Forests worldwide face increasing risks from natural disturbances such as wildfires, insect outbreaks, and windstorms. These disturbances are highly sensitive to changes in the climate system and have already increased in many parts of the world recently. Changes in disturbance regimes can substantially alter ecosystems, for example, by modifying their population dynamics and habitat value, as well as the ecosystem services they provide to society. Therefore, anticipating potential future changes in disturbances is crucial for forest policy and management. However, projecting future disturbance regimes remains complex because of the intricate interactions between individual disturbance agents, and the feedback loops between vegetation development and disturbance changes could significantly dampen or amplify climate impacts.

JUSTIFICATION

Here we present a modeling framework to simulate future forest disturbance trajectories at high spatial resolution (100 × 100 meters) and across a large spatial extent (187 million hectares of forest in Europe). We leveraged a deep learning-based simulation framework to integrate a large body of local projections generated by process-based forest models with climate-sensitive disturbance modules for wildfire outbreaks, windblown vegetation, and bark beetles. Our modeling framework is designed to capture crucial disturbance processes such as the spatial spread of fire and bark beetles across forest landscapes and incorporates disturbance and vegetation feedback interactions. Our specific objectives were to quantify the potential changes in forest disturbances that replace stands in Europe up to the end of the 21st century under different climate change scenarios and to assess the impacts of disturbance change on European forest demography.

RESULTS:

Forest disturbances in Europe are likely to increase in the coming decades. Simulated levels of future disturbance were higher than those observed between 1986 and 2020 under all climate scenarios. Under unchecked climate change scenarios, the simulated disturbed area more than doubled by the end of the century (+122%). In scenarios assuming effective emissions reductions, peak disturbance was reached by mid-century. Forest fire was the disturbance agent most sensitive to changes in the climate system, severely impacting Mediterranean areas but also expanding into temperate and boreal regions. Vegetation feedbacks buffered the climate-induced disturbance change but could not completely mitigate the increase in disturbance. We project profound implications of future disturbance changes for European forest demography, with an increase in the proportion of young forests to 14% and a decrease in old-growth forests to 3% compared to simulations without changes to climate or disturbance regimes.

CONCLUSION
Large-scale changes in forest disturbance regimes projected for the coming decades have significant implications for biodiversity and the ecosystem services provided by forests. For example, they could hinder policy objectives of using nature-based solutions for climate change mitigation, further amplifying climate change. Consequently, forest policy and management must plan for a future with increased disturbances. However, our results highlight that mitigating anthropogenic climate change remains a powerful lever for limiting future disturbance risk and protecting forests and their services to society.

SCIENCE