Evaluation of the triggering potential of seismic landslides in Italy

Abstract

Landslides often occur as a consequence of natural hazards among which earthquakes are one of the main triggering factors. The effects of earthquake-induced ground shaking are often sufficient to cause the failure of slopes that were marginally to moderately stable before the earthquake. In this study, we define screening maps for Italy that classify sites in terms of their potentiality of triggering earthquake-induced landslides based on seismic hazard. To this end, we analyze seismic hazard maps and hazard disaggregation results on a national scale. First, as instabilities occur for acceleration values exceeding critical acceleration, we compare surface peak ground acceleration values derived from national hazard maps with critical acceleration thresholds proposed in the scientific literature. Then, magnitude-distance (M-R) scenarios from hazard disaggregation are analyzed in relation to upper-bound M-R curves for seismic landslide triggering. Landslide triggering can not be discounted if the value of the source-to-site distance R associated with magnitude M is lower than the reference upper-bound value and surface peak ground acceleration exceeds a given critical acceleration value. Most of the work concerns the analysis of hazard disaggregation results to define the controlling M-R scenarios. First, joint probability mass functions (PMFs) of magnitude and distance are analyzed to identify all modal scenarios (i.e., local maxima). To this end, we treat each PMF as an image and apply morphological image processing techniques to find local maxima. Specifically, the maximum (dilation) filter operation is applied. Local maxima are detected by checking for element-wise equality between the original and filtered matrices. Then, for each computation node, mean and modal M-R scenarios are compared to upper-bound M-R curves for earthquake-induced landslides selected from the scientific literature and the preferred M-R pair is selected as follows: if all M-R pairs stand above the reference upper-bound curve, then the triggering of earthquake-induced landslides can be neglected. if at least one M-R pair is below the reference upper-bound curve, then the triggering of earthquake-induced landslides can not be discounted. if more than one M-R pair lies below the reference upper-bound curve, then the triggering of earthquake-induced landslides can not be excluded and the M-R scenario that contributes the most to the hazard (i.e., the M-R pair with the largest PMF value) is selected as the preferred magnitude. As sites respond at specific characteristic frequencies (depending on local geological characteristics) and disaggregation results may vary with response period (T), the previous procedure is repeated considering disaggregation results associated with different spectral periods (i.e., spectral acceleration for different response periods). This allows us to define the controlling M-R pair for each site in relation to geological conditions (through site classification). The entire workflow is replicated for three types of landslides (disrupted slides and falls, coherent slides, and lateral spreads and flows), thus leading to three maps that show areas in Italy where the triggering of landslides due to seismic activity can not be excluded. The reliability of our results is finally checked by comparing them with observations of past seismic landslides in Italy.