Northern high latitudes could become a net carbon source below 2°C global warming

Terrestrial ecosystems in the northern high latitudes have historically acted as a net carbon sink, mitigating anthropogenic CO₂ emissions. However, the long-term stability of this net sink is uncertain due to complex carbon cycle feedbacks in response to future climate change. In this presentation, we will show how the PRIME framework can be used to probabilistically quantify if and when this region will transition from a net carbon sink to a carbon source in a range of plausible future climate scenarios (SSP1-2.6, SSP2-4.5, SSP5-8.5), including overshoot (SSP5-3.4-OS). PRIME incorporates the JULES land surface model, which can explicitly represent permafrost physics, dynamic vegetation, and fire, enabling the simulation of key high-latitude processes that remain uncoupled in most Earth system models. In a low emission scenario, permafrost carbon emissions are shown to increase the risk of a net carbon source by more 50% at 2°C of warming, and at greater levels of warming in high emission scenarios. Conversely, in all emission scenarios dynamic vegetation is found to limit the sink-to-source transition at all warming levels by enhancing the carbon sink. Fire emissions can further weaken the sink by reducing its resilience to warming. In the high temperature overshoot scenario, post-peak cooling leads to less favourable conditions for vegetation growth, limiting recovery of the carbon sink. These results highlight the dominant role of vegetation dynamics in regulating the strength and resilience of the Arctic terrestrial carbon sink under warming. They also emphasise the importance of representing coupled permafrost, vegetation, and fire processes in Earth system models to improve projections of land carbon–climate feedbacks across future climate trajectories.