Morpho-structural domains characterization by spaceborne SAR data: application on two deep-seated phenomena in Aosta Valley (north-western Italy).

Abstract

Deep-seated Gravitational Slope Deformations (DsGSDs) are widespread phenomena across alpine arch. Despite the slow evolution, long-lasting deformation trend of these huge phenomena can represent a relevant geo-hazard, variably affecting human settlements and infrastructures. Given the complexity and spatial heterogeneity of these phenomena, DsGSDs behavior can feature distinct deformation sectors, highlighted by distinctive morpho-structural lineaments. To assess the internal variability in terms of kinematics, deformation trends and style of activity of a deep-seated phenomenon, a local scale analysis is needed. Notoriously, spaceborne radar interferometry have proven to be suitable to characterize ground deformation displacement of very-slow phenomena as DsGSDs, although DInSAR techniques application remain challenging, especially in mountain areas due to complex topography, abundant vegetation and snow cover. In this study, a methodology for the characterization of DsGSDs, exploiting Sentinel-1 dataset on both ascending and descending orbits, is proposed. The Sentinel-1 images are processed with the multi-resolution Component extrAction and sElection SAR- Detector (CAESAR -D), which allows increasing the monitored area density via a spatially variable multilook. Subsequently, operating in a GIS environment, a post-processing and a dedicated analysis of the obtained measured points is implemented. Morpho-structural domains were mainly defined on the basis of geomorphological criteria, leveraging on DEM derivative products (e.g., slope, aspect and hillshade), orthophoto analysis and taking in account the information available in the Italian Landslide Inventory (IFFI). For each recognized domain, firstly, an analysis on the PSs coverage was performed in order to identify the proper distribution and density of the SAR-derived measurement points for a correct definition of the state of activity. Then, we operated filtering the available SAR datasets from possible anomalous values mainly related to the slope orientation to the satellite line of sight (LOS), in order to obtain suitable dataset for the ground deformation analysis. Finally, the filtered measured points were interpolated with the Inverse Distance Weighting (IDW) technique, with the aim of produce diverse ground deformation maps depending on the orientation of the analyzed phenomenon. The combination of ascending and descending geometries allowed to obtain east-west and vertical components of velocity. The projection on the VLOS along the slope allowed to partially reduce the limitation of the topography on SAR sensitivity. This allowed us to analyze the displacement pattern of DsGSDs in more reliable way. We tested the procedure on two variably oriented DsGSDs phenomena, located in the alpine region Aosta Valley, the Croix de Fana DsGSD, mainly north-south oriented, and the Valtournenche DsGSD, mainly east-west oriented. The variations of the kinematic behavior between the morpho-structural sectors is detected, also considering any other phenomenon as secondary landslide or talus, superimposed on the DsGSD. Overall, the implemented methodology allows to a rapid and low-cost generation of ground deformation maps able to spatially analyze and characterize the morpho-structural domains of DsGSDs, providing an effective tool suitable for the definition of DsGSDs impact on the diverse anthropic elements and a proper land use planning in mountainous territories. This research was carried out in the framework of the ASI contract n. 2021-10-U.0 CUP F65F21000630005 MEFISTO