Widespread observations and studies confirm that the active layer (the seasonally thawed soil above permafrost) has generally deepened across northern permafrost regions since the early 2000s, consistent with climate-driven permafrost thaw.

Researchers affiliated with George Washington University, notably Dmitry A. Streletskiy (Department of Geography), have contributed extensively to this field through the Circumpolar Active Layer Monitoring (CALM) program, Global Terrestrial Network for Permafrost (GTN- P), long-term field measurements, and synthesis papers on thaw, ground temperatures, and related impacts such as thaw subsidence.

A machine-learning study using in situ measurements produced 1 km resolution ALT estimates for northern permafrost regions (2003– 2020). About 65% of the area showed deepening, with a mean trend of ~0.11 ± 0.35 cm/yr; roughly 80% of CALM evaluation sites also showed deepening. Trends varied spatially and were influenced by factors such as fire.

Long-term monitoring (including CALM and related networks) indicates predominantly positive ALT trends since the 1990s and 2000s across much of the Arctic and sub-Arctic, with greater increases often in discontinuous or warmer permafrost and regional hotspots (e.g., parts of western Siberia, Interior Alaska). Rates commonly range from millimeters to a few centimeters per year, though they can be higher locally; some continuous-permafrost or specific sites show smaller or temporarily muted changes.

Reports covering recent years (including data into 2023–2024) document continued ALT increases in many regions, with record or above-average values in places such as Interior Alaska, Mackenzie Valley, Greenland, Svalbard, and parts of Europe and West Siberia in recent warm years. Mountain permafrost (e.g., European Alps, Qinghai-Tibet Plateau) has also shown substantial thickening at numerous sites.

Broader Northern Hemisphere observational syntheses report average ALT increases (with large spatial variability by permafrost type, latitude, and local conditions) and ongoing ground warming.

The active layer deepens as summer thaw penetrates farther into previously frozen ground, driven primarily by rising air and ground temperatures (Arctic amplification). This process can release previously frozen organic carbon (contributing to greenhouse-gas feedbacks), alter hydrology and ecosystems, increase ground instability/subsidence (especially where ground ice is abundant), and affect infrastructure. Rates and impacts are highly variable depending on ice content, soil properties, vegetation, snow cover, topography, and disturbances such as wildfire.

Not every location shows uniform deepening every year, some sites exhibit temporary stabilization, smaller trends, or (rarely) slight decreases linked to local climate or snow variations but the overall multi-decadal pattern from monitoring networks and modeling is one of progressive thaw and ALT increase across large parts of the permafrost domain.

These conclusions draw from observational networks (CALM, GTN-P), remote sensing/reanalysis products, and peer-reviewed analyses spanning roughly 2000–2024 (with some records longer). Ongoing monitoring continues to track these changes.

_____________________________________________________________________________________

Permafrost carbon feedback (PCF) is a positive (amplifying) climate feedback in which warming thaws permafrost, enabling microbial decomposition of previously frozen organic matter and the release of greenhouse gases (primarily CO₂ and CH₄, with some N₂O). These gases increase atmospheric concentrations, causing further warming that drives additional thaw.

Northern Hemisphere permafrost regions store roughly 1,460– 1,600 PgC (petagrams of carbon; 1 PgC = 1 billion tonnes) in soils, about twice the amount currently in the atmosphere, with substantial additional stocks in deeper deposits (e.g., Yedoma, peatlands) and subsea permafrost. Much of this carbon accumulated over millennia under cold, frozen conditions that limited decomposition.

Core mechanisms

1. Gradual (top-down) thaw and active-layer deepening
Rising air and ground temperatures deepen the seasonally thawed active layer and raise the permafrost table. Organic carbon that was previously frozen becomes available to microbes during the thaw season (and eventually year-round if taliks form). Decomposition rates increase with temperature, moisture, and oxygen availability. This process is relatively slow and widespread, releasing carbon over decades to centuries.

2. Abrupt thaw
In ice-rich permafrost, thaw causes ground collapse (thermokarst), thaw slumps, thermo-erosion gullies, and lake formation or drainage. These processes can expose meters of carbon-rich material rapidly (over years to decades), create new wetlands or lakes that favor anaerobic conditions, and accelerate lateral carbon transport. Abrupt thaw can disproportionately contribute to emissions relative to the area affected and is often underrepresented in models.

