Assessment of the landslide dams in Western Austria, Bavaria and Northern Italy (part of the Eastern Alps): Data inventory development and application of geomorphic indices

Abstract

Landslide dams pose significant hazards towards their upstream and downstream areas due to damming and flooding. Case studies and the use of geomorphic indices, which are developed based on large datasets, are two common approaches in the stability assessment of landslide dams. To gain an insight into the evolution of landslide dams within a part of the Eastern Alps and to provide a basis for further studies on this topic, a data inventory including 73 landslide dams is presented. The database covers sites in Western Austria, Bavaria and Northern Italy, containing 39 descriptive and quantitative parameters gathered in five categories of location, landslide, dam, lake, and catchment. The methodologies utilized for parameter definition follow a simple and consistent procedure described in detail for one case study, the Wiese landslide in Tyrol (Austria). Statistical assessment of the geometric parameters indicated an acceptable collation between mean values of the developed inventory and dataset of other geographical regions. Moreover, by developing a correlation matrix of the quantitative fields, landslide dam parameters showed a fairly strong correlation with the landslide and backwater lake parameters. However, they appear to be statistically independent of the catchment area characteristics. The applicability of the previously developed geomorphic indices is evaluated by plotting the current data and determining the upper and lower bounds of the stable and unstable domains regarding a 95 % confidence level of the mean value. The best applicability is obtained out of the blockage and dimensionless blockage indices. Further, a set of catchment ruggedness-based indices are developed based on the collected data. The indices act upon the catchment properties as driving forces and three different sets of characteristics of the dam and landslide as resisting forces. The reliability of these indices is confirmed by obtaining the adjusted R2 within 70–85 %, narrowing down the uncertain domain on the graphs in comparison to the reviewed literature, and the applicability of them on the current database. Each catchment ruggedness-based index is suitable to predict the evolution of the dams and backwater lakes at the time of occurrence depending on the available data.