Permafrost emissions are on track to overtake India’s human emissions
Arctic Amplification, driving temperature increases up to 4x faster than the rest of the world, is melting Permafrost at alarming rates that will soon see it’s emissions eclipse even large countries.

What is permafrost and why is it melting?
Permafrost is basically ground that remains frozen for at least two years. It contains a mix of soil, rock, sand and importantly dead organic matter such as plant, animal and microbe remains. The mix is bound together by ice which can keep the organic matter stable for thousands of years. The carbon is essentially locked into the ground until it melts.
Most of the world’s stock of permafrost is located in the Arctic. In fact 22% of the northern hemisphere land mass is covered in permafrost. Here there is more than twice the amount of locked away carbon as there currently is in the atmosphere. 1,500Gt distributed between the frozen land and under the sea in the shallow continental shelve of Siberia.1
As Permafrost melts, microbes are mobilised which digest the organic matter, releasing carbon dioxide and methane into the atmosphere. Dry permafrost tends to release carbon dioxide (CO2), but it also produces smaller yet significant levels of methane (CH4) in waterlogged or submerged regions. CH4 is 84 times more potent as a greenhouse gas than CO2 over a 20 year period, so this is a big deal.
Due to a phenomena called Arctic Amplification, in which the reduction in sea ice and snow cover reduces the amount of solar radiation reflected back into space, the Arctic region is warming 3-4 times faster than the global average. Some regions are warming even faster, up to 7 times faster.
There are many mechanisms for permafrost melt. The most common is redial thaw in which the surface warms enough to melt the ice gradually, extending deeper and deeper below the surface. The second is through melt pools and is more abrupt. As water pools on the surface it penetrates deeper, melting the ice from below as the ground subsides. These lakes grow by melting wider and deeper below the surface accelerating the melt rate and the emissions. The waterlogged melting increases the proportion of methane released. Fires can also contribute to abrupt thaw by burning away the top insulating layer that protects the ice below from warming. More and more tundra wildfires are occurring as the region warms.
River bank collapse and coastal erosion also cause abrupt thawing. Monitored areas are seeing very rapid acceleration of coastal erosion rates. On average the Arctic coastline is retreating by 1m every year. Given that the coastline is 45,390km long (28,200 miles) that equates to a huge volume of soil erosion, much of it permafrost.
Permafrost emissions
As described above, when the ice that binds the organic matter in the soil melts, microbes get to work decomposing it. As they do they respire a mix of CO2 and CH4 depending on conditions and the types of microbes involved.
Already today, permafrost thaw releases annual carbon emissions equal to those of a top 10 greenhouse gas emitter such as Japan (about 0.6Gt CO2-equivalent per year). This may only be equivalent to 1.5% of human emissions but with increased warming comes rapidly accelerating melt and emissions.
When global temperatures pass 1.5ºC, as they will do very soon, permafrost emissions will increase to 2.5Gt/year, around the same as India. At 2ºC this increases to 4Gt/year, roughly the same as the 38 OECD countries combined. At 3ºC, emissions rise to over 5Gt/year, heading towards 10Gy/year and would remain at that annual rate for centuries.
You’ll notice that the rate of emissions is not linear with temperature, but rises abruptly from 1.5ºC, becoming shallower at the higher temperatures. This is due to the geographic position of the permafrost. The area exposed to melting at 1.5ºC is huge, the areas subsequently exposed at 2ºC and 3ºC become smaller. So as temperatures rise, it’s lower volumes of remaining ice that become exposed and melts per degree and therefore lower emissions growth. See the small graph in the upper right of the figure below.

This is good news and bad news. The good news is that emissions become capped at very high warming levels, helping to reduce the risk of runaway warming. The bad news is that the emissions growth over the next couple of decades as the Earth warms towards 2ºC are at their highest, just at the time we need to be reducing them as much as possible.
Sub-sea permafrost is melting more slowly since it takes longer for the ocean waters to warm. Estimates state that current emissions from sub-sea permafrost melt are one order of magnitude less than land based permafrost.
Putting these projections in context, the next chart below compares permafrost emissions with the 2025 fossil fuel country emissions according to the 2025 Global Carbon Budget2. At 1.5ºC permafrost emissions will overtake India. At 2ºC they will overtake even the USA, and at 3ºC they will overtake global land use change.

