Melt ponds of turquoise water spread across white sea ice in the Arctic Ocean, with two crew members in red walking between them.
Melt ponds on Arctic sea ice, NASA ICESCAPE mission, July 2011. Photo: NASA Goddard Space Flight Center, Public domain

57% of all the carbon dioxide from fossil fuels went up after 1990

Who put it there, the US or China, depends on how you count.

Added up since 1750, the United States has emitted the most CO₂. Today China emits two and a half times as much as America, and per person the order changes again.

GeoLens team · 14 Nov 2026 · data to 2025

Film cover: a timeline from 1750 to 2025 over a red relief map of Europe and North Africa in 2025. The 35 years after 1990 hold 57% of all fossil CO₂ since 1750, the 241 years before them 43%.Watch the film 1:05

Since 1750 the US has emitted more than the whole EU27

Fossil CO₂ here means carbon dioxide from coal, oil, gas, cement and gas flaring, counted in the country where it is emitted. International flights and shipping count only in the world total, and CO₂ from clearing forests is left out.

From 1750 to 2025 the world released 1,887 billion tonnes of it, and 1,082 billion of that after 1990.

Add up every year and the United States comes first, with 439 billion tonnes, 23% of the total. By the end of 2025 the EU27, counted as one bloc, was second with 303, and China third with 298.

On the Global Carbon Project's 2026 projection, China passes the EU27 during 2026.

Bar chart: fossil CO₂ added up from 1750 to 2025, in billion tonnes. United States 439, EU27 303, China 298, Russia 124, UK 81, Japan 71, India 69, and Saudi Arabia 19 in grey.
Fossil CO₂ added up from 1750 to 2025, in billion tonnes, inside today's borders. Land use not included.Source: Global Carbon Budget 2026 (Nov 2026) via Our World in Data. Still from the GeoLens film.
The numbers
Figure Value Year Source and release
World, fossil CO₂ added up 1,887 bn t 1750 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
World, of which after 1990 1,082 bn t (57%) 1991 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
Coal, oil and gas alone, no cement or flaring, share after 1990 56.5% 1991 to 2025 GeoLens calc on Global Carbon Budget 2026 via OWID (Nov 2026)
United States, added up 439 bn t (23%) 1750 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
EU27 as one bloc, added up 303 bn t 1750 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
China, added up 298 bn t 1750 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
Russia, UK, Japan, India, added up 124, 81, 71, 69 bn t 1750 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
Saudi Arabia, added up (grey in the chart) 19 bn t 1750 to 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
China passes the EU27, projected during 2026 2026 GeoLens sum on the Global Carbon Project 2026 projection (10 Nov 2026)

Source, unless a row names its own: Global Carbon Budget 2026 via OWID (Nov 2026)

In 2025 China emitted more than the US, India and the EU27 together

One year tells another story. In 2025 China emitted 12.4 billion tonnes, two and a half times the US figure of 5.0.

India came third, with 3.2 billion tonnes, ahead of the EU27 bloc at 2.4.

The same bar chart re-sorted for 2025 alone, in billion tonnes: China 12.4, United States 5.0, India 3.2, EU27 2.4, Russia 1.8, Japan 1.0, Saudi Arabia 0.6 in grey, UK 0.3.
The same eight, counting 2025 alone, in billion tonnes.Source: Global Carbon Budget 2026 (Nov 2026) via Our World in Data. Still from the GeoLens film.
The numbers
Figure Value Year Source and release
China 12.4 bn t 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
United States 5.0 bn t 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
China against the US 2.5 times (2.47) 2025 GeoLens calc on Global Carbon Budget 2026 via OWID (Nov 2026)
India 3.2 bn t 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
EU27 as one bloc 2.4 bn t 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
US, India and EU27 together 10.6 bn t 2025 GeoLens sum on Global Carbon Budget 2026 via OWID (Nov 2026)
Russia, Japan, Saudi Arabia, UK 1.8, 1.0, 0.6, 0.3 bn t 2025 Global Carbon Budget 2026 via OWID (Nov 2026)

Source, unless a row names its own: Global Carbon Budget 2026 via OWID (Nov 2026)

Per person, the US is ahead of China, and China is ahead of the EU27

China has four times as many people as the US. In 2025 the average American emitted 14.5 tonnes, a person in China 8.8 and a person in the EU27 5.4.

