The decomposition of thawing permafrost organic matter (OM) to the greenhouse gases (GHG) carbon dioxide (CO2) and methane forms a positive feedback to global climate change. Data on in situ GHG fluxes from thawing permafrost OM are scarce and OM degradability is largely unknown, causing high uncertainties in the permafrost-carbon climate feedback. We combined in situ CO2 and methane flux measurements at an abrupt permafrost thaw feature with laboratory incubations and dynamic modeling to quantify annual CO2 release from thawing permafrost OM, estimate its in situ degradability and evaluate the explanatory power of incubation experiments. In July 2016 and 2019, CO2 fluxes ranged between 0.24 and 2.6 g CO2-C m(-2) d(-1). Methane fluxes were low, which coincided with the absence of active methanogens in the Pleistocene permafrost. CO2 fluxes were lower three years after initial thaw after normalizing these fluxes to thawed carbon, indicating the depletion of labile carbon. Higher CO2 fluxes from thawing Pleistocene permafrost than from Holocene permafrost indicate OM preservation for millennia and give evidence that microbial activity in the permafrost was not substantial. Short-term incubations overestimated in situ CO2 fluxes but underestimated methane fluxes. Two independent models simulated median annual CO2 fluxes of 160 and 184 g CO2-C m(-2) from the thaw slump, which include 25%-31% CO2 emissions during winter. Annual CO2 fluxes represent 0.8% of the carbon pool thawed in the surface soil. Our results demonstrate the potential of abrupt thaw processes to transform the tundra from carbon neutral into a substantial GHG source.

Permafrost regions with high soil organic carbon (SOC) storage are extremely vulnerable to global warming. However, our understanding of the temperature sensitivity of SOC decomposition in permafrost regions remains limited, leading to considerable uncertainties in predicting carbon-climate feedback magnitude and direction in these regions. Here, we investigate general patterns and underlying mechanisms of SOC decomposition rate and its temperature sensitivity (Q(10)) at different soil depths across Tibetan permafrost regions. Soils were collected at two depths (0-10 and 20-30 cm) from 91 sites across Tibetan permafrost regions. SOC decomposition rate and Q(10) value were estimated using a continuous-flow incubation system. We found that the SOC decomposition rate in the upper layer (0-10 cm) was significantly greater than that in the lower layer (20-30 cm). The SOC content governed spatial variations in decomposition rates in both soil layers. However, the Q(10) value in the upper layer was significantly lower than that in the lower layer. Soil pH and SOC decomposability had the greatest predictive power for spatial variations in Q(10) value within the upper and lower layers, respectively. Owing to the greater temperature sensitivity in the lower layer, our results imply that subsurface soil carbon is at high risk of loss, and that soil carbon sequestration potential might decrease in these regions in a warming world.