Anomalous far-field geodetic signature related to the 2009 L’Aquila (central Italy) earthquake Simone Atzori, 1 Claudio Chiarabba, 1 Roberto Devoti, 1 Manuela Bonano 2 and Riccardo Lanari 2 1 Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Rome 00143, Italy; 2 Istituto per il Rilevamento Elettromagneti- co dell’Ambiente, CNR, Viale Diocleziano 328, Naples, 80124, Italy ABSTRACT The broad availability of geodetic measurements for the M w 6.3 April 6th 2009 L’Aquila earthquake allowed an unprece- dented description of the co- and post-seismic ground defor- mations, leading to the definition of the Paganica fault geometry and kinematics. Through DInSAR, we found, in a wide area of 20 kilometres on the Paganica hangingwall, a further displacement up to 7 cm, which might have occurred in the earthquake proximity. In this study, we explore the possibility of the co-, post- and pre-seismic alternative scenar- ios. Although our data are not sufficient to undoubtedly prove that this signal occurred before the main event, this seems to be the most likely hypothesis based on tectonics constraints and image acquisition times. The nature of this deformation remains unclear, but we speculate that deep fluids played a role. These results can drive ad hoc require- ments for future space-based missions and design of the GPS network. Terra Nova, 25, 343–351, 2013 Introduction The M w 6.3 L’Aquila earthquake occurred in the early morning of April 6th, 2009 and was the most destructive seismic event in Italy since the 1980 Irpinia earthquake. Despite its moderate magnitude, it had a strong resonance and political implications for the high number of casualties (308), the economic loss (more than 10 billion Euros, as assessed by the Italian Government for the European Commission) and the inestimable damage to the histor- ical and artistic heritage. This earthquake is the best docu- mented normal faulting event in terms of seismological, geological and geodetic data. Relocated after- shocks (Chiarabba et al., 2009a; Di Stefano et al., 2011) describe a com- plex fault system and the activation of nearby segments with M w > 5 earthquakes in the days that fol- lowed the main event (Fig. 1). Geo- detic data, from GPS to radar satellites, allow the identification of the fault extent, location and orienta- tion with unprecedented accuracy. The coseismic slip has a predominant normal dislocation on a ~16-km-long fault, with a maximum dislocation of ~1 m located at a depth of 7–8 km, few kilometres S–E from the main- shock hypocentre (Atzori et al., 2009a; Anzidei et al., 2009; Walters et al., 2009; Cirella et al., 2009; Che- loni et al., 2010; Cirella et al., 2012). Shortly before the earthquake, a change in V p /V s was observed by Lu- cente et al. (2010), and interpreted as dilatancy and opening of fluid-filled cracks. Recently, acceleration of slip prior to large earthquakes was hypothesized to explain sudden changes in tectonic tremors and repeating earthquakes (Shelly, 2009; Bouchon et al., 2011). However, the direct observation of an increased deformation before earthquakes has never been revealed by geodetic data until now. In this study, we show new results obtained by means of the satellite- based Differential Synthetic Aperture Radar (SAR) Interferometry (DIn- SAR) technique, focusing the atten- tion on the spatial and temporal characteristics of a subtle and broad deformation pattern that occurred in the fault hangingwall, relatively far from the epicentre. A large number of DInSAR displacement maps, adjusted to the reference frame of the local GPS network, are used to describe timing and origins of such deformation; additional DInSAR time-series are also exploited to catch possible signatures in time of the investigated signal. Although our data are not sufficient to undoubtedly prove the signal occurrence before the event, this seems to be the most likely hypothesis based on constraints imposed by tectonics and the image acquisition times. DInSAR Data Processing The DInSAR technique has gained a key role in geodesy during the last decades, thanks to its ability to detect even small deformations affecting very large areas, by comput- ing the phase difference (interfero- gram) between two temporally separated SAR images acquired over the same illuminated scene (Masson- net and Feigl, 1998; Franceschetti and Lanari, 1999). The displacement is measured in the ground-to-satellite direction (the line-of-sight, LOS here- inafter), which is inclined tens of degrees from the vertical and is therefore sensitive to the vertical and horizontal components of the retrieved displacement, although they can be discriminated only combining measurements from different LOSs (Manzo et al., 2006). The interferometric phase is sensi- tive to several factors: topography, deformation occurred during the observed time interval, signal delay due to the tropospheric and iono- spheric contents, Earth curvature, inaccuracy in the orbit information, instrumental artefacts and some other minor (Massonnet and Feigl, Correspondence: Simone Atzori, Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, 00143 Rome, Italy. Tel.: +39 0651860351; fax: +39 0651860351; e-mail: simone.atzori@ ingv.it © 2013 John Wiley & Sons Ltd 343 doi: 10.1111/ter.12040