Search

SUMMARYThe scaling of earthquake parameters with seismic moment and its interpretation in terms of self-similarity is still debated in the literature. We address this question by examining a worldwide compilation of corner frequency-based and elastic rebound theory (ERT)-based fault slip, area and stress drop values for earthquakes ranging in magnitude from −0.7 to 7.8. We find that corner frequency estimates of slip (and stress drop) scale differently than those inferred from the ERT approach, where the latter deviates from the generally accepted constant stress drop behaviour of so-called self-similar scaling models. We also find that average slips from finite-source models are consistent with corner frequency scaling, whereas peak slip values are more consistent with the ERT scaling. The different scaling of corner frequency- and ERT-based estimates of slip and stress drop with earthquake size is interpreted in terms of heterogeneity of the rupture process. ERT-based estimates of stress drop decrease with seismic moment suggesting a self-affine behaviour. Despite the inferred heterogeneity at all scales, we do not observe a clear effect on the Brune stress drop scaling with earthquake size.

SUMMARY In this study, we investigate the rupture history of the 2009 April 6 (Mw 6.1) L’Aquila normal faulting earthquake by using a non-linear inversion of strong motion, GPS and DInSAR data. Both the separate and joint inversion solutions reveal a complex rupture process and a heterogeneous slip distribution. Slip is concentrated in two main asperities: a smaller shallow patch of slip located updip from the hypocentre and a second deeper and larger asperity located southeastwards along-strike direction. The key feature of the source process emerging from our inverted models concerns the rupture history, which is characterized by two distinct stages. The first stage begins with rupture nucleation and with updip propagation at relatively high (∼4.0 km s−1), but still subshear, rupture velocity. The second stage starts nearly 2.0–2.5 s after nucleation and it is characterized by the along-strike rupture propagation. The largest and deeper asperity fails during this stage of the rupture process. The rupture velocity is larger in the updip than in the along-strike direction. The updip and along-strike rupture propagation are separated in time and associated with a Mode II and a Mode III crack, respectively. The comparison between the source models inferred in this study with the Poisson ratio anomalies in the crustal volume containing the fault plane allows the interpretation of the delay in along-strike rupture propagation in terms of a structural control of the rupture history. Our results show that the L’Aquila earthquake featured a very complex rupture, with strong spatial and temporal heterogeneities suggesting a strong frictional and/or structural control of the rupture process.

SUMMARY In this study, we aim to understand the dependence of the critical slip weakening distance (D c) on the final slip (D tot) during the propagation of a dynamic rupture and the consistency of their inferred correlation. To achieve this goal we have performed a series of numerical tests suitably designed to validate the adopted numerical procedure and to verify the actual capability in measuring D c. We have retrieved two kinematic rupture histories from spontaneous dynamic rupture models governed by a slip weakening law in which a constant D c distribution on the fault plane as well as a constant D c/D tot ratio are assumed, respectively. The slip velocity and the shear traction time histories represent the synthetic ‘real’ target data which we aim to reproduce. We use a 3-D traction-at-split nodes numerical procedure to image the dynamic traction evolution by assuming our modelled slip velocity as a boundary condition on the fault plane. We assume a regularized Yoffe function as source time function in our modelling attempts and we measure the critical slip weakening distance from the inferred traction versus slip curves at each point on the fault. We compare the inferred values with those of the target dynamic models. Our numerical tests show that fitting the slip velocity functions of the target models at each point on the fault plane is not enough to retrieve good traction evolution curves and to obtain reliable measures of D c. We find that the estimation of D c is very sensitive to any small variation of the slip velocity function. An artificial correlation between D c/D tot is obtained when a fixed shape of slip velocity is assumed on the fault (i.e. constant rise time and constant time for positive acceleration) which differs from that of the target model. We point out that the estimation of fracture energy (breakdown work) on the fault is not affected by biases in measuring D c.

SUMMARY We measure crustal anisotropy parameters from several hundreds of aftershocks (M L > 2.5) of the 1997 Umbria‐Marche seismic sequence which occurred in a carbonatic fold and thrust belt in the shallow crust of central Apennines (Italy). The analysis of shear wave polarization shows clear S-wave splitting with prevalent fast direction ∼140 ◦ N and average delay times of 0.06 s. The observed fast direction is parallel to the strike of the activated normal-fault system and to the maximum horizontal stress (σ 2) active in the region. This is explained by the presence of stressaligned microcracks or stress-opened fluid-filled cracks and fractures within the sedimentary coverage, even if the role of structural anisotropy cannot be completely ruled out since the maximum horizontal stress is subparallel to the major structural features of the area (main thrusts and normal faults). The peculiar spatio-temporal evolution of the seismic sequence gives us also the opportunity to investigate temporal variations of anisotropic parameters. We analyse those seismograms whose ray paths sample the crustal volume containing two of the major fault zones, before and after the occurrence of normal faulting mainshocks (M w > 5). We observe variations of the anisotropic parameters during the days before and after the occurrence of mainshocks and we interpret them in terms of temporal variations of anisotropic parameters. This interpretation is consistent with temporal variations of the local stress condition and of the fluid pressure in the studied crustal volume proposed in the literature. However, since the spatial sampling of the selected ray paths varies with time, we cannot exclude the contribution of spatial variations of anisotropic parameters.

