Download (21MB) | Preview
[error in script] [error in script]
Item Type: Tesi di dottorato
Resource language: English
Date: 10 December 2018
Number of Pages: 154
Institution: Università degli Studi di Napoli Federico II
Department: Scienze della Terra, dell'Ambiente e delle Risorse
Dottorato: Scienze della Terra, dell'ambiente e delle risorse
Ciclo di dottorato: 31
Coordinatore del Corso di dottorato:
Ascione, AlessandraUNSPECIFIED
Date: 10 December 2018
Number of Pages: 154
Settori scientifico-disciplinari del MIUR: Area 04 - Scienze della terra > GEO/04 - Geografia fisica e geomorfologia
Date Deposited: 11 Jan 2019 12:46
Last Modified: 23 Jun 2020 14:22

Collection description

The core of the study herein has been the analysis of PS-InSAR datasets aimed at providing new constraints to the active tectonics framework, and seismotectonics, of several regions of the Apennines. The analysed Permanent Scatterers datasets result from processing of large amounts of temporally continuous series of radar images acquired with the ERS (1992-2000), ENVISAT (2003-2010) and COSMO SKYMED (2011-2014) satellite missions. Such datasets, which are available in the cartographic website (Geoportale Nazionale) of the Italian Ministry of Environment (MATTM) have been collected through time by the MATTM in the frame of the "Extraordinary Remote Sensing Plan" (Piano Straordinario di Telerilevamento Ambientale, PST-A, law n. 179/2002 - article 27), with the aim of supporting local administrations in the field of environmental policies. The database was realized through three phases: the first one (2008-2009), which involved the interferometric processing of SAR images acquired throughout the country by the ERS1/ERS2 and ENVISAT satellites in both ascending and descending orbits, from 1992 to 2008; the second one (2010-2011) integrated the existing database with the processing of the SAR images acquired by the ENVISAT satellite from 2008 to 2010; the third phase (2013-2015) provided an upgrading and updating of the previously developed database on critical areas, based on StripMap H image acquired with a 16-day recurrence, either in ascending or descending orbit, using the Italian national satellite system, the COSMO SKYMED. With this study, a massive use of Permanent Scatterer datasets is applied for the first time at assessment of ground deformation of large (hundreds of km2 wide) regions of Italy over the last decades, in order to unravelling their current tectonic behaviour. To date in the field of tectonics – in particular, of earthquake geology - the SAR images have been used essentially through the DinSAR technique (comparison between two images, acquired pre- and post-event) in order to constrain the co- and post-seismic deformation (Massonet et al., 1993; Peltzer et al., 1996, 1998; Stramondo et al., 1999; Atzori et al., 2009; Copley and Reynolds, 2014), while the approach that has been used in the case studies that are the object of the research herein is based on analyses of data that (with the exception of the Lunigiana case study) cover an about 20-year long time window. The opportunity of analysing so long, continuous SAR records has allowed detection of both coseismic displacement of moderate earthquakes (i.e., the M 6.3 2009 L’Aquila earthquake, and the M 5 2013 Lunigiana earthquake), and subdued ground displacements - and acceleration – on time scale ranging from yearly to decades. The specific approach used in this study rests on a combination of various techniques of analysis and processing of the PS datasets. In general, as the analyses that have been carried out aimed at identifying motion values with wide areal extent, a statistical filtering has been applied to PSs velocity values in order to discard from the initial, “native”, dataset fast-moving PSs that may be associated with the occurrence of local-scale phenomena (e.g., landslides, sediment compaction, water extraction, etc.). Furthermore, an in depth inspection of time series of PSs from all of the investigated areas has been carried out with the aim of identifying changing (LoS-oriented) motion trends over the analysed time windows. A distinctive feature of this study was the estimation of vertical ground displacements. In fact, while most studies on ground deformation are based on analysis of SAR data recorded along either ascending or descending satellite orbits (thus based on LoS-oriented motions), a specific focus of this study was to obtain - starting from LoS-oriented PS velocity values - displacement values in the vertical plane oriented west-east. In order to evaluate vertical displacements, a geometrical relationship was applied to ascending - descending PSs pairs. As PS from ascending and descending tracks are neither spatially coincident nor synchronous, each image pair was obtained by selecting ascending-descending radar images with a time separation within one month. In the L’Aquila region case study, the combination of data recorded along both the ascending and descending satellite orbits has been crucial to the identification of pre-seismic ground motions, undetected in previous works that – similarly – had addressed assessment of possible pre-seismic satellite-recorded signals. In the various case studies, different kinds of GIS-aided geostatistical analyses were used to extract and synthesise information on ground deformation through the construction of both raster maps of displacement values for the ascending and descending LoS, respectively, and maps of the vertical (z, up - down) component of the “real” displacement vector. In the Campania plain case study, the PS-InSAR data analysis and processing have been integrated by detail scale geomorphological-stratigraphical analysis. Results of analyses of the two independent data sets are consistent, and point to