GNGTS 2013 - Atti del 32° Convegno Nazionale

Some very thin clayey interbeds were identified in the upper part of the Fonzaso Formation (unit a’ ) and indicated by Hendron and Patton (1985) as the stratigraphic level where the initial sliding happened. There are just few outcrops of this unit as during the failure the unit was dragged and crunched along the sliding surface. The bottom level of unit a’’ is represented by the Ammonitico Rosso Formation (Rossi and Semenza, 1965) while the body of unit a’’ is comprised of an alternation of layered cherty limestone and marly limestones. Unit b is represented by a thin conglomerate layer and it is very important because it represents an isochron surface and it’s a marker level within the landslide. Units c , d and e are comprised of massive limestones grading to layered marly and cherty limestones. Finally unit f represents the top of the Soccher Formation and it is comprised of layered marly cherty limestones. A review of available structural data, associated with some new field observations, has been recently completed (Massironi et al. , 2013). According to these data the E-W trending Erto syncline (Giudici and Semenza, 1960) is further folded by a N-S trending syncline with it’s axis elongated along the pre-landslide Massalezza valley. The interference of these two sets is exposed on the sliding surface and in some cases the undulations generated steps that interrupted the continuity of the surface itself. The ensemble of these structural features appears to be a major constrain in the kinematics of the landslide. In particular the association of the stratigraphy and of the N-trending bedding planes with the curved shape of the sliding surface converging on the Massalezza valley behave as a track for the 1963 failure. The physical database. The success of the geophysical experiment, in reconstructing the landslide stratigraphy and structure, strongly depends on the possibility of measuring some key properties of the in-situ formations. A reliable and accurate physical reference model of the upper part of the Fonzaso Formation and of the Soccher Limestone Formation was then constructed to reduce the uncertainty in correlating physical properties (e.g. velocity and resistivity) to specific lithological units. This robust tie could be also a constrain while attempting to estimate the degree of fracturing of the landslide mass itself. The first indicator of physical properties is the rock coherency. Hard rocks (HR) will be probably characterized both by high velocity and high resistivity while soft rocks (SR) are generally low velocity and low resistivity. In this simplified approach units a’ , a’’ and f could be classified as moderately soft rocks while units b , c , d and e are mostly referable to moderately hard rocks. Some of the boreholes drilled on the landslide accumulations where also used as major constrains in assisting geophysical data analysis, processing and interpretation. The majority of these boreholes reached the depth of the in-situ bedrock. The few geophysical data available for the left side of the Vajont valley, before the failure, are represented by four seismic profiles collected immediately after the discovery of the paleolandslide (Caloi and Spadea, 1960, 1961). Results from a first seismic survey indicated how the P-wave velocity in the uppermost layers was higher than 5000 m/s. Some authors (Semenza, 1965; 2001; Selli and Trevisan, 1964) pointed out that these values appear to be too high also for compact limestones. A second seismic survey (Semenza, 2001) carried out after the discovery of the perimetrical crack, showed more realistic P-wave velocities around 2500- 3000 m/s. The Court of Belluno to collect new evidences for the trail during the preparation of the case required a new seismic survey. The major purpose of this new survey was to collect data inside and outside the landslide area. Seismic velocities were measured at few locations (Morelli and Carabelli, 2005) inside and outside the landslide area. Unfortunately the measurements on the landslide body were carried out in boreholes mostly below the sliding surface making this new data not directly comparable with the previous surveys. Other measurements carried out on the Cretaceous sequence outside the landslide show P-wave velocities ranging from 2100 m/s to 3000 m/s. These numbers are more and less similar to the values measured during the second seismic survey before the failure. 191 GNGTS 2013 S essione 3.3