Adding natural disturbances to a large-scale forest scenario model and a case study for Switzerland

In this study we assessed the impact of climate change and the role of natural disturbances on the development of forest resources. A module dealing with natural disturbances was developed for the European forest information scenario model (EFISCEN), based on the observed distribution of damages in the past and the dependency of disturbances on forest characteristics. To put the model to the proof, the development of Swiss forest resources was projected until 2048 under three alternative scenarios. The first scenario consisted of a run without the new module (the base run), in the second scenario the model was run with the new module under current climate (natural disturbance run) and the third scenario was a run with the new module under a changed climate (natural disturbance and climate change run), where outcomes of the process-based model TREEDYN3 were used to simulate the influence of a changing climate on the volume increment and where the frequency of disturbances was increased to simulate the effect of climate change on natural disturbances. Incorporating natural disturbance dynamics in the EFISCEN model resulted in a more realistic simulation of the total fellings and natural mortality, due to the fact that killed, but unrecovered, timber is also taken into account, resulting in a better simulation of volume and increment development. When the natural disturbances module was used in the simulation, the growing stock increased from 366 m(3) ha(-1) in 1984 to 460 m(3) ha(-1) in 2048, while without disturbances it increased up to 592 m(3) ha(-1). The simulation under current climate showed an increase in damage due to natural disturbances of 40% over the period 2004-2048, due to an increase of growing stock and a higher proportion of older stands. Under a simulated climate change scenario, the frequency of disturbances was assumed to increase, which resulted in 25% higher damages. However, the increment increased more than the damage done by disturbances, which resulted in a simulated growing stock volume of 530 m(3) ha(-1) in 2048. The increase in damage must be attributed to an increasing average standing volume. Because of uncertainties caused by the assumptions made in the model and the stochastic character of the disturbances, the results of this study must be seen as merely indicative. (C) 2002 Elsevier Science B.V. All rights reserved.