3. Microbial processes and gas pathways

• Aerobic conditions (well-drained soils): Primarily CO₂ production.
• Anaerobic conditions (waterlogged soils, wetlands, lake sediments): Methanogenesis produces CH₄ (higher global warming potential over shorter timescales) alongside CO₂. Some CH₄ is oxidized to CO₂ before reaching the atmosphere; vegetation (e.g., sedges) can act as conduits that bypass oxidation.
  Moisture, nutrient availability (including nitrogen released by thaw), and organic matter quality strongly influence rates and the CO₂:CH₄ ratio.

4. Interacting biophysical and ecological processes

• Vegetation shifts (greening, shrubification, or forest expansion/loss) can increase carbon uptake via photosynthesis and litter inputs, partially offsetting soil losses, while also altering snow insulation, albedo, and soil thermal regimes.
• Hydrology changes (wetting or drying) control oxygen availability and CH₄ emissions.
• Wildfires accelerate thaw by removing insulating vegetation and organic layers and can combust surface carbon directly.
• Lateral export of dissolved and particulate organic carbon to rivers, lakes, and the ocean.
• Canopy buffering in forests can help keep deeper carbon frozen; canopy loss can mobilize large additional stocks.

5. Hysteresis and irreversibility
Once thawed, soils warm and organic matter decomposes with thermal inertia. Recovery (re-freezing or re-accumulation of carbon) is slow, especially in organic-rich soils. Emissions can continue even under net-zero or negative emissions scenarios, and the CH₄:CO₂ ratio may shift.

Expert assessments and models generally project a gradual but prolonged release rather than an instantaneous “tipping point” runaway. Cumulative 21st- century emissions from gradual thaw are often estimated in the tens to low hundreds of PgC under high- warming scenarios (with large ranges), corresponding to additional warming on the order of 0.1– 0.3 °C or more by 2100 depending on assumptions; including abrupt thaw and wildfire increases the feedback strength.

Recent work incorporating deeper carbon pools suggests earlier shifts from sink to source in northern soils under high-emission pathways and larger net releases than some prior models. The region as a whole remains a weak CO₂ sink or near-neutral in some recent budgets when excluding certain processes, but becomes a clearer source when CH₄, N₂O, fires, and abrupt thaw are included; parts of the tundra already show source behavior.

Key uncertainties include the fraction of thawed carbon that is labile versus recalcitrant, the role of deep (>3 m) carbon, future hydrology and vegetation responses, the contribution of abrupt vs. gradual thaw, methane oxidation efficiency, and representation in Earth system models (many still incompletely capture these processes). Subsea and Antarctic permafrost add further uncertainty.

PCF reduces remaining carbon budgets for climate targets (e.g., 1.5 °C or 2 °C), increases the risk or accelerates timing of other tipping elements through added warming, and interacts with Arctic amplification, wildfires, and infrastructure/ecosystem impacts. It is not typically framed as a single global tipping point but as a progressive, multi-century feedback whose strength scales with the amount and duration of warming. Limiting peak warming reduces the total carbon eventually mobilized.

The dominant pathway is temperature-driven thaw enabling microbial greenhouse-gas production, modulated by hydrology, vegetation, disturbances, and carbon quality. The feedback is positive and long-lasting but occurs gradually, with abrupt processes adding important regional pulses. Continued monitoring, process studies, and improved model representation remain essential for refining projections.

_____________________________________________________________________________________

Arctic methane hydrates (also called methane clathrates or “fire ice”) are crystalline, ice-like structures in which methane molecules are trapped within cages of water molecules. They form and remain stable only under specific conditions of relatively low temperature and moderate-to-high pressure. In the Arctic, these deposits occur primarily in two settings: (1) within and beneath onshore and subsea permafrost, and (2) in marine sediments on continental slopes and shelves.

Global methane hydrate inventories are estimated in the hundreds to low thousands of petagrams of carbon (PgC), with the Arctic holding a notable share, especially on shallow continental shelves and slopes where colder bottom waters allow stability at shallower depths than in lower latitudes.

Estimates for Arctic marine and subglacial hydrates vary widely (e.g., tens to hundreds of PgC in polar regions overall), reflecting large uncertainties in distribution, saturation, and measurement. Subsea permafrost on the East Siberian Arctic Shelf (ESAS) and similar areas is associated with both hydrates and free gas trapped beneath or within frozen sediments.

