Thomas Pugh

European forests have served as a critical carbon sink, absorbing approximately 10% of annual anthropogenic CO₂ emissions in the EU, as a result of both increased forest area and growing stock since 1950. However, this sink has declined by one-third over the past decade, based on the EU National Greenhouse Gas Inventory. This decline has been proposed to be due to multiple interacting factors: climate conditions that now constrain rather than promote tree growth, increased ecosystem respiration, forest aging, more frequent disturbances, and rising harvest rates. Yet, the spatiotemporal dynamics of the carbon sink-source shift and the underlying drivers remain unclear.
In this study, we employ a European-focused version of the dynamic vegetation model (DVM) LPJ-GUESS to simulate forests in 12 EU countries, comprising 71.79% of the EU forest area, over 2011-2023. The model incorporates newly parameterized plant functional types (PFTs) for common European species, satellite-derived forest disturbance inputs, a harvest module based on National Forest Inventories, and forest age aligned with observed age distributions.
Our results show a decreasing carbon sink trend from 2011 to 2023 across European forests, with pronounced interannual variability. The magnitude and direction of sink-source changes vary across countries and regions over time. In identified hotspot areas, different drivers dominate or interact – such as growth reduction, increased soil respiration, intensified harvesting, fire, and other disturbances including bark beetle outbreaks and windthrow.
By comparing living biomass changes from model simulations with country-reported data from Land Use, Land-Use Change and Forestry (LULUCF) inventories, our results indicate the model’s ability to reproduce national-level forest carbon sink trends and its potential to further assist LULUCF reporting by providing a consistent continental-scale framework to explore drivers of sink-source changes and identify priority areas for actions. Our findings suggest that well-constrained DVM simulations can provide powerful insights into both long-term ecosystem trajectories and short-term regional events, contributing to forest management and climate policy development.