A Permafrost tipping point
The Arctic permafrost abrupt melt is one of the recognised global tipping elements, with the latest estimates putting its central risk at 1.5ºC. However the system is highly complex due to different geography, hydrology and weather conditions. A more accurate analysis is to consider the system as containing a large set of localised tipping elements. When a melt pool starts to burrow deeper, this is a local tipping point which will melt the ground underneath and in a wider and wider area with no further surface warming. Likewise coastal and river bank erosion and melting will be locally constrained by soil type, water flow etc. leading to local tipping points starting abrupt erosion events.
When considered this way, a full permafrost tipping point is less helpful. It is more like passing a temperature threshold where more and more localised events are inevitable. This acceleration then creates more emissions and more fresh water runoff that sustain and reinforce continued melting in a self reinforcing feedback loop.3
Paleoclimate evidence of previous melting and emissions
In a paper published in September, Amelie Lindgren at al.4 presented evidence of how much permafrost melted as the Earth emerged from the last glaciation 12,000 years ago. They not only examined the extent of melt but also the carbon content of the remaining thawed soils. They found that of the 366Gt of carbon stored at the last glacial maximum which melted during the deglaciation, only 48Gt remains in the unfrozen ground. The released 318Gt of carbon makes 1,165Gt of CO2. Although this is equivalent to 30 years of our current fossil fuel emissions, remember this was released over a period of 8,000 years or more. Today’s warming rate, melting rate and emissions rates are much higher.
Another study by Anton Vaks et al.5 found evidence that the northern hemisphere was permafrost free during the late Miocene approximately 7.3 million years ago. Temperatures were approximately 4.5ºC above pre-industrial and CO2 was around 350ppm - significantly less than todays 425ppm.
There is therefore no doubt that the northern permafrost is capable of melting fully and releasing the vast majority of its stored carbon into the atmosphere unless temperatures are reduced, not just capped.
Impact on remaining carbon budgets
Committed permafrost melt is 108-235Gt CO2e. This is not included in remaining carbon budgets and has the effect of lowering them by 120Gt for 1.5ºC and 250Gt for 2ºC. Given that the latest Global Carbon Project remaining budget for 1.5ºC is just 170Gt, this blows it completely (if it wasn’t already). Even the budget for 2ºC at 1,055Gt takes a serious hit of almost 25%.
The effect get worse however, as the emissions keep coming, so need to be accounted for in global net-zero calculations. To achieve neutrality, Carbon Dioxide Removal (CDR) methods will be required to counter these emissions at whatever temperature we stabilise at for hundreds of years, just to stand still. So if we are exceptionally lucky, and manage to stabilise temperatures at 3ºC*, carbon capture of at least 5Gt per year will be required to balance continued permafrost emissions and prevent further warming towards 4ºC. This is on top of balancing ‘hard to mitigate’ sectors and any other natural carbon sink declines such as forest die back or ocean sink reductions, let alone bring temperatures back down through net-negative emission efforts.
*Obviously landing at 3ºC will not be lucky at all, it will be catastrophic, what I mean is that on our current trajectory and policy mix, we will be lucky to stop it at 3ºC.
Timing
Given that permafrost emissions are temperature dependant and temperature rise rate is related to our emissions, plus other feedbacks, how quickly could they rise?
The current warming rate projected linearly into the future passes 1.5ºC in 2026, passes 2ºC in 2037 and reaches 3ºC in 20606. At the same time emissions are expected to drop according to the IEA 2025 report7 for both stated policy and current policy projection for energy use and the trend towards transportation emissions drop as the world embraces electric vehicles and trucks.
Detailed trajectories are not available for whole economy emissions forecasts apart from an incomplete selection of Nationally Determined Contributions presented at COP30. However the IEA data for 2025 has both stated policy trajectories and current policy trajectories for major economies for fossil fuel emissions in energy production. The plot below shows how the permafrost emissions, based on linear temperature rise compares to stated policy energy emissions from countries. By the early 2040s permafrost emissions will be greater than any countries energy emissions.

Conclusions
Permafrost emissions of mainly carbon dioxide, but with significant quantities of methane are about to accelerate as the world passes through 1.5ºC of warming. These emissions will rise rapidly from todays low value of about 0.6GtCO2 a year today, to 2.5Gt a year by the end of this decade and go on to 4Gt or more per year by 2040.
These rates will see permafrost emissions overtake those of India with only the US and China emitting larger amounts by 2030. Where as those two countries will see lowering emissions going forward, permafrost emissions will continue to increase through mid century and never decline (on any relevant human timescale) until temperatures are lowered.
This has major implications for remaining carbon budgets since in order to stabilise temperatures, Carbon Dioxide Removal (CDR) will need to remove these emissions in addition to ‘hard to abate’ human emissions. This could realistically lead to a doubling of requirements for net-negative or CDR campaigns.
In a scenario whereby humanity decides to manage a temperature overshoot, deploying CDR to limit the peak temperature and return to 1.5ºC or lower, the peak temperature reached and the duration of overshoot will be negatively affected by permafrost melt.
Given the imminent acceleration of permafrost emissions as we pass through 1.5ºC the sooner measures are taken to decarbonise and slow the acceleration of warming the better. Having said that, as other natural feedbacks engage, such as albedo reduction, and air quality improvements unmask hidden warming potential, the chances of avoiding 2ºC and the acceleration of these emissions is very slim indeed.
I haven’t mentioned the other impacts of permafrost melt such as infrastructure damage but these are also very relevant for the coming years and decades, especially for indigenous communities.
Like Antarctica, permafrost is often dismissed as a local issue, far far away, but the implications are global and coming for all of us.
ICCI, 2025. State of the Cryosphere 2025: Ice Loss = Global Damage. International Cryosphere Climate Initiative (ICCI), Stockholm, Sweden. 52 pp.
Friedlingstein, P., et al. Global Carbon Budget 2025, Earth Syst. Sci. Data Discuss. [preprint], https://doi.org/10.5194/essd-2025-659, in review, 2025
T. M. Lenton, D.I. Armstrong McKay, S. Loriani, J.F. Abrams, S.J. Lade, J.F. Donges, M. Milkoreit, T. Powell, S.R. Smith, C. Zimm, J.E. Buxton, E. Bailey, L. Laybourn, A. Ghadiali, J.G. Dyke (eds), 2023, The Global Tipping Points Report 2023. University of Exeter, Exeter, UK.
Amelie Lindgren et al. ,Massive losses and gains of northern land carbon stocks since the Last Glacial Maximum.Sci. Adv.11,eadt6231(2025).DOI:10.1126/sciadv.adt6231
Vaks, A., Mason, A., Breitenbach, S.F.M. et al. Arctic speleothems reveal nearly permafrost-free Northern Hemisphere in the late Miocene. Nat Commun 16, 5483 (2025). https://doi.org/10.1038/s41467-025-60381-5
Stefan Rahmstorf, Grant Foster. Global Warming has Accelerated Significantly, 03 March 2025, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-6079807/v1]
IEA, World Energy Outlook 2025, https://www.iea.org/reports/world-energy-outlook-2025



Detailed informative article.
..meant OAE in one shot..