Hydrate stability depends on the intersection of the local geothermal gradient, pressure (water depth or overburden), and temperature. In the Arctic:

• Shallow shelf and upper-slope deposits (often <300–500 m water depth, or associated with thin subsea permafrost) are most sensitive to bottom-water warming.
• Deeper deposits respond far more slowly because heat diffusion through sediments takes centuries to millennia.
• Subsea permafrost itself, formed during lower sea levels of the last glacial period and subsequently inundated, acts as a cap that can trap free gas and stabilize overlying or interbedded hydrates. Thaw of this permafrost (driven by geothermal heat from below and ocean warming from above) can eventually allow gas escape.

Modeling indicates that significant shrinkage of the methane hydrate stability zone (MHSZ) on Arctic shelves occurs mainly from the base upward over millennial timescales. Anthropogenic warming can accelerate permafrost and hydrate loss, but complete disappearance of the MHSZ on middle and inner shelves is not expected for thousands of years under most scenarios.

Methane seeps, bubble plumes, and elevated water-column concentrations are documented on the ESAS (Laptev and East Siberian seas), Beaufort Sea, and other Arctic margins. Sources include:

• Dissociation of hydrates.
• Free gas previously trapped by permafrost.
• Microbial production from organic matter in thawing sediments.
• Deeper thermogenic gas.

Not all seafloor methane reaches the atmosphere. Key sinks include anaerobic oxidation of methane (AOM) in sediments (often coupled to sulfate reduction), aerobic oxidation in the water column, and dissolution. Bubble transport can bypass some oxidation, especially in shallow water or during storms that enhance mixing and ventilation. Sea ice can temporarily trap or redirect bubbles.

Flux estimates from the ESAS vary widely (a few to ~17 Tg CH₄ yr⁻¹ in some studies), but atmospheric observations and more recent analyses generally support lower values and indicate that many early high estimates overstated the hydrate contribution. Current Arctic marine methane emissions are detectable but not yet a dominant global source compared with wetlands, agriculture, or fossil fuels.

Release of hydrate-derived methane could contribute to a positive climate feedback because CH₄ is a potent greenhouse gas (global warming potential roughly 28–34 times that of CO₂ over 100 years, higher over shorter horizons). However, the scientific consensus is that a catastrophic, near-term “methane bomb” is unlikely:

• Heat propagation into sediments is slow.
• Oxidation and dissolution strongly attenuate fluxes to the atmosphere.
• Only a small fraction of the global (and Arctic) hydrate inventory is vulnerable on century timescales.
• Paleoclimate records (e.g., ice-core methane isotopes during deglaciations) do not show large hydrate-driven spikes.

Models typically project limited additional warming from hydrate dissociation this century (often <0.1–0.5 °C in global mean, depending on assumptions), with larger effects possible over millennia if substantial dissociation occurs and methane is oxidized to CO₂ that remains in the atmosphere. Arctic upper-slope and thin-permafrost- associated deposits are the most relevant near- term sources.

Major uncertainties remain in hydrate inventory size and distribution, the relative contributions of hydrates vs. free gas vs. microbial methane, the efficiency of microbial and physical sinks, the rate of subsea permafrost thaw, and the response of seepage to continued ocean warming and sea- ice decline. Ongoing work uses seismic data, sediment cores, acoustic surveys of bubble plumes, water-column chemistry, atmospheric monitoring, and coupled sediment– ocean– atmosphere models.

Arctic methane hydrates represent a large, climate-sensitive carbon reservoir concentrated in permafrost and shallow marine settings. They are already experiencing some dissociation and seepage linked to long-term and ongoing warming, but physical and biogeochemical barriers limit rapid atmospheric release. Their contribution to near- term climate change is expected to be modest relative to anthropogenic emissions and terrestrial permafrost carbon feedback, though they remain an important process to monitor on multi-century to millennial timescales.

_____________________________________________________________________________________

Long- term monitoring of active layer thickness confirms global permafrost degradation

Long- term monitoring of active layer thickness (ALT) provides clear observational evidence of widespread permafrost degradation across the Northern Hemisphere (and to a lesser extent in Antarctica and mountain regions).

The active layer is the near- surface soil that freezes and thaws seasonally above the permafrost table. An increase in its maximum seasonal thickness (ALT) is a primary indicator of warming-driven thaw, as greater summer heat penetration deepens the thawed zone and raises the top of the remaining permafrost.

The primary sources are the Circumpolar Active Layer Monitoring (CALM) program and the broader Global Terrestrial Network for Permafrost (GTN- P), which track ALT via mechanical probing, thaw tubes, borehole temperatures, and related methods at hundreds of sites. Recent data compilations cover roughly 35 years (e.g., 1990–2024) across latitudinal and elevational gradients.

