Automatic landslide prioritization at regional scale through PS-InSAR cluster analysis and socio-economic impacts

Abstract

Listen

Translate article

The present study is part of an initiative to revise the PAI landslide inventories for post-seismic reconstruction in the Central Apennines (Italy), affected by the 2016-2017 seismic sequence. We introduce an innovative automated workflow for classifying and prioritizing landslide-prone areas using multi-sensor Persistent Scatterers (PS). This approach aims to support post-seismic reconstruction efforts by government institutions through a comprehensive and up-to-date ranking of landslides. The methodology integrates post-seismic multi-sensor interferometric datasets, advanced clustering techniques, and socio-economic factors to establish a standard procedure for monitoring hazardous areas and prioritizing resource allocation. Key findings demonstrate the efficacy of the method in identifying critical regions that require immediate intervention, thereby enhancing the effectiveness of field inspections and territorial planning activities. Our multi-sensor analysis reveals that approximately 6% of landslides are not detectable by interferometric technique, 45% show stability with no detected deformation via PS data, and 19% are accurately mapped with deformation confined within their boundaries. Notably,30% of analyzed landslides exhibit displacement beyond their mapped perimeters, indicating potential expansion or underestimation of their extent. The integration of data from multiple sources provides a more robust understanding of potential risks across the study area, promoting a structured approach to hazard management that ensures interventions are both scientifically grounded and economically justified.