The different hazard levels in avalanche hazard maps are defined on the basis of the impact pressures of the design avalanche with a specific return period. Avalanche dynamics models are needed to define such limits. For a single path, the definition of the potential release area (PRA) of the design avalanche and the other model input parameters is generally made on an expert-based assessment. For large areas, such expert-based approach is too time consuming and therefore not feasible.
A procedure to automatically produce avalanche hazard indication maps is here applied in a test site in Aosta Valley in the Western Italian Alps. The large scale hazard mapping (LSHM) procedure is described in Bühler et al. (2018). In this work, we present the comparison of the results of the LSHM procedure with the already existing avalanche hazard maps in the upper Gressoney's Valley, in order to verify the potential of such automatic procedure.
We made a first run of the procedure completely automatically. Then, we performed a second run, after a manual expert-based selection of the automatically identified PRAs. To summarize, the main differences between the limits of the hazard zones in the official hazard maps and the results of the procedure arose mainly from the definition of the PRAs and of the forested areas.
In conclusion, the automatic procedure generally produced hazard zones in good agreement with those in the official maps. This case study illustrates the high potential of the automatic, not time-consuming, procedure for large areas.