Northern permafrost regions (2003– 2020 and longer records): Machine-learning and observational analyses show deepening ALT trends at ~65% of the area, with a regional mean of about 0.11 cm yr⁻¹ (highly variable spatially; interquartile range includes both modest increases and some decreases). Site-level CALM evaluations indicate deepening at ~80% of locations. Greater rates occur in discontinuous/warm permafrost and certain Siberian or Interior Alaska sites (sometimes 1–2 cm yr⁻¹ or more).

Broader Northern Hemisphere: Observational syntheses report average ALT increases (e.g., ~3.6 cm yr⁻¹ in one multi-site compilation, though rates vary widely by region and method). Mean ALT values are larger in transitional/sporadic permafrost and smaller in continuous high-latitude zones, but the overall trend is thickening.

Recent years (into 2023– 2024): State-of-the-climate reports document continued ALT increases in many Arctic regions (more pronounced in discontinuous permafrost). Record or above-average values occurred in places such as Interior Alaska, the Mackenzie Valley, Greenland, and Svalbard. European mountain sites (e.g., Swiss Alps) show multi-meter thickening over decades, with some locations no longer fully refreezing in winter. Qinghai- Tibet Plateau sites also exhibit clear multi-decadal increases.

Spatial and temporal variability: Trends are not uniform. Some continuous- permafrost or specific East Siberian sites show smaller changes or short-term decreases linked to local climate, snow, or vegetation. Antarctica’s ice- free areas (e.g., McMurdo Dry Valleys) display interannual variability with limited long-term thickening in some records. Fire disturbance can accelerate local deepening.

Complementary evidence includes rising mean annual ground temperatures (often 0.1– 1 °C decade⁻¹ or more, faster in colder permafrost), ground ice loss, thaw subsidence, and expansion of thermokarst features. These confirm degradation beyond simple ALT changes.

ALT deepening exposes previously frozen organic carbon and ice, contributing to carbon- climate feedback, altered hydrology, ecosystem shifts, infrastructure damage, and landscape instability. Because monitoring networks provide multi-decadal, in situ records, they form a robust basis for detecting the signal of climate-driven change amid natural variability and for validating models and remote-sensing products (e.g., InSAR-derived thaw or satellite- based ALT estimates).

the long-term observational record from CALM, GTN- P, and related programs consistently shows that active layers are deepening over large parts of the global permafrost domain, confirming ongoing degradation driven by Arctic and high-mountain warming. Continued and expanded monitoring remains essential for tracking rates and regional differences.

Published:  Communications Earth & Environment

DOI: 10.1038/s43247-026-03824-1

Provided: George Washington University

Authors: Dmitry A. Streletskiy, 
Kelsey E. Nyland, 
Nikolay I. Shiklomanov, 
Emma Haggerty, 
Michael L. Mann, 
Vasily Tolmanov, 
Frederick E. Nelson, 
Anna E. Klene, 
Felix Pretis, 
Lin Chen, 
Hanne Christiansen, 
Miguel Angel de Pablo, 
Dmitry Fyodorov-Davydov, 
Mauro Guglielmin, 
Filip Hrbáček, 
Ketil Isaksen, 
Margareta Johansson, 
Marina Leibman, 
Dario Trombotto Liaudat, 
Alexey Maslakov, 
Mikhail Mastepanov, 
Jeannette Noetzli, 
Vladimir E. Romanovsky, 
Anarmaa Sharkhuu, 
…
Anna Irrgang 

Abstract

Active layer increase may contribute to greenhouse gas emissions, ecosystem change, and increased hazards. Here, we show the results of field measurements of active-layer thickness from 156 monitoring sites in Arctic, Antarctic and mountain permafrost regions. Active layer thickness increased significantly at 55% and 38% of sites in the Arctic and Antarctic regions, at more than 90% European mountain and high elevation Asian sites, and at sites in South America, demonstrating worldwide permafrost degradation during the first quarter of this century (2000–2024). The largest changes were observed in mountain regions where active layer doubled at several sites. The smallest changes were observed at sites with thick surficial organic horizons and high ground ice content. In continuous permafrost, sites generally exhibited smaller increases compared to sites in discontinuous permafrost. Regression analysis indicates Arctic active layer changes are attributable to increasing thawing degree-days, followed by increases in total rainfall. Other permafrost regions require more sites and longer time-series to draw conclusions regarding active layer change attribution.