Landslide-bridge interaction: Insights from an extensive database of Italian case studies

Despite the wealth of documented case studies, systematic approaches to correlate landslide characteristics with the damage they cause to bridges are rare. The correlation is challenging due to the complexity of landslides, which can vary in movement types, volume, velocities, materials, and orientations. Additionally, the lack of universally applicable models to forecast bridge responses in case of landslide interaction adds complexity. Recognizing the urgency of addressing this challenge, various countries, including Italy, have introduced guidelines and strategies to manage infrastructure risks and enhance safety. Efforts are underway to develop practical tools for authorities and infrastructure managers, encompassing factors influencing bridge response, especially under the action of natural hazards. This article presents a database of landslide-bridge interactions in Italy, developed under the FABRE Consortium. The database was compiled by analysing 382 bridges across 12 Italian regions. The article explores correlations between landslide characteristics and risk classification for bridges, defined as "Landslide Class of Attention" (L-CoA). The analysis shows that landslide volume is directly correlated with L-CoA severity, with larger volumes leading to higher classifications. Very slow-moving landslides are prevalent in high-risk L-CoA categories, suggesting they are associated with significant volumes and severe consequences. Complete interference between landslides and infrastructure poses the highest risk, while partial interference also contributes significantly. Combined landslides tend to result in more severe L-CoA classifications. The findings underscore the importance of better understanding the interactions between landslides and bridges, to develop predictive models and mitigate the risks posed by landslides to infrastructure in Italy and beyond.