🌍 Climate
India Adds 141 Million Tonnes and the Fossil Decline Never Reaches the Total
Fossil CO2 falls 158 million tonnes. Land use rises 700 million. The total hits a record 43.3 billion, and a 150 billion tonne 1.5°C budget lasts 3.46 years.
500 million tonnes more.
Fossil CO2 is projected to fall 158 million tonnes in 2026, from 38.4 billion tonnes to 38.2, while fossil plus land use is projected to rise 500 million tonnes, from 42.8 billion to a record 43.3, and the fossil range on that decline still runs from a 1.5 percent fall to a 0.6 percent rise. The Global Carbon Project printed the fossil line in a briefing dated 10 November, the morning its preprint opened for discussion, the morning its Tsukuba office said the update went out at COP31, and the morning Our World in Data repeated the 0.4 percent drop. Oil is the only fuel whose range sits entirely below zero: down 1.9 percent, somewhere between 2.6 and 1.2.
Land use moves the other way. Carbon Brief, co-authored by Pierre Friedlingstein, puts land-use emissions at 4.4 billion tonnes in 2025 and 5.1 billion in 2026, and says that holding land use flat would have left the total near 42.6, which is just 38.2 plus 4.4, while 38.2 plus 5.1 is the printed 43.3. Divide 158 by 700 and the fossil dip offsets 22.6 percent of that land-use increase. Divide 500 by 158 and the net rise is 3.16 times the dip the fossil headline is carrying.
| Line | 2025 | 2026 projection | Change |
|---|---|---|---|
| Fossil, including the cement carbonation sink | 38.4 | 38.2 | minus 0.158 |
| Land use | 4.4 | 5.1 | plus 0.7 |
| Total | 42.8 | 43.3 | plus 0.5 |
India's row is almost the whole dip. Table 1 has India up 141 million tonnes, 4.4 percent, in a range from 1.7 to 7.1 that does not include a decline, coal is 126 of the 141, 141 divided by 158 replaces 89.2 percent of the global fossil decline, and 126 divided by 158 means Indian coal alone replaces 79.7 percent. Subtract 141 from 158 and the rest of the fossil accounts fall 299 million tonnes, 1.89 times the headline dip. The named rows other than India sum to minus 330, and minus 330 plus the 32 million tonne smaller carbonation sink is minus 298, one million from that 299. A decline on most of the map is still not a decline in the sum.
| Piece | Change | What it does to the dip |
|---|---|---|
| World fossil | minus 158 | the printed 0.4 percent |
| India, of which coal is plus 126 | plus 141 | puts back 89.2 percent |
| World total minus India | minus 299 | 1.89 times the headline dip |
| Coal, oil, gas, cement, before the sink adjustment | minus 188 | plus 90, minus 243, minus 27, minus 8 |
| Fossil low case, minus 1.5 percent, plus land use of 5.1 | 42.924 total | 0.124 above 2025 |
| Fossil high case, plus 0.6 percent, plus land use of 5.1 | 43.730 total | 0.930 above 2025 |
Coal, oil, gas and cement sum to minus 188 million tonnes and the country rows sum to minus 189, while the footnote says the cement carbonation sink decreased by 32 million tonnes, a smaller sink and therefore higher net emissions, so minus 188 plus 32 is minus 156 and minus 189 plus 32 is minus 157, one or two million short of the printed minus 158, rounding rather than a second sink.
No detection follows from that. The preprint dataset puts 2025 land-use emissions at 1.2 plus or minus 0.7 gigatonnes of carbon against a year of 11.7 gigatonnes of carbon, or 42.8 billion tonnes of CO2, so 0.7 times 42.8 divided by 11.7 is 2.56 billion tonnes, five times the 0.5 billion rise, using the methods in the 2025 budget. Low case: 38.4 times 0.985 plus 5.1 equals 42.924. High case: 38.4 times 1.006 plus 5.1 equals 43.730. Both clear 42.8, and the 2.56 band swallows both.
What actually cut the fossil number
Clean energy did not produce the dip. Le Quéré, Friedlingstein and coauthors have demand-linked drivers adding about 0.78 billion tonnes in 2026 and solar, wind and electric vehicles offsetting 0.54, which is 69 percent, leaving a rise of 0.24 that must still fall 0.398 to reach minus 0.158. They assign the residual to prices and supply limits after the Iran conflict, and Carbon Brief names the Hormuz closure. Inputs are Ember, the IEA EV outlook and the IMF's October outlook. The authors say this counterfactual, with code on Zenodo, does not demonstrate causation.
Oil's 243 million tonne drop is only partly the cars: avoided EV emissions rise from 0.19 billion tonnes to 0.24, and 0.05 divided by 0.243 is 20.6 percent, so about four-fifths is price and missing barrels. Solar and wind avoided 4.6 billion tonnes in 2026, against 4.1 and against 0.28 in 2010, about 95 percent of the 2025 combined total, consistent with the IEA World Energy Outlook 2025 Stated Policies case. Data centers in the IEA energy-and-AI note added about 0.03 billion tonnes: 70 terawatt-hours times 0.43 is 30.1 million tonnes, and 0.0301 divided by 1.39 is 2.2 percent.
