Application of DInSAR data for slow-moving landslides monitoring

Landslides are among the most important and most frequent natural calamities. After earthquakes, they cause the highest number of victims and of damages to man-built structures. Thus, landslides risk assessment and prevention is the key step to correct territory planning and management, as witnessed by growing scientific evidence in this field. Landslides' studies rely first of all on territory monitoring, essential for the detection of at-risk phenomena. Territory monitoring is a complex and dynamic process, which requires continuous technological adjustments in order to obtain the most precise and up-to-date information on natural events and on their evolution. Identification of the most efficacious monitoring system is difficult and it is based on the one that best meets technical, logistical and economic requirements of the local community and needing careful cost-benefits analyses. Owing to the high number of contributing factors, the optimization of monitoring processes is known to be quite complex. Integrated monitoring networks for the assessment of mass movements are thus being developed in order to obtain useful information on both triggering factors and magnitude of the studied phenomenon. Such networks also include technical and logistical aspects. Final comparison of the results obtained from the application of different methods allows to achieve their validation. The present study develops in the above-mentioned field. In detail, this work focuses on Remote Sensing, on its potential as an additional monitoring technique and as an alternative method, at least in preliminary phenomena evaluations, to traditional ground monitoring systems. Among the different types of Remote sensing techniques, one that very well meets dynamic processes' requirements is the Interferometry SAR (InSAR). This branch of Remote sensing uses active monitoring sensors that measure the backscattering radiation produced by the sensor-originated impulse directed to the Earth's surface. Such systems use radar antenna that are able to produce and receive electromagnetic signals; in detail, modern satellites are equipped with Synthetic Aperture Radar (SAR) sensors which can obtain good ground resolution. The images obtained from a SAR is composed by the amplitude which depends on the scene reflectivity and the phase, among others, is proportional to the two-way distance from satellite to ground. InSAR is the method used to process SAR generated images and it is based on the combination between one or more pairs of satellite images whose orbital parameters are all known. The combination of two SAR images of the same scene acquired from different orbits, i.e., incidence angles, produces an interferogram. The interferogram is obtained by multiplying one image by the complex conjugate of the other and contains, on a pixel by pixel basis, the phase difference between the two acquisitions (Massonet et al.,1993). This phase difference can be exploited in combination with the orbital information for each acquisition to derive a Digital Elevation Model (DEM) of the scene. If the latter is already known, InSAR allows to calculate the terrain displacement. The aim of this work is to evaluate the applicability of satellite interferometry monitoring in the center-southern Italian Appennine. This area is characterized by structurally complex geological formations that give rise to slow to extremely slow velocity deformative phenomena, which can be well studied with SAR-based monitoring technique. A preliminary, detailed study of the different SAR data processing techniques and available algorithms was made in order to detect limits and potentials of each one. Velocity maps and deformation time series were then elaborated in order to evaluate phenomena extension and activity and consequently to define the events' evolution. Finally, the obtained results were compared with ancillary data, where available, for validation. Ancillary data included inclinometer data, GPS, topographic measurements and rain data which, in some cases, allowed a better phenomena interpretation. Studied sites displayed representative phenomena and were sufficiently supplied with traditional monitoring data. Four experimental study sites were identified and are listed below (Fig. 1.2): - Calitri (Campania Region - Avellino Province), interested by the presence of a huge complex landslide, reactivated after the main shock of the M = 6.9 November 23, 1980 earthquake; - Moio della Civiltella (Campania Region - Salerno Province) where two urban centres, Moio and Pellare are interested by slide-flows and roto-translational slides; - Potenza, Costa della Gaveta zone (Basilicata Region) involved by many complex mass movements occurred on the northern slope of the Basento River valley; - Agnone (Molise Region - Isernia Province) where in the catchment of the Vallone S. Nicola, a complex slope movement, to be considered as a reactivation of a pre-existing landslide, took place as a consequence of an intense rainfall event occurred on January 2003. For each of these sites, available data have been elaborated by means of the processing procedures that best allowed objective data interpretation.