Permafrost—ground that remains frozen for at least two consecutive years—stores an estimated 1,400 to 1,600 gigatonnes of organic carbon, roughly twice the amount currently in the atmosphere. As Arctic temperatures rise at four times the global average rate, this frozen carbon is thawing and becoming available for microbial decomposition, potentially releasing CO₂ and methane at scales that could significantly amplify global warming through a positive feedback loop that climate models are only beginning to capture.
Freitas, Walter Anthony, and Lenz (2025) reveal a previously overlooked emission pathway: deep thermokarst lake sediments in the Arctic. Published in Nature Geoscience, their study demonstrates that as permafrost thaws beneath these lakes, ancient carbon stored in deep sediments—some tens of thousands of years old—is being mobilized and released as both CO₂ and methane. The emissions from deep sediments are substantial and have been systematically excluded from most permafrost carbon budgets, which focused on surface-layer thaw. The finding effectively increases the estimated permafrost carbon feedback because it adds a previously unaccounted emission source that is particularly methane-rich—and methane has roughly an order of magnitude the warming potential of CO₂ over a 20-year timeframe.
Song, Rousseau, and Song (2024) provide a comprehensive review of ecological processes and carbon feedback in permafrost wetlands. Their synthesis identifies a cascade of interconnected changes triggered by warming: permafrost thaw increases active-layer depth, alters hydrological patterns (converting some wetlands from waterlogged sinks to drained sources), shifts vegetation communities (with implications for soil carbon decomposition and greenhouse gas emissions), and changes microbial community composition in ways that favor methanogenesis. The review emphasizes that these processes interact nonlinearly—warming does not simply speed up existing carbon cycling but fundamentally reorganizes the ecosystem in ways that can shift wetlands from net carbon sinks to net sources.
Bao, Xu, and Jia (2025) add an important comparative dimension, examining climate-carbon feedbacks in both Arctic and alpine permafrost systems. Their experimental synthesis reveals a surprising tradeoff: warming increases greenhouse gas emissions from some permafrost types while increasing carbon uptake by vegetation in others, with the net balance depending on temperature magnitude, moisture availability, and ecosystem type. In alpine permafrost, warming weakened the GHG sink, increasing its global warming potential by 13%. In arctic permafrost, warming actually strengthened the GHG sink, decreasing its global warming potential by 10%, as enhanced vegetation growth partially offsets the emissions, creating a weaker net feedback. This heterogeneity means that global estimates of the permafrost carbon feedback based solely on Arctic data may not capture the full picture—but the Arctic contribution alone is sufficient to meaningfully accelerate warming.
The policy dimension is daunting. The permafrost carbon feedback is largely beyond human control once initiated—unlike fossil fuel emissions, which can be reduced through policy, permafrost emissions are driven by temperature thresholds that, once crossed, cannot be reversed on human timescales. This makes aggressive mitigation of controllable emissions all the more urgent: every fraction of a degree of warming avoided is a fraction of permafrost carbon that remains frozen.