Two budgets, one temperature
Divide 150 by 43.3 and the briefing's 1.5 degree budget lasts 3.464 years, which Carbon Brief dates from 1 January 2027. Spread 43.3 evenly across a 365.25-day year and that is 1,265.3 days: three calendar years reach 1 January 2030 in 1,096 days because 2028 has a leap day, and the leftover 169.3 days cover January through May (151 days) plus 18.3 days after 1 June: 19 June 2030. 220 divided by 43.3 is 5.08 years, their "about five" for a tenth of a degree, and 0.13 degrees from 1.37 to 1.50, times 2,200 billion tonnes per degree, is 286, which is 136 above the printed 150.
| Worksheet | Billion tonnes left | Years at 43.3 | Where it lands |
|---|---|---|---|
| Briefing, from 1 January 2027 | 150 | 3.46 | 19 June 2030 |
| Straight line, 0.13 degrees times 2,200 per degree | 286 | 6.60 | 136 billion above the briefing |
| Forster's 130, depleted by 2026 | 86.7 | 2.00 | 63.3 billion below the briefing |
| 1.7 degrees, as printed | 510 | 11.78 | briefing says 12 years |
| 2 degrees, as printed | 1,035 | 23.90 | briefing says 24 years |
Forster and coauthors, cited for the 1.37 degree figure, refuse the straight line: their worksheet uses the 2016-2025 decade mean of 1.24 degrees, takes off 0.10 degrees of future warming from gases other than CO2, applies 0.45 degrees per 1,000 billion tonnes, then removes emissions already out since mid-decade plus about 7 billion tonnes of Earth-system feedback. What they print is 130 billion tonnes from 1 January 2026, the same 50 percent figure on the Copernicus indicators page and in the Potsdam Institute's summary. Subtract this year's projected 43.3 and 86.7 remain on 1 January 2027, 63.3 billion tonnes and 1.46 years short of 150. Even subtracting only the fossil 38.2 leaves 91.8. A smaller cement-carbonation sink, which Carbon Brief says raised historical fossil estimates about 0.3 billion tonnes a year versus last year's budget, would shrink a rolled-forward number rather than enlarge it. The briefing does not show the term that moved.
Novel removal stays a rounding error: about 0.002 billion tonnes from vegetation and 0.00001 from everything else, so 0.00201 divided by 38.4, times 525,960 minutes, is 27.5 minutes of 2025 fossil emissions, the briefing's "less than 30 minutes," and one day of those emissions, 38.4 divided by 365.25, is 0.105 billion tonnes, about 52 times larger.
What You Can Do
If a headline says emissions fell, ask which stack and which bound. Only the fossil central estimate is a decline. A peak would need the fossil upper bound below zero and a land-use change no larger than the fossil drop, and 2026 fails both on the printed figures. Even the accelerated case offsets 2.04 of a 2.59 billion tonne demand increase from 2025 to 2028, 79 percent, and Carbon Brief still has 2028 above 2025 once the shock is left out.
Until the five terms behind the 150 are printed, spend Forster's worksheet, not the briefing sentence: 130 from 1 January 2026, minus this year's 43.3, is 86.7 when 2027 starts. Watch the next indicators update and the version open for comment through 17 December. Worth waiting on is a full year whose fossil upper bound is negative and whose land-use line no longer exceeds the fossil drop. Novel removal at 27.5 minutes is not that year.
Limitations
This arithmetic does not prove that 2026 emissions rose, that India caused the path, or that either budget is a misprint. Both 2026 lines are projections, clean-energy months mostly end in September, and the land-use line has the lower confidence, so a 2.56 billion tonne band that swallows a 0.5 billion rise is not a resolved increase. The carbonation footnote says the sink decreased and does not print the identity that brings the columns within 1 or 2 million tonnes of minus 158. Minus 0.398 inherits Carbon Brief's rounding of 0.78 and 0.54. Forster's 130 is not a tank: non-CO2 pathways, committed warming and a revision can move it, and the five terms are not in the briefing, so this piece cannot see which term moved. The 19 June date freezes emissions at 43.3, which the companion paper says breaks if the oil shock ends, and 427.7 is a tenth above 425.6 plus 2.2.
The Strongest Case Against This Analysis
Grant the optimistic reading at full strength. Fossil emissions come from energy statistics, land use is a bookkeeping term whose wiggle sits inside its error bar, oil's range is negative in every major region, 40 economies cut fossil CO2 while growing over 2016-2025 (26 percent of fossil emissions, against 22 countries the decade before), and solar and wind avoided 4.6 billion tonnes while only about 10 percent of cars are electric, 6 percent without ordinary hybrids. Adding a low-confidence land-use line to a preliminary fossil line and calling the sum a record confuses a central estimate with a measurement. The 63.3 billion tonne gap may be a revision nobody itemized.
That reading can be right about the noise and still leave the printed totals alone: land use held flat gives 42.6, a 0.2 dip whose fossil piece the authors attribute to the oil price, while the fossil upper bound and India's range are both increases, and a budget that will not roll forward from the temperature paper cited beside it is not yet a number to spend. Once the shock fades, the companion paper's own warning has 2028 above 2025.
The Bottom Line
Fossil CO2 is down 158 million tonnes on the central estimate, India puts 141 million back, and land use adds 700 million, so the total is up 500 million tonnes, at 43.3 billion. Divide 150 by that total and a flat-emissions 1.5 degree budget runs to the morning of 19 June 2030. The worksheet that shows its terms runs out 1.46 years sooner. A full year with the fossil upper bound below zero, and land use no longer larger than the fossil drop, is what would move the chart.