GNGTS 2016 - Atti del 35° Convegno Nazionale

GNGTS 2016 S essione 1.2 195 Geodetic evidence of a regional uplift in Northern Apennines (Italy) from integration of GPS and InSAR measurements L. Anderlini 1 , E. Serpelloni 1 , A. Pepe 2 , G. Solaro 2 , M. Polcari 3 , C. Bignami 3 , M. Moro 3 , S. Stramondo 3 , L. Chiaraluce 3 1 INGV, CNT, Bologna, Italy 2 CNR-IREA, Napoli, Italy 3 INGV, CNT, Roma, Italy Introduction. The northern Apennines are characterized by a complex structural setting, due to the tectonic and geodynamic evolution of the orogeny with the superposition of different processes, including a compressional phase, forming E-NE verging thrust and folds, and an extensional phase with chain-normal extension. This extensional phase is the clearest ongoing tectonic processes in the Apennines, being well documented by earthquake focal mechanisms (Pondrelli et al. , 2006) and geodetic deformation measurements (e.g., Anderlini et al. , 2016). Evidence for active compression is more elusive, although compressional crustal stresses are observed at the front of the northern Apennines (Montone and Mariucci, 2016) and GPS data provide fair indications of active shortening (e.g., Devoti et al. , 2008; Bennett et al. , 2012). In this work we combined Global Positioning System (GPS) three-dimensional ground velocities and Interferometric Synthetic Aperture Radar (InSAR) line-of-sight (LOS) ground velocities for the Umbria-Marche sector of the northernApennines. We made use of the accurate and precise, but sparser, GPS velocity estimates in order to adjust the InSAR velocities to a local GPS reference frame. This operation includes the estimate of offsets between the two geodetic measurements and of residual orbital ramps. The corrected InSAR velocities provide a first high- resolution image of relative vertical motions between the inner sector of the Apennines and the external sector, with the first one moving upward with respect to the second one, according to the GPS observations. This vertical deformation signal, characterized by a wavelength of 40 km (i.e., change between positive and negative values), has been used to develop an interseismic dislocation model, providing preliminary indications of active thrusting in this sector of the northern Apennines, coexisting with the well-known chain-normal extension. In this work we describe the data and methods used to combine InSAR and GPS velocity observations and describe the dislocation model used and its limitations. GPS and InSAR Data. The 3D GPS velocity field has been obtained analyzing data from several continuous GPS (cGPS) networks operating in the study area with an observation period longer than 2.5 years in the 1998-2015 time-span, following a three-step approach (see Serpelloni et al. , 2013 for more details). We refer GPS velocities with respect to the same reference point of InSAR data (Fig. 1a), in order to identify, and eventually minimize, any possible discrepancy between the two ground velocity measurement techniques. InSAR velocities, measured along the satellite line of sight, have been estimated from multi-temporal analysis of ENVISAT images, acquired along the ascending orbit, using the SBAS technique (Berardino et al., 2002). Topography removal using SRTM DEM has been applied; in addition the atmospheric phase screen is estimated and filtered out. For the sake of comparison, the InSAR time series have been referred to the same reference point and the GPS measurements, reproduced over the same time interval of ENVISAT acquisitions (2003-2010), have been projected along the radar line of sight. The two velocity datasets have been then compared projecting GPS velocities along the satellite line-of-sight (LOS) direction (same color palette in Fig. 1b), presenting a visible disagreement in many portions of the study region. These differences have been corrected estimating, and subsequently removing, a planar ramp from the InSAR velocity field (Fig. 1c), ascribable to a residual orbital error, using ground-based GPS velocities. In particular, the ramp and offset have been estimated in a least-squares sense by minimizing the differences between GPS LOS velocities and InSAR LOS velocities at nearby points,