Global Emissions Trends

Last updated: 2026-05-20 · PlainEmissions

The long arc

Source: EU EDGAR v8.0, Joint Research Centre, European Commission (2024). CC BY 4.0.

Global greenhouse-gas emissions have grown roughly threefold since 1970, from about 28 GtCO2e/year to roughly 50-55 GtCO2e/year by the early 2020s, the precise figure depending on whether LULUCF is included and which source you read. The first two decades of that arc (1970-1990) were dominated by industrialization in already-developed economies. The next two decades (1990-2010) were dominated by China's rise and the emerging economies of South and Southeast Asia. The most recent decade (2010-2024) shows a more fragmented picture: declining emissions in most OECD countries, continued growth in India and Southeast Asia, and a Chinese plateau that may or may not have peaked.

Has China peaked?

China's CO2 emissions plateau has been one of the most-watched questions in climate accounting for several years. EDGAR figures show roughly flat Chinese CO2 emissions from 2013 to 2020, a brief decline during COVID-19, and a partial rebound thereafter. UNFCCC figures (Chinese submissions are less frequent than Annex I countries) tell a broadly similar story with some discrepancy on the absolute level. Whether 2024-2026 mark a definitive peak depends on coal demand, renewables build-out pace, and economic structure, none of which is fully visible yet in published data.

What is reasonably clear: Chinese coal-fired power emissions, which were rising at 5-7%/year through 2013, have been roughly flat for a decade. The growth in Chinese emissions since 2013 has been driven primarily by industrial process emissions (cement, steel) and not by power-sector coal. Whether the recent dip is structural or cyclical is the policy question that will shape global trajectory through 2030.

Sources: EU EDGAR v8.0 (US, China, India, UK, computed live); UNFCCC National Inventory Submissions (2024) for EU-27.

EU and UK: real declines

EU-27 production-based CO2 emissions are down roughly 30% from 1990 levels by 2023, driven primarily by power-sector decarbonization (coal-to-gas-to-renewables switching) and structural shift away from heavy industry. UK figures show even sharper declines, roughly 50% from 1990, also driven primarily by power-sector changes. Both regions' consumption-based emissions have declined less, reflecting imported manufactured goods from China and other emerging economies.

These declines are real (the underlying physical-fuel-consumption data is well-measured) and they refute the framing sometimes heard that "no major economy has decoupled emissions from GDP growth." Several have. The harder question is whether the decline rate is fast enough relative to the Paris Agreement's trajectory, and whether the recent decoupling is fully sustainable given electrification of transport and heating yet to come.

United States: structural shift

U.S. emissions grew from 1990 through a 2007 peak, then declined, ending 2023 only modestly below the 1990 level (the small net change shown in the chart above). Measured from the 2007 peak specifically, the decline is much larger, roughly 17%, driven primarily by the coal-to-gas switch in power generation (cheap shale gas displacing coal) and secondarily by demand-side improvement (vehicle fuel economy, building efficiency). Per-capita U.S. emissions remain among the highest in the world but have declined from a peak around 25 tCO2e/person to roughly 16-17 by 2023. The trajectory since 2007 is real, but the U.S. has made far less progress than the EU or UK when measured from the same 1990 starting line.

India: rising but trajectory uncertain

India's emissions have roughly tripled since 1990 and are expected to continue rising through 2030 as the country industrializes and urbanizes. Per-capita Indian emissions remain among the lowest in any major economy (roughly 2.5 tCO2e/person in 2023), and India's growth pattern through 2050 will be one of the most important determinants of global trajectory. Whether India's growth bends toward renewables-led electrification (similar to China's 2020s) or coal-led industrialization (similar to China's 2000s) is genuinely uncertain.

Methane: the satellite revolution

Recent years have seen a significant gap open between bottom-up methane inventories and top-down satellite observations. Climate TRACE, TROPOMI, and Carbon Mapper have repeatedly identified methane plumes from oil-and-gas operations and landfills that are not reflected in country inventories. The IEA estimates that global methane emissions from fossil-fuel operations are roughly 70% higher than countries currently report. As this satellite-derived signal matures, expect inventories to be revised upward over the next several years, making the apparent global emissions trajectory worse, not because emissions have risen but because the previously-unreported portion is finally being measured.

Source: Climate TRACE 2023 release; IEA Global Methane Tracker (2024). CC BY 4.0.

What the trend doesn't tell you

Trend analysis is most reliable for the production-based, LULUCF-excluded, total greenhouse-gas series at the country level for countries with mature inventories. It is less reliable for: LULUCF-included totals (where methodology revisions can swing the figure by 20%+); consumption-based figures (which require trade-flow modelling); methane figures pre-2018 (before satellite verification was operational); and sub-national breakdowns for any country except the EU and the OECD.

Reading the trend, hold the framing assumptions in mind. PlainEmissions's guide on reading emissions time series walks through the six framing choices that materially shape the story.