A methodological approach for the analysis of shallow landslides in non-collapsible soils

Abstract

Shallow landslides in non-collapsible, fine-grained soils are natural phenomena which, due to their simultaneous occurrence over large areas, often cause agricultural production loss, with enormous economic damage. Although the scientific community has offered in depth analyses of landslides in collapsible soils, shallow landslides in non-collapsible soils, have been paid very little attention in spite of their consequences. The few works available in scientific literature are based on monodisciplinary approaches which provide useful specific information on these phenomena; yet, at the same time, they also oversimplify the problem, often neglecting some relevant issues. Such issues can be best highlighted and addressed following a multisisciplinary and multiscalar (deductive or inductive) approach, as some authors have evidenced in very different geological and environmental contexts affected by medium- to deep -seated landslides. Within such a context, rich in approaches but lacking in references to shallow landslides in non-collapsible soils, the present thesis seeks to offer a methodological contribution for creating a homogeneous reference framework for the phenomena examined. At the same time, this work seeks to answer specific questions concerning the ways in which these phenomena are triggered. The proposed methodology has been tested and validated over an area in southern Italy, in which shallow landslides are very frequent. The predisposing factors and the triggering causes have been analysed going from small scale, through geological models, to large scale by means of geotechnical analyses performed at slope and large scale. On the one hand, this approach allowed to identify the predisposing factors of the morphological evolution of the reliefs in the main tectonic structures, drainage lines and in the main lithology; on the other hand it allowed the characterization of large scale and detailed scale landslides by identifying their triggering mechanisms, and by investigating specific aspects such as the hydraulic and mechanic properties of the soil and the pore water pressure regime in situ. In particular, three main triggering mechanisms have been identified and analysed at large scale which differ in terms of morphometric characteristics, volumes involved, and which do not show a casual slope distribution. Subsequently, a (saturated and unsaturated) modelling of groundwater regime for these mechanisms has been performed which provided the input data for stability analysis carried out through physically based and limit-equilibrium methods. The overall validity of the results obtained highlights the efficacy of the methodology applied and recommends its use in geo-environmental contexts which differ from the one in which this methodology has been tested and validated. For this purpose, at the end of the present thesis, a series of general considerations have been discussed by making use of preliminary applications to internationally well-known case studies available in scientific literature. [edited by author]