Large stocks of peatland carbon and nitrogen are vulnerable to permafrost thaw

Abstract

Northern peatlands have accumulated large stocks of organic carbon (C) and nitrogen (N), but their spatial distribution and vulnerability to climate warming remain uncertain. Here, we used machine-learning techniques with extensive peat core data (n > 7,000) to create observation-based maps of northern peatland C and N stocks, and to assess their response to warming and permafrost thaw. We estimate that northern peatlands cover 3.7 ± 0.5 million km² and store 415 ± 150 Pg C and 10 ± 7 Pg N. Nearly half of the peatland area and peat C stocks are permafrost affected. Using modeled global warming stabilization scenarios (from 1.5 to 6 °C warming), we project that the current sink of atmospheric C (0.10 ± 0.02 Pg C⋅y^-1) in northern peatlands will shift to a C source as 0.8 to 1.9 million km² of permafrost-affected peatlands thaw. The projected thaw would cause peatland greenhouse gas emissions equal to ∼1% of anthropogenic radiative forcing in this century. The main forcing is from methane emissions (0.7 to 3 Pg cumulative CH₄-C) with smaller carbon dioxide forcing (1 to 2 Pg CO₂-C) and minor nitrous oxide losses. We project that initial CO₂-C losses reverse after ∼200 y, as warming strengthens peatland C-sinks. We project substantial, but highly uncertain, additional losses of peat into fluvial systems of 10 to 30 Pg C and 0.4 to 0.9 Pg N. The combined gaseous and fluvial peatland C loss estimated here adds 30 to 50% onto previous estimates of permafrost-thaw C losses, with southern permafrost regions being the most vulnerable.

Keywords: carbon stocks; greenhouse gas fluxes; nitrogen stocks; northern peatlands; permafrost thaw.

Fig. 1.

Peatland data and properties north of 23°N latitude. (A) Estimated areal coverage (in percentage) of peatlands based on the national soil inventory maps and SoilGrids250m. (B) Estimated areal coverage (in percentage) of permafrost in mapped peatlands based on the national soil inventory maps and SoilGrids250m, including a maximum threshold for permafrost at MAAT +1 °C (use the same legend as in A). (C) Spatial distribution of peat core sites with peat depth data (n = 7,111) and peat organic C storage (n = 782) over a map of biome distributions (biomes adapted from ref. 32). Sites with peat N stock data (n = 105) are not shown in the map (see Dataset S6), but are predominantly located in boreal forest and tundra biomes. (D) Sites with peat organic C storage data, with the size of site symbols proportional to measured peat organic C storage, over a map of permafrost zonation (33). (E) Estimated total peatland C storage and (F) permafrost peatland C storage.

Fig. 2.

Projected permafrost loss from peatlands and the conceptual model of permafrost thaw impacts on GHG fluxes. (A) Map of projected permafrost loss from peatlands. Colors indicate temperature thresholds when equilibrium permafrost extent drops to less than 10%. (B) Projected equilibrium extent of permafrost in peatlands for global warming stabilization scenarios. Note that zero degree reflects preindustrial warming, with present-day climate already being close to +1 °C. (C) Spatial model of peatland transitions, including the properties and GHG balances of the different degradation and recovery stages. (D) Schematic mean (±SD) annual peatland GHG balances for the different stages of thaw and recovery (weighted average of all pixels). Negative numbers indicate C loss from peatlands, that is, a flux to the atmosphere (upward in the figure). Fluxes of CO₂ for the transient permafrost thaw stages, as well as CH₄ and N₂O fluxes, are synthesized from observed fluxes in field or experimental studies (Datasets S1 and S5).

Fig. 3.

Projected GHG flux and radiative forcing from peatland permafrost thaw (calculated from the net change in GHG flux relative to stable peatlands) under different global warming stabilization scenarios. (A and B) Projected radiative forcing from GHG fluxes for three centuries for +2 and +4 °C global warming. (C) Added radiative forcing from peatlands in relation to human GHG emissions for this century, assuming that the +0.5 °C degree warming was passed in 1990 and starting the peatland thaw scenarios from that year (human radiative forcing consistent with the different warming trajectories) (60). (D) Projected gross lateral losses of peat C into aquatic systems (mainly as DOC/POC) inferred from permafrost thaw chronosequences. (E and F) Net peatland C-balance following the active layer deepening and young thermokarst thaw stages at +2 and +4 °C global warming.