India, the most populous country, emitted 2.2 tonnes a head, under the world average of 4.7. Russia, at 12.3, sits between the US and China.

Peers are close. Australia (14.0) and Canada (13.6) are near the US; South Korea (11.2) and Japan (7.8) sit either side of China.

Inside the EU, Poland emits 7.4 tonnes a head, Germany 6.8 and France 3.8.

The same bar chart per person in 2025, in tonnes: Saudi Arabia 18.5 in grey, United States 14.5, Russia 12.3, China 8.8, Japan 7.8, EU27 5.4, UK 4.6, India 2.2, with the world average of 4.7 as a dashed rule.
Tonnes per person in 2025, counted where the CO₂ is emitted, not where the goods end up. The dashed rule is the world average.Source: Global Carbon Budget 2026 (Nov 2026); population UN WPP 2024; both via Our World in Data. Still from the GeoLens film.

Line chart of CO₂ emissions per person from 1950 to 2025 for the United States, Russia, Japan, China, the EU27, India and the world. The US runs highest, peaking at 21.5 tonnes in 1973 and falling to 14.5 in 2025. China rises from 0.1 to 8.8 and passes the EU27 in 2013, as the EU27 falls from 10.2 in 1979 to 5.4. India reaches 2.2 and the world 4.7.

Edit this chart
The numbers
Figure Value Year Source and release
Population, China and US 1,416 m and 347 m 2025 UN WPP 2024 via OWID
Population, India 1,464 m 2025 UN WPP 2024 via OWID
Per person: US, China, EU27 14.5, 8.8, 5.4 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Per person: India, Russia, World 2.2, 12.3, 4.7 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Per person: UK, Japan 4.6, 7.8 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Per person: Saudi Arabia (grey in the chart) 18.5 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Per person: Australia, Canada, South Korea 14.0, 13.6, 11.2 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Per person: Poland, Germany, France 7.4, 6.8, 3.8 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
China passes the EU27, per person 7.3 against 7.2 t 2013 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
US per person, at its peak 21.5 t 1973 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)

Source, unless a row names its own: Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)

In 2024 the world's yearly average passed 1.5 °C for the first time

The IPCC finds the human influence on the warming unequivocal.

The spiral below draws world temperature month by month since 1850. It measures each month against the 1850 to 1900 average, which the IPCC uses for the years before industry.

The Met Office puts the 2024 average at 1.53 °C above 1850 to 1900, Copernicus at 1.60 and the WMO at 1.55.

NASA (1.47) and NOAA (1.46) put it just under. The WMO, averaging six datasets, calls 2024 likely the first year past 1.5 °C.

Not every month was past the line: in the Met Office series, May, June, July and December 2024 stayed under it.

Nor does one year break the Paris Agreement's 1.5 °C goal: the IPCC measures warming as a 20-year average.

A spiral of world temperature, one turn a year since 1850, winding outward from blue to red. A ring marks 1.5 °C above the 1850 to 1900 average; a white mark tagged "year average" sits just outside it, and the centre reads 2024, +1.53. Foot: Met Office HadCRUT5, Copernicus 1.60, WMO 1.55.
World temperature each month, January 1850 to December 2024. The white mark is the 2024 average; the spiral is after Ed Hawkins.Source: Met Office Hadley Centre, HadCRUT 5.1.0.0 (Open Government Licence v3), moved to the 1850 to 1900 baseline by GeoLens. Still from the GeoLens film.
The numbers
Figure Value Year Source and release
Human influence on the warming unequivocal 2021 IPCC AR6 WG1, SPM (2021)
World against 1850 to 1900, Met Office +1.53 °C (1.45 to 1.62) 2024 Met Office, HadCRUT5 (2025)
World against 1850 to 1900, Copernicus +1.60 °C 2024 C3S Global Climate Highlights 2025 (Jan 2026)
World against 1850 to 1900, WMO +1.55 °C (+/- 0.13) 2024 WMO State of the Global Climate 2024 (Mar 2025)
World against 1850 to 1900, NASA +1.47 °C 2024 NASA GISS (Jan 2025)
World against 1850 to 1900, NOAA +1.46 °C 2024 NOAA NCEI, Global Climate Report 2024 (Jan 2025)
Months under 1.5 °C, Met Office series 4 of 12 (May, Jun, Jul, Dec) 2024 GeoLens calc, HadCRUT 5.1.0.0 (read 14 Nov 2026)
Months drawn in the spiral Jan 1850 to Dec 2024 1850 to 2024 Met Office HadCRUT 5.1.0.0 (read 14 Nov 2026)
Paris goal, measured as a 20-year average 1.5 °C goal 2015, measure 2021 Paris Agreement (2015); IPCC AR6 WG1 (2021)

