The role of the lithological setting on the landslide pattern and distribution

• A detailed lithological map was constructed at a scale of 1:10,000.
• The methodology used was the stereoscopic interpretation of aerial photographs.
• The field validation of the detailed lithological map was done.
• The landslide distribution was analysed with respect to the detailed lithological map.
• A comparison between the official and the detailed lithological data was made.

In landslide susceptibility mapping studies, the quality of whole inputs such as geological setting and DEM derivatives should be as high as possible. Given the strong relationship between landslide occurrences and geology the use of a detailed lithological map is certainly an alternative preferable to the most commonly available rough geological map. Therefore, this study focuses on this reality and proposes a procedure for improving the available geological map in a study area located 90 km North of Lisbon.Based on a 1:50,000 scaled official geological map and according to the litho-stratigraphic characteristics, the map improvement was performed using the interpretation of aerial photographs by stereoscopic analysis and further field validation.In order to assess the relevance of the lithological improvement regarding the relationships with landslide distribution, a comparison was performed between the results obtained with the official lithological map and the improved detailed lithological map using the Conditional Probability (CP), the Accountability (ACC) and Reliability (RLB) indexes. A landslide inventory, including shallow and deep-seated landslides, was prepared for the study area and landslides were overlapped with both lithological maps. The ACC and RLB indexes obtained with the detailed geological map are higher than equivalent features obtained with the official lithological map. From these procedures it was possible to conclude that the detailed lithological map has a better discriminating power than the official lithological map, enabling a better quality of the landslide susceptibility analysis with the separation of the relevant lithological classes.