01 Evaluating the effect of rock shape and deadwood on induced rockfalls in forests

Ringenbach Adrian, Caviezel Andrin, Bebi Peter, Bartelt Perry

  1. WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland

Forests play an important role in rockfall mitigation, due to the rock interaction with living and fallen trees – so-called “deadwood”. Recent rockfall experiments in open terrain (Caviezel et al., 2019) show that the rock shape has a major influence on the lateral dispersion of rock trajectories. Rock-tree interaction modeling demonstrates the importance of rock shape (Lu et al., accepted). Consequently, novel experimental rockfall studies in forested areas aim to close this knowledge gap. Of major interest are forest effects to the rockfall trajectory arising from obvious rock tree interactions, ramping effects due to tree stumps, and general forest soil elasticity. Additionally, the braking effect of deadwood is scrutinized. As increasing rates of deadwood production by natural disturbances are expected, this effect may become a primary mitigation measure to reduce rockfall runout.

We present results of real scale experiments with different natural shaped and artificial, instrumented rocks in forests with (n=96) and without (n≥34) deadwood. The experimental site is located in Schiers in a beech-silver-fir-spruce forest. The deadwood originates from the Burglind storm in January 2018. We find that clusters of deadwood stop a significant number of released rocks. In the segment of the larger rock masses (250 kg – 840 kg) logs stop 2.5 times more rocks (n=16) then living trees (n=6). Platy rock shapes generally feature shorter travel distances than compact shapes. Platy rocks feature a unique optimal traveling configuration, i.e. rolling like a wheel, which can be disturbed at every tree impact. Compact shapes are more likely to restart rolling motion after a tree impact due to the absence of a larger flat surface. The preliminary results of trajectory reconstructions confirm the expectations of higher velocities without deadwood (n=2) compared to those with deadwood (n=7). This unique dataset is used to calibrate the forest representation within the RAMMS::ROCKFALL model.