In this work SamosAT was analyzed and evaluated with eld data. The newly developed velocity evaluation approach showed to be highly applicable for two-dimensional simulation results in natural terrain. A large number of simulation runs was analyzed according to the extended AIMEC approach. The introduced discrepancy indicator appeared to be a good representative of the dierence between simulation results and eld observations. The evaluation revealed that the simulations underestimate velocities, in comparison to the corresponding measurements, by ..16:10% (average) and ..10:41% (best t) over all investigated avalanches. It has to be noted that the evaluation has been performed for dierent types of avalanches, which are mostly accompanied by low return periods. For large scale avalanches from the investigated sample, which are assumed to correspond to higher return periods, the velocities were underestimated by up to ..40% to ..50% averaged along the path. Furthermore it is important to note that the discrepancy indicator is not reecting the maximum velocity discrepancy that might occur locally along the path. Comparing pressure outlines of the best tted velocity simulation runs to mapped deposition and impacted areas showed that the spatial extend along the track was too large especially for smaller avalanches. Overall the presented approach provides new operational simulation concepts and evaluation methods explicitly developed for two-dimensional simulation results. The analysis and description of the SamosAT entrainment modules showed the potential of the implemented modules and highlighted the importance of new modelling concepts to properly account for entrainment. A calibration, including entrainment should only be performed with respect to a recalibration of SamosAT, based on an advanced simulation concept, such as the newly introduced mountain snow cover. The possibility of evaluating simulated time series with eld data appears promising. Pressure values were in reasonable ranges. Special care has to be taken to distinguish physical and numerical owe ects. For a detailed analysis new evaluation approaches are necessary.
116 (Hydrologie. Pflege des Wasserhaushalts, Bodenerhaltung (soil conservation) und Erosion. Bewirtschaftung von Einzugsgebieten) 384.1 (Verbauungen gegen Lawinen, Erdrutsche usw.)
