Attributing trends in the land carbon cycle using process-based DGVMs and global scale observations

The global land carbon sink has increased in parallel with anthropogenic CO2 emissions over the last several decades, taking up ~25% of these emissions, and acting as a strong negative feedback to mitigate climate change. However, we have a limited ability to confidently attribute past changes. Here we use the suite of Dynamic Global Vegetation Models (DGVMs) from the Global Carbon Budget to develop a process-attribution framework to identify where models agree and, just as importantly, disagree, and thus guide future modelling efforts. We take a holistic approach to answer the following questions:

What are the 1) external drivers (concurrent rises in atmospheric CO2 and nitrogen deposition, climate, land-use and land-cover change (LULCC)), 2) main regions (tropics, extra tropics), and 3) processes (production vs turnover) primarily responsible for the changes in the net land carbon sink?

We find the observed global net land carbon sink is captured by current land models. However, there is a lack of consensus in the partitioning of the sink between vegetation and soil, where models do not even agree on the direction of change in carbon stocks over the past 60 years. This uncertainty is driven by plant productivity, allocation, and turnover response to atmospheric CO2 (and to a smaller extent to LULCC), and the response of soil to LULCC (and to a lesser extent climate). Overall, differences in turnover explain ~70% of model spread in both vegetation and soil carbon changes.

Using a top-down constraint of net land-atmosphere carbon exchange from atmospheric inversions and remote-sensed products of vegetation functioning, we show that DGVMs underestimate carbon uptake in northern latitudes. A large portion of model error can be explained by the simulated LULCC flux. In tropical lands, models likely overestimate net carbon uptake due to too strong CO2 fertilisation, which can, in part, beexplained by too high modelled forest area and carbon densities.