|
In this paper the landslide hazard of a rocky cliff has been assessed using the methodology known as Slope Mass Rating [Romana, 1985], in order to estimate its potentiality, its difficulty of application and its suitability to specific geomorphological contexts. The chosen area for applying the method is the Valley of the Pardu River, placed in the eastern-central part of Sardinia, Italy, which, on the basis of the Census of the Hydrogeological Disasters Damaged Areas, turns out to be one of more hazardous zone in Sardinia for the presence of ancient and current landslides, whose causes are to be searched in the structural and lithological characteristics, the extreme meteorological conditions and the anthropic factors [Ardau et al. 2005].
The particular geomorphological shape of the Gairo Taquisara hillside, the general lack of vegetation, the high steepness of the cliff and the presence of many joint families in the dolomitic limestones give rise to landslide phenomena like falls and topples.
When the landslide hazard evaluation involves walls of bared rocks or strongly steep, substantially threatened by landslides phenomena like falls, topples, slides of wedges or slabs of rock, the methods commonly used for the hazard assessment, such as statistical methods or Overlay Mapping methods, don't seem entirely appropriate to represent the real conditions of the landslide susceptibility, because the phenomena of instability are in these cases strongly influenced by the geostructural characteristics of the rock formations and the relationship between the orientation of the rock walls and the joints systems. It was therefore thought to utilize the Romana method, which assesses the stability conditions of the cliffs through the Slope Mass Rating (SMR) index, which calls into account, in addition to the geomechanic strength of the intact rock, both the geostructural style, through the assessment of the characteristic parameters of the joints, and the relationship between the orientation of the rock walls and the joints systems For computing the SMR index of the different areas of the cliff, the strength index of the rocks and the basic parameters of the joints, like orientation of the main joints families, orientation of the rock walls, joints opening, joints spacing and RQD index, have been surveyed.
To verify the results of the analysis, the map of the landslide hazard of the cliff obtained with the RMR method has been compared with the corresponding map obtained with the Neural Network method. The comparison highlighted as the proposed method has given very satisfactory results, identifying in high or very high hazard classes most of the areas identified as very instable by the neural method.
|