Before 1800, CO₂ stayed under 300 ppm for 800,000 years. In 2025 it averaged 427

Air trapped in Antarctic ice holds 800,000 years of CO₂, rising and falling with the ice ages. Before 1800 it never passed 300 parts per million (ppm): the highest was 298.6, about 335,000 years ago.

In 1750, before industry, the air held 278 ppm. The 2025 average at Mauna Loa, Hawaii, 427 ppm, is 54% above that ice-core value.

The IPCC found the 2019 world average, 410 ppm, already the highest in at least 2 million years.

When the last ice age ended, CO₂ rose about 80 ppm in about 7,000 years. From 1985 to 2025 it rose 81 ppm, in 40 years.

Ice cores smooth fast changes, but the IPCC, allowing for that, finds CO₂ rose at least 10 times faster from 1900 to 2019 than at any other time in 800,000 years.

Line chart of CO₂ in the air over the last 800,000 years. The line rises and falls in slow teeth between about 174 and 299 ppm, always under a dashed line at 300 ppm, then shoots straight up at the right edge to 427 ppm. One bracket marks about +80 ppm in about 7,000 years at the end of the last ice age; a second marks +81 ppm from 1985 to 2025.
CO₂ in the air, in ppm, from 800,000 years ago to 2025. On this axis the years since 1800 are thinner than a pixel, so the rise is a wall.Source: Ice cores: Bereiter et al. 2015 (NOAA NCEI), via EPA and Our World in Data. Air: NOAA Global Monitoring Laboratory, Mauna Loa. Still from the GeoLens film.
The numbers
Figure Value Year Source and release
Length of the ice-core record about 800,000 years 803,719 BCE to 1958 EPA via Our World in Data (Nov 2026)
Highest CO₂ before 1800 298.6 ppm (+/- 3.0) about 335,000 years ago Bereiter et al. 2015 via EPA and OWID (Nov 2026)
CO₂ before industry 278.3 ppm (+/- 2.9) 1750 IPCC AR6 WG1, ch. 2 (2021)
CO₂ in the air, Mauna Loa 427.35 ppm 2025 NOAA Global Monitoring Laboratory (2026)
Mauna Loa 2025 against the 1750 ice-core value +54% 1750 to 2025 GeoLens calc on NOAA GML (2026) and IPCC AR6 (2021)
CO₂ in 2019, world average, highest in at least 2 million years (410 ppm) 2019 IPCC AR6 WG1, SPM A.2.1 (2021)
Rise at the end of the last ice age about +80 ppm (+76.5) in about 7,000 years about 18,000 to 11,000 years ago Bereiter et al. 2015 (NOAA NCEI)
Rise at Mauna Loa +81 ppm (346.35 to 427.35) 1985 to 2025 NOAA Global Monitoring Laboratory (2026)
CO₂ rate against any other time in 800,000 years at least 10 times faster 1900 to 2019 IPCC AR6 WG1, ch. 5 (2021)

Source, unless a row names its own: NOAA Global Monitoring Laboratory (2026)

In 2024 all 187 countries and territories in the data ran warmer than their 1951 to 1980 average

The film's map colours each country and territory by how far its year ran from its own 1951 to 1980 average, in Copernicus ERA5 data that start in 1940. In 1944, 59% were cooler than that average; by 2000, 9% were.

In 2024, 93% were at least 1 °C warmer. In 2025, 78% were.

The coolest in 2024, Uruguay, still ran 0.46 °C above its average.

Copernicus and the WMO name Europe the fastest-warming continent, at about twice the global average. One hot year can be luck.

Decades of red are not.

Relief map of Europe, North Africa and the Middle East in 2025, almost every country in shades of red, under a legend that runs from 2.5 °C cooler or more to 2.5 °C warmer or more than each country's own 1951 to 1980 average.
2025, each country against its own 1951 to 1980 average: blue cooler, red warmer.Source: Copernicus ERA5 via Our World in Data (Nov 2026), baseline moved by GeoLens. Contains modified Copernicus Climate Change Service information 2026. Still from the GeoLens film.

