In commission of the Federal Service for Avalanche and Torrent Control, Austria (WLV), the Federal Research and Training Center for Forests, Natural Hazards and Landscape (BFW), Department of Natural Hazard and Alpine Timberline, Austria carries out a systematic model test of the Austrian avalanche model SamosAT. The present report focuses on the state of the art. and a first parameter check. Preliminary assessment of the performance of SamosAT suggests:
. That runout distances of the dense flow of extreme avalanches might be reasonably well predicted in many cases with the present rheological approach for the dense flow part and the present parameter set. However, with increasing flow height, there is a tendency that runouts may be overestimated.
. Velocities are largely underestimated. This may also lead to an underestimation of the simulated pressures, as the fixed density of 200 kg m-3 in the dense part cannot compensate for the velocity discrepancy. Definitely, an underestimation of the flow velocity also influences coupled dense/powder flow runs as the transition from the dense part into suspension depends on the velocities of both parts. Therefore, underestimation of the velocities must be considered as a crucial fault of SamosAT.
. Predicted peak flow heights of the dense part seem to be on the high side compared to field observations.
. The erosion model (frontal erosion) contributes to the low velocity and is not be counterbalanced by higher flow heights. There are indications that frontal erosion causes numerical instabilities or non-physical behavior, respectively.
Recommendation: The underestimation of the avalanche velocity is a conceptual problem of the rheological model used in SamosAT. Therefore, it is proposed to reconsider the implemented stress (bottom friction) model. As mass entrainment is a decisive factor for the mass balance of avalanches, the retardation due to the common dissipative forces and due to mass intake needs to be considered jointly. This holds also true for the interaction between the dense/fluidized part and the suspension layer. Admittedly, it should be noted that increasing the velocity is accompanied by an increase of the derived impact pressure as long as the density is kept constant. Therefore, it might be necessary to introduce a variable density to be consistent with velocity and impact pressure measurements.
