PSI - Issue 84

Alessandro Brunetti et al. / Procedia Structural Integrity 84 (2026) 97–102

99

Fig. 1. SGAM three-tier workflow for landslide hazard assessment of bridges and linear infrastructures following the Italian National Guidelines.

2.1. Hazard Assessment The SGAM framework implements the P Factor methodology as specified in the Italian guidelines, ensuring national compliance and methodological consistency. The framework was tested in a pilot area, with landslide inventories enriched with ancillary data compilation as summarized in Table 1. The P Factor calculation integrates three fundamental parameters: activity state (P a /P c ), velocity (P v ), and magnitude (P m ). For mapped landslides, the Pa parameter is assigned based on activity classification: active, inactive with movement signs, stabilized. When activity state information is not provided from the SGAM database, PS-InSAR velocity data is used for inference. For potential landslides identified through data-driven susceptibility analysis, the Pc parameter is assigned based on susceptibility class thresholds determined through ROC curve analysis and Gaussian Mixture Model classification. Integration of PS-InSAR data provides velocity assessment for both mapped and potential slow-moving landslides. Fast-moving landslides and rockfalls receive default maximum velocity scores. Volume estimation employs differentiated volume-area relationships based on landslide type. For potential landslides, clump analysis within slope unit constraints identifies contiguous high-susceptibility areas, ensuring spatial consistency between magnitude and velocity calculations. The P Factor = ( + + ) maps to initial hazard classes from low to high with score ranging from 5-25 points. Table 1. Predictor data for SGAM application. Input Dataset Topography DTM, Slope, Curvature, Relative Elevation, TWI Lithology National Lithological Map. 1:100.000 Land Use/Land Cover Corine / Regional data Seismicity PGA and Vs 30 A-DInSAR (PSI) Persistent Scatterers derived from S1 – CSK – SAOCOM