Heatmap of yearly temperature against each country's own 1991 to 2020 average from 1940 to 2025, one row each for the world and 16 countries, the world outlined, blue cooler and red warmer. The rows run mostly blue until the 1980s and turn red from the 2000s; from 2015 to 2025, 95% of the cells are red.

Edit this chart
The numbers
Figure Value Year Source and release
Countries and territories in the data 187 1940 to 2025 Copernicus ERA5 via OWID (Nov 2026)
Cooler than their 1951 to 1980 average 59% 1944 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
Cooler than it 9% 2000 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
1 °C or more above it 93% 2024 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
1 °C or more above it 78% 2025 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
Cooler than it none 2024 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
Coolest, Uruguay +0.46 °C 2024 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
Europe's warming against the global average about 2 times 1980s to 2025 C3S and WMO, European State of the Climate 2025 (Apr 2026)
Years warmer than their 1991 to 2020 average, world and 16 countries (the chart) 2% 1940 to 1979 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
The same, recent years 95% 2015 to 2025 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)

Source, unless a row names its own: GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)

What people ask about the count

Isn't China worse? In one year, yes: in 2025 China emitted 32% of the world's fossil CO₂.

Added up since 1750, the US has emitted more. Per person, an American still emits more than a person in China.

Is the US the worst per person? No.

Eleven places emit more per head, led by Qatar at 31.5 tonnes in 2025. Most of them are small oil and gas states.

Among countries of more than 10 million people, the US is fourth, after the United Arab Emirates, Saudi Arabia and Kazakhstan.

Isn't 1990 cherry-picked? Pick another start and it holds: 69% went up since 1980, and the halfway mark fell in 1996.

Count CO₂ from clearing forests too (that series starts in 1850), and the halfway mark moves back to 1987.

The chart below shows why: yearly emissions have climbed steeply since 1950.

Line chart of the world's CO₂ emissions from fossil fuels and industry, 1850 to 2025. The line is near zero in 1850 at 0.2 billion tonnes, climbs slowly to 5.9 billion in 1950, then steeply, and reaches 39.0 billion tonnes in 2025.

Edit this chart
The numbers
Figure Value Year Source and release
China's share of world fossil CO₂ 32% 2025 Global Carbon Budget 2026 via OWID (Nov 2026)
Places above the US per person 11 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Qatar, per person 31.5 t 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
US rank per person, countries over 10 million people 4th 2025 Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)
Share of all fossil CO₂ that went up since 1980 69% 1980 to 2025 GeoLens calc on Global Carbon Budget 2026 via OWID (Nov 2026)
Halfway mark of all fossil CO₂ during 1996 1750 to 2025 GeoLens calc on Global Carbon Budget 2026 via OWID (Nov 2026)
Halfway mark, land use included during 1987 1850 to 2025 GeoLens calc on Global Carbon Budget 2026 incl. land use via OWID (Nov 2026)
World, fossil CO₂ (the chart) 0.2, 5.9, 39.0 bn t 1850, 1950, 2025 Global Carbon Budget 2026 via OWID (Nov 2026)

Source, unless a row names its own: Global Carbon Budget 2026; population UN WPP 2024; via OWID (Nov 2026)

What people ask about the warming

Was 2024 just one El Niño year? One year could be.

But 2023 to 2025 together averaged 1.52 °C above 1850 to 1900, the first three-year average past 1.5 °C (Copernicus).

Why more than one baseline? The map compares each country with its own 1951 to 1980 average; the 1.5 °C compares the world with 1850 to 1900.

The chart below uses the 1991 to 2020 average that Copernicus publishes, 0.88 °C warmer than 1850 to 1900. Even on that harder test, only Iceland ran cooler in 2024.

Bar chart of 2024 temperature against each country's own 1991 to 2020 average for 20 countries, with the world as the reference row at +0.72 °C and a dashed rule at that value. Every bar is above zero, from Ukraine at +2.39 °C to India at +0.57 °C.

