Correlation of multiplatform sar-data for multitemporal slope instability analysis: the Paupisi case study (Benevento Province, southern Italy)
Keywords:PSInSAR data, slow-moving landslide, remote sensing techniques, state of activity map
The Benevento Province (southern Italy) has been historically affected by soil erosion and ground deformation (i.e. landslides; Revellino et alii, 2019), as testified by documents and reports which describe damage scenario, social effects, surveys and measures carried out. Revellino et alii (2010) highlighted as the landslide index, defined as the ratio of areas affected by landslides over total areas, reaches up to 90%. Earth flows make up about 46% of these landslides, involving structurally complex geologic formations and often responsible for damaging human infrastructures (e.g. roads and service lines; Guerriero et alii, 2016; Maresca et alii, 2022). As a result, the province has widespread problems in management of the landslide hazard, due to the activation (and reactivation) of phenomena connected to both rainfall and seismic events. In order to reconstruct state of activity maps for the Benevento Province, a multitemporal analysis of multiplatform satellite data was carried out, with a resolution of cells of 20×20 m. In particular, PSInSAR (ERS 1&2 for the time-span 1992-2000, ENVISAT for 2002-2010, and RADARSAT for 2003-2007) and ISBAS data (Sentinel-1 for 2017-2020), were first treated by a grid-based approach to uniform the PS data and subsequently, by a multiplatform approach defining weighted average velocity (VWA) maps on the base of the extent of the area with data availability. Finally, a multiplatform activity-matrix approach allowed the definition of landslide activity maps based on a velocity threshold of ± 3 mm/y, derived by VWA maps. Specifically, velocity threshold allowed to discriminate for each cell of the analyzed area i) stable cells (velocities between ± 3 mm/y), ii) unstable cells (velocities greater than ± 3 mm/y), iii) cells with no data (unclassified). The multiplatform activity-matrix approach at provincial scale was used to identify and analyze unstable area (hotspots), currently involved in active deformation processes. Among them, the Paupisi municipality (Benevento Province) was identified as critical for the involvement of built-up areas and infrastructures in landslide deformation processes. For that reason, Paupisi area was selected as study case and a detailed analysis of the PS distribution as well as reconstruction and evaluation of deformation trend were carried out. Finally, a relation between land deformation and rainfall events was also investigated.
How to Cite
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.