Classification of the state of activity of Deep-seated Gravitational Slope Deformation phenomena of the Aosta Valley Region exploiting Sentinel-1 data processed by SAR tomography

Deep-seated Gravitational Slope Deformations (DsGSDs) are, both in terms of frequency and extension, one of the most widespread phenomena in the alpine territory. Their dynamic, although very slow may last for long period, affecting anthropic elements, representing a relevant natural hazard. Moreover, they may feature distinct deformation sectors, showing a spatial heterogeneity represented by typical morpho-structural lineaments variable rock mass fracturing, or presence of secondary landslides. A characterization of the state of activity of these huge phenomena is mandatory to evaluate their impact on anthropic elements. Leveraging on DInSAR techniques, a dedicated procedure to explore the behavior and define the state of activity of the 279 DsGSDs inventoried in the regional landslide inventory, i.e. IFFI Catalogue, of the Aosta Valley Region (Western Italian Alps) has been implemented. The proposed methodology consists in several steps. Firstly, Sentinel-1 data acquired over ascending and descending orbits have been processed through SAR tomography technique tailored to the processing of multipass SAR data at the full available spatial resolution for the analysis of Persistent Scatterers (PSs), i.e. ground targets which are concentrated in space (with respect to the wavelength and spatial resolution) and whose electromagnetic response is sufficiently coherent over all the observation period. Subsequently, an analysis of PSs dataset within DsGSD polygons, devoted to the assessment of Sentinel-1 data coverage has been carried out. In particular, this analysis involves the application of three indexes considering the PS abundance, the computation of voids in points distributions, and the assessment of PS clustering by the computation of skewness of their nearest neighbor distance. Three thresholds are set for each considered index leading to identifying cases with adequate points number and distribution for a suitable definition of the state of activity. Using post-processing algortims we projected the velocity values, measured along the Line of Sight (VLOS), along the steepest slope (VSLOPE) to compare DsGSD with different displacement direction, and the E-W and vertical components are also computed. A VSLOPE threshold has been set to distinguish stable and active phenomena, considering the aforementioned heterogeneities and other active phenomena, superimposed on the DsGSD area that could not be directly related to DsGSD evolution. Overall, the implemented methodology provides a valid instrument to rapidly and remotely define the state of activity of these huge phenomena, often wrongly underestimated or neglected in risk management, useful for a better definition of DsGSDs impacts on anthropic elements for a proper land use planning. This research was carried out in the ASI contract n. 2021-10-U.0 CUP F65F21000630005 MEFISTO, within which COSMO-SkyMed and SAOCOM data have been provided by ASI under license to use agreements.