Forests have a very significant protective function against the impacts of natural hazards. Identifying the protective forests is therefore of crucial importance in order to define proper technical and silvicultural measures able to maintain or improve their function. The active mitigation function of the forests could completely prevent the triggering of an avalanche; otherwise, the passive function (i.e. increasing the friction of the snow mass while moving) is quite limited or null in case of large sized events. In this work, an automated approach for avalanche hazard mapping is used for identifying the forests playing a protective role in avalanche risk mitigation in the Trento Province (Italy). A digital elevation model with a spatial resolution of 5m was used for the terrain classification; a layer identifying the forested areas was calculated by combining a forest map with the information derived from a canopy height model; the climatic snow cover characteristic, necessary to define the fracture heights, were derived by analysing the data of 53 manual weather stations, which have a recorded dataset longer than 17 years. The potential release areas were automatically defined with an algorithm for a scenario with and one without forest. The fracture heights were defined depending on the micro-climatic area and the elevation of each avalanche release area. Then, the avalanche dynamics model RAMMS was applied to simulate the avalanche tracks and deposition areas for an extreme scenario with a return period of approximately 100-300 years. By comparing the results obtained considering both the forest effects and not, the forests playing an active function for avalanche risk mitigation were identified. Even if we considered an extreme scenario only in which the forests playing a passive protection role have small effects, the passive protection forests were still accounted for hazard indication mapping. The obtained results show the applied automated approach is a big help for large-scale identification of the protection forest. Especially in this period of strong climate changes, both the forest cover and snow cover characteristics can change rapidly; thus faster approaches for their evaluation are even much needed. Even if automated approaches are less precise than an expert based hazard map, they are significantly less expensive both money and time - wise, therefore more suitable to be applied on large areas such as for an entire province.