Fault geometry, coseismic-slip distribution and Coulomb stress change associated with the 2009 April 6, Mw 6.3, L’Aquila earthquake from inversion of GPS displacements

Abstract

M w = 6.3 L'Aquila earthquake occurred within a complex system of NW-SE trending normal faults in the Abruzzi Central Apennines (Italy). We analyse the coseismic deformation as measured by >70 global positioning system (GPS) stations, both from continuous and survey-mode networks, providing unprecedented details for a moderate normal faulting earthquake in Italy from GPS measurements. We use rectangular, uniform-slip, dislocations embedded in an elastic, homogeneous and isotropic half-space and a constrained, non-linear optimization algorithm, to solve for the best-fitting rectangular dislocation geometry and coseismic-slip distribution. We use a bootstrap approach to investigate uncertainties in the model parameters and define confidence bounds for all the inverted parameters. The rupture occurred on a N129 +8 −7• E striking and 50 +9 −5• southwestward dipping normal fault, in agreement with geological observations of surface breaks along the Paganica fault. Our distributed slip model exhibits a zone of relatively higher slip (>60 cm) between ∼1.5 and ∼11 km depth, along a roughly downdip, NW-SE elongated patch, confined within the fault plane inverted assuming uniform-slip. The highest slip, of the order of ∼1 m, occurred on a ∼16 km 2 area located at ∼5 km depth, SE of the mainshock epicentre. The analysis of model resolution suggests that slip at depth below ∼5 km can be resolved only at a spatial scale larger than 2 km, so a finer discretization of different asperities within the main patch of coseismic-slip is not allowed by GPS data. We compute the coseismic Coulomb stress changes in the crustal volume affected by the major aftershocks, and compare the results obtained from the uniformslip and the heterogeneous-slip models. We find that most of the large aftershocks occurred in areas of Coulomb stress increase of 0.2-13 bar and that a deepening of the slip distribution down to a depth greater than 6 km in the SE part of the fault plane, in agreement with the inverted slip model, can explain the deepest, April 7, M w 5.3 aftershock.

Horizontal surface velocity and strain patterns near thrust and normal faults during the earthquake cycle: The importance of viscoelastic relaxation in the lower crust and implications for interpreting geodetic data

“…In the hanging wall, the direction of horizontal displacements changes twice with increasing distance from the fault surface break (i.e., from southwestward to northeastward and to southwestward again), which results in zones of shortening and extension, as observed in our model (Figure c). According to our experiments with different fault dips, the presence of a zone of extension indicates that the fault dips steeper than 45° (cf., Figure and Figure S1 in the supporting information), which is consistent with a dip of 50–55° derived from GPS data inversion [ Cheloni et al ., ; Serpelloni et al ., ] and the distribution of aftershocks [ Valoroso et al ., ]. GPS stations in both the hanging and footwall of the Chelungpu thrust fault indicate extension (Figure b), which is again in good agreement with our model results (Figures d and f).…”

“…Typically, coseismic displacement fields are calculated by subtracting the displacement fields some time before and after the earthquake from each other [e.g., Chen et al , ; Cheloni et al ., ]. By inversion of the geodetic data, the coseismic slip on the underlying fault plane at depth may be derived [e.g., Zhang et al , ; Serpelloni et al , ]. If the geodetic survey continues, information about the deformation after the main earthquake can be obtained, which typically includes aftershocks, afterslip on the ruptured fault plane, and flow in the viscoelastic layers of the lithosphere.…”

Control of pore fluid pressure diffusion on fault failure mode: Insights from the 2009 L'Aquila seismic sequence

“…Actually, Coulomb stress changes as low as 0.01 MPa are thought to be large enough to trigger earthquakes [see Harris, 1998, and references therein]. Moreover, several prior studies have in fact concluded that the aftershocks on the Campotosto fault could have been triggered by Coulomb stress changes [e.g., Serpelloni et al, 2012;De Natale et al, 2011].…”

“…Following Lucente et al [2010], the most important portion of the L'Aquila seismic sequence started with a foreshock of M W $ 4 that occurred on March 30, 2009, which is likely to have injected pressurized fluids in the main fault plane, consequently reducing its strength. A fluid-driven cascade of failures culminated with the M W 6.13 main shock of April 6, 2009;in Sibson's [2009] terms, the L'Aquila main shock was preceded, at least in the final stage of its precursory sequence, by hydrofracture dilatancy.…”

Waveguide effects in very high rate GPS record of the 6 April 2009, Mw 6.1 L'Aquila, central Italy earthquake

“…This mechanism is in agreement with the NE‐SW direction of the crustal extension in the Central Apennines, amounting to 2.5–3 mm/yr in a 40–50 km wide belt [ D'Agostino et al ., ]. The GPS‐derived static offsets [ Anzidei et al ., ; Cheloni et al ., ] and the deformation field obtained by InSAR (Interferometric Synthetic Aperture Radar) [ Atzori et al ., ; Walters et al ., ] have been used, separately [ Cheloni et al ., ; Cirella et al ., ; Serpelloni et al ., ] and jointly [ Trasatti et al ., ; Cirella et al ., ; D'Agostino et al ., ], to obtain the coseismic slip distribution on the causative fault. In addition, the kinematic rupture process of the earthquake was determined by a joint inversion of the geodetic results combined with the strong ground motion data [ Cirella et al ., , ] and with the high‐rate (10 Hz and 1 Hz) GPS (HRGPS) time series in the near field [ Avallone et al ., ].…”