Abstract: | We present two case studies regarding the application of Synthetic Aperture Radar (SAR) Persistent Scatterers Interferometry (PSI) techniques to landslide-prone slopes situated in the municipal territories of Caramanico Terme and Volturino (Italy). The analysis of satellite SAR data with PSI techniques poses often problems on sites where, due to the scarcity of human artefacts and the presence of vegetation cover, density of coherent points (PS) is low (< 10 per km2). Moreover, the steep and rough topography typical of landslide-prone areas hamper the interferometric pre-processing, making more difficult the joint estimation of displacements and of DEM errors. Under these conditions the significance of temporal interferometric phase trends can be uncertain and conservative assumptions, necessary to ensure low false detection probabilities, need to be coupled with innovative processing strategies to increase the detection efficiency of PS objects. Here, the SPINUA (Stable Point Interferometry over Un-urbanised Areas) processing technique is applied together with an alternative PS Candidate (PSC) selection procedure based on the use of pixels classified as urban. The cases of Caramanico and Volturino are representative, respectively, of harsh and favourable conditions for PSI applications. The results from Caramanico show clusters of PS exhibiting similar line-of-sight (LOS) deformation behaviour in the period 1995–2000. The locations of moving PS often coincide with distressed buildings and appear consistent with the areal distribution of recent and past landslide activity. The temporal displacement trends, however, are characterised by very low annual average velocities (from 3 to 7 mm/y) and it is uncertain to what extent the PS data reflect true slope movements, local deformations (e.g. settlement of engineering structures) or both. Thanks to the more favourable conditions, the application of the standard SPINUA approach in the Volturino area was sufficient to obtain suitable densities of PS, as well as spatially and temporally consistent displacement results for a period 1992–2000. In particular, a group of moving PS was identified in a peri-urban area, known for the past and recent slope stability problems. The slowly moving PS (from 3 to 5 mm/y) fall in a location that, unlike the remaining part of the town, is characterised by the presence of many distressed buildings and structures. Although the site information confirms the reliability of PS data, in the absence of ground monitoring and detailed records of landslide movements, it is difficult to identify the main mechanism of the detected deformations. In general, in geologically and topographically complex urban/peri-urban settings, the significance of very low-velocity PSI surface displacements should always be considered together with in situ geotechnical controls and ground monitoring data. |