summary for Policymakers | Climate Change 2021: The Physical Science Basis