We used one of the few rockfall models explicitly taking trees into account and compared the results obtained with the 3D simulation model RockyFor with empirical data on tree impacts at three mountain forests in Switzerland. Even though we used model input data with different resolutions at the study sites, RockyFor accurately predicted the spatial distribution of trajectory frequencies at all sites. In contrast, RockyFor underestimated mean impact heights observed on trees at the two sites where high-and medium-resolution input data were available and overestimated them at the site where input data with the lowest resolution data were used. By comparing the results of the simulation scenarios ''current forest cover'' and ''non-forested slope'', we assessed the protective effect of the current stands at all three sites. The number of rocks reaching the bottom parts of the study sites would, on average, almost triple if the ''current forest cover'' were absent.We conclude that RockyFor is able to predict the spatial distribution of rockfall trajectories on forested slopes accurately, based on input data with a resolution of at least 5 m  5 m. With the increasing availability of high-resolution data, it provides a useful tool for assessing the protective effect of mountain forests against rockfall.

André WehrliFederal Office for the Environment (CH) state of the art in rockfall -forest interactionsTo effectively prevent rockfall related disasters below forested slopes, silvicultural, eco-engineering, civil engineering or mixed techniques can be used. To do this in a cost-efficient manner it is necessary to know the following: 1) where rockfall events occur and which magnitudes are likely, 2) to what extent the forest reduces the run-out distances, the jump heights and the energies of rocks falling downslope, and 3) how the protective function of forests could be improved. This paper gives an overview of the current scientific knowledge and methods that are applied by practitioners who deal with rockfall and forests protecting against it. Efficient ways to derive information on the probable magnitude and frequency of future rockfall events from the source and deposit area are described. Subsequently, the scientific knowledge on the energy absorption capacity of single trees and the currently available knowledge on the protective function of forest stands against rockfall are presented. Then easy-to-use tools and simulation models for rockfall hazard assessment on forested slopes are described. Finally, this paper identifies the most important challenges to be tackled in the field of integrated rockfall-forest research.

Simulation models such as forest patch models can be used to forecast the development of forest structural attributes over time. However, predictions of such models with respect to the impact of forest dynamics on the long-term protective effect of mountain forests may be of limited accuracy where tree regeneration is simulated with little detail. For this reason, we improved the establishment submodel of the ForClim forest patch model by implementing a more detailed representation of tree regeneration. Our refined submodel included canopy shading and ungulate browsing, two important constraints to sapling growth in mountain forests. To compare the old and the new establishment submodel of ForClim, we simulated the successional dynamics of the Stotzigwald protection forest in the Swiss Alps over a 60-year period. This forest provides protection for an important traffic route, but currently contains an alarmingly low density of tree regeneration. The comparison yielded a significantly longer regeneration period for the new model version, bringing the simulations into closer agreement with the known slow stand dynamics of mountain forests. In addition, the new model version was applied to forecast the future ability of the Stotzigwald forest to buffer the valley below from rockfall disturbance. Two scenarios were simulated: (1) canopy shading but no browsing impact, and (2) canopy shading and high browsing impact. The simulated stand structures were then compared to stand structure targets for rockfall protection, in order to assess their long-term protective effects. Under both scenarios, the initial sparse level of tree regeneration affected the long-term protective effect of the forest, which considerably declined during the first 40 years. In the complete absence of browsing, the density of small trees increased slightly after 60 years, raising hope for an eventual recovery of the protective effect. In the scenario that included browsing, however, the density of small trees remained at very low levels. With our improved establishment submodel, we provide an enhanced tool for studying the impacts of structural dynamics on the long-term protective effect of mountain forests. For certain purposes, it is important that predictive models of forest dynamics adequately represent critical processes for tree regeneration, such as sapling responses to low light levels and high browsing pressure.

No abstract