Edit this chart
The numbers
Figure Value Year Source and release
World, three-year average, against 1850 to 1900 +1.52 °C 2023 to 2025 C3S Global Climate Highlights 2025 (Jan 2026)
1991 to 2020 against 1850 to 1900 +0.88 °C ERA5, as of Jan 2026 C3S Global Climate Highlights 2025 (Jan 2026)
Only country cooler than its 1991 to 2020 average, Iceland -0.96 °C 2024 GeoLens calc, Copernicus ERA5 via OWID (Nov 2026)
World against its 1991 to 2020 average (the chart) +0.72 °C 2024 Copernicus ERA5 via OWID (Nov 2026)

Source, unless a row names its own: C3S Global Climate Highlights 2025 (Jan 2026)

What people ask about the ice cores

Is it just a natural cycle? The ice-age cycles are real but slow: Earth's orbit paces them over tens of thousands of years, and the next natural ice age is at least 10,000 years away.

The IPCC finds natural drivers, the Sun and volcanoes included, moved global temperature by -0.1 to +0.1 °C between 1850 to 1900 and 2010 to 2019.

Did you glue Mauna Loa onto the ice cores? The ice-core record runs to 1958 and the air measurements at Mauna Loa start in 1959.

At the join they agree: 315.3 ppm in the ice, 316.0 in the air. The film's ice line stops at 1799; in the ice, CO₂ first passes 300 ppm in 1909.

The numbers
Figure Value Year Source and release
Next natural ice age, at the earliest about 10,000 years away from today IPCC AR5 WG1 (2013); Ganopolski et al. 2016; Barker et al. 2025
Natural drivers, effect on global temperature -0.1 to +0.1 °C 1850-1900 to 2010-2019 IPCC AR6 WG1, SPM A.1.3 (2021)
CO₂ in the ice cores at the join 315.3 ppm 1958 EPA via Our World in Data (Nov 2026)
CO₂ in the air at the join, Mauna Loa 316.0 ppm 1959 NOAA Global Monitoring Laboratory (2026)
First year above 300 ppm in the ice 1909 (301.5 ppm) 1909 EPA via Our World in Data (Nov 2026)

Source, unless a row names its own: EPA via Our World in Data (Nov 2026)

Sources

  1. Fossil CO₂ by country and the world, 1750 to 2025, added up, yearly and per person: Global Carbon Budget 2026 (10 Nov 2026) via Our World in Data; population UN WPP 2024 via Our World in Data
  2. China passing the EU27 in 2026: GeoLens sum on the Global Carbon Project 2026 projection (10 Nov 2026)
  3. Shares after 1979 and 1990, and the halfway marks: GeoLens calc on the same data
  4. World temperature each month, 1850 to 2024: Met Office Hadley Centre, HadCRUT 5.1.0.0 (read 14 Nov 2026), Open Government Licence v3; 2024 at 1.53 °C: Met Office (2025). Spiral after Ed Hawkins
  5. The world against 1850 to 1900 in 2023 to 2025: C3S, Global Climate Highlights 2025 (Jan 2026)
  6. WMO, State of the Global Climate 2024 (Mar 2025) and 2025 (Mar 2026)
  7. 2024 in NASA GISS (1.47 °C) and NOAA NCEI (1.46 °C) against 1850 to 1900: their January 2025 statements
  8. Europe the fastest-warming continent: C3S and WMO, European State of the Climate 2025 (Apr 2026)
  9. Temperature by country, 1940 to 2025: Copernicus Climate Change Service, ERA5, via Our World in Data (updated 11 Nov 2026), read 14 Nov 2026; rebased to each country's 1951 to 1980 average, shares of countries: GeoLens calc. Contains modified Copernicus Climate Change Service information 2026
  10. CO₂ in ice cores, 803,719 BCE to 1958: EPA via Our World in Data (Nov 2026, CC BY 4.0); Bereiter et al. 2015, Geophysical Research Letters
  11. CO₂ in the air, 1959 to 2025: NOAA Global Monitoring Laboratory, Mauna Loa (2026)
  12. Human influence, CO₂ before industry, CO₂ in 2 million years, the CO₂ rate, natural drivers, warming as a 20-year average: IPCC AR6 Working Group I (2021)
  13. Ice-age cycles and the next ice age: Barker et al. 2025, Science; Ganopolski et al. 2016, Nature; IPCC AR5 WG1, ch. 5 (2013)
  14. Relief in the film: NOAA, USGS

Chart the data