Methane, a potent greenhouse gas, contributes significantly to radiative forcing. High-latitude wetlands in the Arctic tundra are a major source of methane, but uncertainties remain due to limited observational data and the complex interplay of factors influencing methane production, transport, and oxidation. Arctic warming, at least twice the global average, is expected to accelerate permafrost thaw and increase methane emissions through the decomposition of previously frozen organic matter, extending the thawing season and deepening the active layer. This positive feedback loop could exacerbate climate change. While projections anticipate increased methane emissions from thawing permafrost, observational evidence has been lacking. This study addresses this gap by analyzing a long-term (16-year) eddy covariance methane flux dataset from the Lena River Delta, providing the first trend analysis of directly observed methane fluxes in an Arctic ecosystem.

Current estimates of the global methane budget show a significant contribution from high-latitude wetlands, but large uncertainties exist due to the sparsity of long-term, direct flux observations. Previous studies have provided indirect indications of increased carbon loss from northern permafrost regions, but clear trends in methane emissions have not been established through atmospheric measurements, inversion models, or biospheric models. The impact of rising temperatures on the counteracting processes of methane production and oxidation remains unclear. Previous work has highlighted the vulnerability of permafrost carbon to climate change, emphasizing the potential for significant methane feedbacks to the global climate system.

This research analyzed a 16-year (2002-2019) eddy covariance methane flux dataset from a permafrost site in the North Siberian Lena River Delta. This dataset represents the longest such record from the Arctic, enabling a trend analysis of directly observed methane fluxes. The analysis encompassed the four-month thawing season (June-September), accounting for varying data availability across years. Gap-filled time series of daily mean methane fluxes were analyzed to determine trends using modified Mann-Kendall tests. The study examined flux dynamics, estimated annual and seasonal flux budgets, and identified key flux drivers and proxies to explain observed fluxes across various timescales (diurnal, intra-annual, and inter-annual). Flux drivers considered included air and soil temperature, thaw depth, growing degree days (GDD), surface albedo, and friction velocity. Relationships between methane fluxes and these drivers were investigated using statistical methods, including regression analysis.

Statistically significant increases in methane emissions were observed in June and July (0.41 ± 0.16 mmol m² yr⁻¹ and 0.44 ± 0.15 mmol m² yr⁻¹, respectively). These increases correspond to a relative increase of 2.1 ± 0.8% yr⁻¹ and 1.7 ± 0.6% yr⁻¹, respectively. The increase in early summer emissions is linked to a significant air temperature rise of 0.3 ± 0.1 °C yr⁻¹ in June, which advanced the growing season by approximately 11 days. This warming likely affected methanogenesis more strongly than methanotrophy. No significant trends were detected in soil temperature or near-surface hydrology in June, suggesting that atmospheric warming primarily drives the earlier onset of vegetation. The increased early summer methane emissions are likely due to an earlier snowmelt and vegetation development, leading to earlier release of labile organic substances from plants providing substrates for methanogenesis and facilitating methane transport through plant aerenchyma. In contrast, the maximum source strength in August remained unchanged. A marginal decline in September was detected, but its reliability was questioned due to limited data availability. The mean annual methane budget was estimated at 171.5 ± 12.3 mmol m⁻² yr⁻¹, with the thawing season contributing roughly 61%. Diurnal variations in methane fluxes were primarily driven by friction velocity, which also influenced multi-day variations through pressure pumping and methane bursts during storms. On an intra-annual scale, carbon dioxide fluxes served as a proxy for methane fluxes during the non-thawing season, while thaw depth was an indicator of methane flux in the mid-thawing season. On the annual scale, air and soil temperatures were major predictors of methane fluxes, with air temperature showing greater predictive power in early summer. A temperature sensitivity (Q10) of 1.9 was observed from March to July, and 1.5 from August to February.

This study provides the first observational evidence of increasing methane emissions from an Arctic tundra site, linked to atmospheric warming. The increase in early summer emissions, despite no change in the peak August fluxes, underscores the sensitivity of permafrost ecosystems to even moderate warming. The strong correlation between air temperature and early summer methane fluxes suggests that atmospheric warming's effect on vegetation phenology and organic matter availability is a key driver of increased emissions. The lack of significant changes in soil temperature or hydrology highlights the importance of considering atmospheric warming's direct impact on vegetation and microbial processes in models of permafrost carbon cycling. These findings have significant implications for refining predictions of future methane emissions from thawing permafrost and understanding the positive feedback loop between climate warming and Arctic methane release.

This study provides compelling observational evidence of increasing early summer methane emissions from a Siberian tundra site, linked to atmospheric warming. The findings highlight the vulnerability of permafrost ecosystems to climate change and the importance of considering the direct effects of atmospheric warming on vegetation and soil processes. Further research should focus on expanding long-term monitoring efforts across diverse Arctic regions to better understand the spatial and temporal variability of methane emissions and to improve the accuracy of global methane budget estimates. This includes investigating the role of different plant communities and soil types and exploring the potential for future tipping points in permafrost methane release.

The study focuses on a single site in the Lena River Delta, limiting the generalizability of the findings to other Arctic regions. The gap-filling procedure used for the methane flux data might introduce some uncertainties, although the results were consistent using both gap-filled and non-gap-filled data. The analysis focuses on methane fluxes; a more comprehensive study that accounts for methane oxidation would improve understanding of the carbon budget. Limited data availability in September restricts the conclusions drawn from the September data.