PSI - Issue 84

Mirza Adeel Zeb et al. / Procedia Structural Integrity 84 (2026) 248–255

252

flatareas (0–10°) and mountainous areas (>10°). Of the 250 bridges considered, 183 are located inflat areas and 67 in mountainousregions.Subsequently, to account for the different spatial influence of landslide processes, buffer zones of 300 m for flat areas and 500 m for mountainous areas were applied. These distances were selected following a reasoned and precautionary approach, aimed atidentifyingthe largest possible number of bridges potentially affected by landslide-related issues.Based on this approach, 44 bridges were identified as potentially exposed, including 20 bridges in flat areas and 24 in mountainous areas. The analysis was performed by importing the IFFI vector dataset into a Geographic Information System (GIS) environment using QGIS. Landslides were classified according to their documented activity status and assigned numerical values on a semi-quantitative scale: active landslide received 5, inactive ones a score of 3, and stabilized 1, following the framework reported in the Italian guidelines MIMS (2022). The analytical workflow has been implemented by performing a spatial join operation in which each of the 250 bridge locations has been associated with the nearest IFFI landslide polygon. Through this operation, each bridge was assigned a corresponding P a score based on the activity state of the closest documented landslide, as illustrated in Fig. 2 a. Applying the adopted criteria, theoutcomesindicatethat 10% of landslides are classified as active, 88% as dormant, and 2% as stabilized. The results further reveal a higher relative concentration of active and potentially active landslides in mountainous areas, confirming the strong control exerted by slope steepness and geomorphological setting on landslide dynamics. 3.4. Landslide magnitude (P m ) assessment The parameter landslide magnitude ( P m ) has also been derived from the IFFI inventory. The main limitation of this dataset is that it provides information on landslide area ( A L ) but not on the corresponding volume. To address this, a well-established and scientifically validated empirical relationship proposed by Guzzetti et al. (2009) was employed to estimate landslide volume ( V L ) as a function of area. This approach enables a consistent and reproducible estimation of potential landslide magnitude, even in the absence of direct volume measurements. The use of recognized empirical model ensures that the analysis remains objective and data-driven, which is a critical requirement of this study. The estimated volume ( V L ) has been then classified into five categories; each assigned a score ranging from 1 to 15 according to the Italian Guidelines (2022). With higher values corresponding to larger landslide volume classes defined by specific VL threshold intervals. The result shows that the landslide magnitude ( P m ) in the study area varies from small (10 2 < M ≤ 10 4 ) to extreme-large (M > 10 6 ) (Fig. 3). It is worth noting that bridges located outside the defined buffer zones were assigned the minimum Pm value, ensuring a precautionary assessment while maintaining consistency within the overall methodology.

(a)

(b)

Fig. 3. (a) Landslide magnitude ( P m ) assessment map; (b) Landslide magnitude ( P m ) assessment graph.