REGIONAL SCALE IDENTIFICATION OF SLOPE SUSCEPTIBILITY TO CO-SEISMIC LANDSLIDES IN URBAN AND PERI-URBAN AREAS OF DAUNIA (APULIA REGION – SOUTHERN ITALY)

The aim of this thesis is to propose and test advances in developing an expeditious procedure for identifying the slopes most susceptible to earthquake-induced landslides on a regional scale, in order to determine the areas most prone to develop critical issues following a seismic event. This procedure could provide useful information for improving preparedness for future seismic scenarios and seismic emergency management. To this end, we revisited a probabilistic approach proposed by Del Gaudio et al. in 2003 for estimating the resistance demand posed to slopes by seismicity. Based on the Newmark’s model of 1965, it estimates the critical acceleration a slope should have to keep within a precautionary limit the probability of exceeding critical values of Newmark’s displacement during seismic shaking with an energy level measured by Arias Intensity. Compared to previous implementations of this approach, the main novelty introduced in the present work is the different way in which the site effect is incorporated into the assessment of the resistance demand. Previously it was introduced at the stage of evaluating seismic shaking exceedance probability by including, within the adopted ground motion prediction model, a site factor dependent on a qualitative classification of the soil type or the average of S-waves velocity in the first 30 meters of subsoil (VS30). In this work the site effect is introduced by means of a site-specific amplification factor expressed in terms of Arias intensity. An average value of this amplification factor and the relative standard deviation were calculated for multiple seismic inputs representative of relevant seismic scenarios by using 1D lithostratigraphic modeling of the site response. For this modeling data deriving from seismic microzonation studies were used with a major role played by advanced analysis of ambient noise recordings, while neglecting topographic effect, which was assumed to have a lesser influence. Average and standard deviation of the amplification factors were then used to modify the exceedance probability of different level of Arias intensity expected for unamplified site conditions. As further refinement, to calculate the slope resistance demand to seismic shaking, a new simple empirical relationship was used to estimate the Newmark displacement, which was specifically calibrated on data from the study area. This relationship proved to have better predictive capability than other more general relationships calibrated on data from different regions, even those characterized by a more complex functional form. While, however, the use of more specific predictive equation of Newmark’s displacement was found to have a limited influence on the assessment of the resistance demand, incorporating a site-specific amplification factor produced a significant increase. Compared to using VS30-based amplification factors, those obtained by incorporating a site-specific value calculated using the procedure tested in this work led to values greater by a factor of up to 4 and the resulting resistance demand was found to increase by a factor of up to about 3. The topography-related effect was also investigated at a site in the study area using 2D modeling of the site response. The test results confirmed – at least for the site analyzed in this work – that the pure topographic effect produces a lower effect, increasing shaking energy at most by 60%.