tectonically-controlled ground displacements in a large part of the northern part of the study area (Volturno plain) during the 1992-2010 analysed time span. In particular, the integrated data sets show that the boundaries of the area affected by current subsidence follow fault scarps formed in the 39 ka old Campania Ignimbrite, while the horst blocks of such faults are substantially stable (or slightly uplifting) during the analysed time window. Furthermore, mean rates of current subsidence and long-term (Late Pleistocene to present) mean subsidence rates are comparable, pointing to current vertical displacement assessed through the PS-InSAR data analysis as the expression of the recent tectonics of the analysed sectors of the Campania plain. The Campania plain substantially lacks strong historical seismicity. Such evidence suggests that the detected surface displacements result at least in part from aseismic fault activity. The Monte Marzano case study has allowed assessment of subdued deformation along both the major structures that were activated with the Irpinia 1980 earthquake, i.e. the NE-dipping Monte Marzano fault and the SW-dipping Conza fault, respectively. Ground deformation associated with such structures appears decreasing from the time window covered by the ERS satellites (1992-2000) to that covered by the ENVISAT (2003-2010). These data suggest that post-seismic slip of the M 6.9 has continued until 20 years after the main shock to become very weak in the following ten years. Furthermore, the PS-InSAR data analysis has shown that wide areas located between the Monte Marzano and Conza faults (i.e., in the one that is recognised as the graben structure bounded by those structures) show uplift in the range of 0-2 mm/yr, more evident in the period surveyed by the ERS satellites (1992-2000) and less evident in the 2003-2010 time span (ENVISAT). Such uplift might be related to the occurrence, at depth, of a fluid reservoir that has been independently identified by seismic tomography (Amoroso et al., 2014). In depth analysis of pre-seismic periods have been carried out in three study areas, i.e. those of the 1997 Colfiorito earthquake, of the 2009 L’Aquila earthquake and of the 2013 Lunigiana earthquake. The Colfiorito case study has not provided any significant information on possible pre-seismic ground deformation, most probably due to the PS spatial distribution in that region too much discontinuous to allow identification of both net signals from inspection of the rare and sparse PS time series, and statistically meaningful surface displacement patterns. Both in the L’Aquila and Lunigiana case studies, ground deformation signals in the pre-seismic period have been detected from inspection of PS time series. Pre-seismic ground deformation signals detected in the Lunigiana area (which was affected by a strike-slip faulting earthquake; Eva et al., 2014, Pezzo et al., 2014, Stramondo et al., 2014) are questionable, as they are quite complex and difficult to be interpreted and framed within the local tectonic scenario. Conversely, very clear and net pre-seismic signals have been identified in the region hit by the L’Aquila normal faulting earthquake. There, in the time span predating of some four years the 6th April 2009 main shock, ground deformation with distinct spatial patterns, and orientations, have been detected. In particular, the PS-InSAR analysis has shown that the hanging wall block of the Paganica fault (the surface expression of the structure activated with the main shock; e.g., Galli et al., 2010) has been subject to slow uplift and eastward horizontal motion from 2005 to September/October 2008, and then (October 2008-March 2009) subject to subsidence and westward oriented horizontal motion. Following coseismic collapse, in the early post-seismic period (April-May 2009), subsidence extended eastwards beyond the Paganica fault trace. The region affected by opposite pre-seismic motions covers the area in which the 6th April main shock and most of both foreshocks and aftershocks (Valoroso et al., 2013) were recorded, while the inversion of the pre-seismic displacements is coeval with onset of the foreshocks (October 2008; Di Luccio et al., 2010). In addition, such a region includes both topographic highs and lows. All of such features point to a correlation of the detected motions with the seismic phenomena, and suggest a deep-seated causative mechanism, such as volume changes in response to vertical/lateral fluids migration and fracturing processes at depth, with all phenomena having been documented in connection with the 2009 earthquake in the study region (e.g., Di Luccio et al., 2010; Lucente et al., 2010; Moro et al., 2017). Pre-seismic ground deformation that has been detected in the L’Aquila region could represent a precursor signal of the 2009, M 6.3 earthquake. Such a hypothesis should be tested, in the future, through the continuous monitoring through SAR satellites, but also high-resolution geodetic techniques, of seismically active regions worldwide aimed at detecting the possible occurrence of pre-seismic signals. However, the results of this study point to the long-term (yearly scale) PS-InSAR technique as a tool crucial to the detection of ground deformation in areas struck by recent earthquakes, and to monitoring active – possibly aseismic - structures. Such knowledge may strongly support strategies addressed at territorial planning and mitigation of seismic hazard, and represent an important sustenance for actions ruled by Civil Protection. On the other hand, the results of this study highlight the importance of the existing PS database, and the importance of continuing implementing such an instrument in the future.


Downloads per month over past year

Actions (login required)

View Item View Item