Simulating wind disturbance impacts on forest landscapes: Tree-level heterogeneity matters

Wind is the most detrimental disturbance agent in Europe's forest ecosystems. In recent years, disturbance frequency and severity have been increasing at continental scale, a trend that is expected to continue under future anthropogenic climate change. Disturbance management is thus increasingly important for sustainable stewardship of forests, and requires tools to evaluate the effects of management alternatives and climatic changes on disturbance risk and ecosystem services. We here present a process-based model of wind disturbance impacts on forest ecosystems, integrated into the dynamic landscape simulation model iLand. The model operates at the level of individual trees and simulates wind disturbance events iteratively, i.e., dynamically accounting for changes in forest structure and newly created edges during the course of a storm. Both upwind gap size and local shelter from neighboring trees are considered in this regard, and critical wind speeds for uprooting and stem breakage are distinguished. The simulated disturbance size, pattern, and severity are thus emergent properties of the model. We evaluated the new simulation tool against satellite-derived data on the impact of the storm Gudrun (January 2005) on a 1391 ha forest landscape in south central Sweden. Both the overall damage percentage (observation: 21.7%, simulation: 21.4%) as well as the comparison of spatial damage patterns showed good correspondence between observations and predictions (prediction accuracy: 60.4%) if the full satellite-derived structural and spatial heterogeneity of the landscape was taken into account. Neglecting within-stand heterogeneity in forest conditions, i.e., the state-of-the-art in many stand-level risk models, resulted in a considerable underestimation of simulated damage, supporting the notion that tree-level complexity matters for assessing and modeling large-scale disturbances. A sensitivity analysis further showed that changes in wind speed and soil freezing could have potentially large impacts on disturbed area and patch size. The model presented here is available as open source. It can be used to study the effects of different silvicultural systems and future climates on wind risk, as well as to quantify the impacts of wind disturbance on ecosystem services such as carbon sequestration. It thus contributes to improving our capacity to address changing disturbance regimes in ecosystem management.