The release of dry-snow slab avalanches is often associated with faceted crystals around crusts. This type of persistent weak layer mainly arises due to a dry-wet (DW) faceting mechanism. Based on pattern recognition of this snow stratigraphy signature the Avalanche Warning Services Tirol introduced the so-called Danger Pattern Four (dp.4) - cold following warm, warm following cold. Since detecting this signature within snow profiles can be challenging for inexperienced observers, this thesis presents an approach to improve the accessibility of existing data by employing a pattern recognition algorithm and incorporating simulated profiles in further analyses. These analyses rely on a recently developed snow profile alignment algorithm which has the potential to detect groups of snow profiles in which faceted crystals have an impact on the interpretation of snowpack stability and consequently avalanche hazard assessment. Patterns and characteristics within those groups are investigated. To enhance the accessibility of information related to dp.4 characteristics within these groups, a specific signature is derived. This is done by aligning a set of snow profile pairs selected according to certain criteria. These pairs are either composed of observed profiles or a combination of observed and modeled profiles. The signature is derived from a statistical evaluation of similarity values based on differences in relevant snowpack stability indices, e.g. relative threshold sum approach, grain type and layer hardness. Applying the signature on the results of a clustering algorithm addresses the question, at what time and terrain location dp.4 characteristics developed.
Despite the low accuracy of the signature, its application to clusters of snow profiles tends to detect dp.4 characteristics in a large number of profiles. Considering these significant clusters, including observed and modeled profiles, the profile characteristics of the clusters suggest possible spatial patterns described by regions, elevation, or aspects where dp.4 conditions may occur. In other words, applying the signature on a set of snow profiles can help to point out regions with hazardous DW faceting conditions. With that information, we provide an approach to support so far human-based decisions with machine-based quantitative statements.