SUMMARY We model the time histories of coseismic stress changes produced by an earthquake rupture on an extended fault in a layered elastic half-space. We compute the stress perturbations caused by earthquakes using the discrete wavenumber and the reflectivity methods. We investigate the influence on the computed Coulomb stress changes of the adopted crustal velocity model, the rupture history and slip duration as well as the poro-elastic model assumed to calculate the pore-pressure changes. The comparison between the spatio-temporal evolution of the induced stress field, calculated for different crustal models and rupture histories, allows us to assess the depth and frequency dependence of the induced stress perturbation and the relative weight of dynamic and static stress changes. Our results show that the rise time and the rupture directivity are more important for characterizing the simulated stress time histories than the details of the rupture history. The adopted poro-elastic model affects the static stress changes more than the transient stress perturbations. The results of this study should be taken into account to model fault interaction through elastic stress transfer.

Summary We study the dynamic initiation, propagation and arrest of a 2-D in-plane shear rupture by solving the elastodynamic equation by using both a boundary integral equation method and a finite difference approach. For both methods we adopt different constitutive laws: a slip-weakening (SW) law, with constant weakening rate, and rate- and state-dependent friction laws (Dieterich–Ruina). Our numerical procedures allow the use of heterogeneous distributions of constitutive parameters along the fault for both formulations. We first compare the two solution methods with an SW law, emphasizing the required stability conditions to achieve a good resolution of the cohesive zone and to avoid artificial complexity in the solutions. Our modelling results show that the two methods provide very similar time histories of dynamic source parameters. We point out that, if a careful control of resolution and stability is performed, the two methods yield identical solutions. We have also compared the rupture evolution resulting from an SW and a rate- and state-dependent friction law. This comparison shows that despite the different constitutive formulations, a similar behaviour is simulated during the rupture propagation and arrest. We also observe a crack tip bifurcation and a jump in rupture velocity (approaching the P-wave speed) with the Dieterich–Ruina (DR) law. The rupture arrest at a barrier (high strength zone) and the barrier-healing mechanism are also reproduced by this law. However, this constitutive formulation allows the simulation of a more general and complex variety of rupture behaviours. By assuming different heterogeneous distributions of the initial constitutive parameters, we are able to model a barrier-healing as well as a self-healing process. This result suggests that if the heterogeneity of the constitutive parameters is taken into account, the different healing mechanisms can be simulated. We also study the nucleation phase duration Tn, defined as the time necessary for the crack to reach the half-length lc. We compare the Tn values resulting from distinct simulations calculated using different constitutive laws and different sets of constitutive parameters. Our results confirm that the DR law provides a different description of the nucleation process than the SW law adopted in this study. We emphasize that the DR law yields a complete description of the rupture process, which includes the most prominent features of SW.

SUMMARY In central Mexico some significant normal faulting events have occurred within the subducted oceanic Cocos plate, just below or near the down-dip edge of the strongly coupled interface. These normal faulting shocks followed large shallow thrust earthquakes. In other subduction zones such events generally precede the up-dip thrust events. A vertical 2-D finite element modelling has been used to simulate the subduction of the Cocos plate beneath the North American plate when the slab is driven by an active convergence velocity or slab pull. We find that the latter mechanism plays only a minor role due to shallow subduction. The modelling results show that the stress pattern is very sensitive to the geometry of the plates. In particular, normal faulting earthquakes that follow large thrust events can be explained on the basis of the flexural response of the overriding and subducting plates to the peculiar geometry of this subduction zone, where the subducting slab becomes horizontal at about 100 km from the trench. This horizontal part of the subducting plate, down-dip with respect to the main thrust zone, is under an extensional stress field. This provides an alternative explanation to the slab pull for the occurrence of normal faulting intraplate earthquakes. In order for normal faulting earthquakes to occur in the early part of the seismic cycle, it is necessary that the large up-dip thrust events have a partial stress drop. We find that for small fractional stress drop, a wide region of extension remains below the down-dip edge of the main fault plane following a large thrust earthquake. Thus, the main thrust earthquakes do not invert the polarity of the active stress field, which is compressional and extensional up-dip and down-dip, respectively, with respect to the main thrust fault. Larger fractional stress drops result in larger delays in the occurrence of normal faulting events after the main thrust events.

SUMMARY We propose a dislocation model for the two normal faulting earthquakes that struckthe central Apennines (Umbria^Marche, Italy) on 1997 September 26 at 00:33 (Mw 5.7) and 09:40 GMT (Mw 6.0). We ¢t coseismic horizontal and vertical displacements resulting from GPS measurements at several monuments of the IGMI (Istituto Geogra¢co Militare Italiano) by means of a dislocation model in an elastic, homogeneous, isotropic half-space. Our best-¢tting model consists of two normal faults whose mechanisms and seismic moments have been taken from CMT solutions; it is consistent with other seismological and geophysical observations. The ¢rst fault, which is 6 km long and 7 km wide, ruptured during the 00:33 event with a unilateral rupture towards the SE and an average slip of 27 cm. The second fault is 12 km long and 10 km wide, and ruptured during the 09:40 event with a nearly unilateral rupture towards the NW. Slip distribution on this second fault is non-uniform and is concentrated in its SE portion (maximum slip is 65 cm), where rupture initiated. The 00:33 fault is deeper than the 09:40 one: the top of the ¢rst rupture is deeper than 1.7 km; the top of the second is 0.6 km deep. In order to interpret the observed epicentral subsidence we have also considered the contributions of two further moderate-magnitude earthquakes that occurred on 1997 October 3 (Mw 5.2) and 6 (Mw 5.4), immediately before the GPS survey, and were located very close to the 09:40 event of September 26. We compare the pattern of vertical displacements resulting from our forward modelling of GPS data with that derived from SAR interferograms: the ¢t to SAR data is very good, con¢rming the reliability of the proposed dislocation model.