This indicates that the system does not depend critically on details, and that the inversion is robust for the damping we have chosen. (1995) and Hitchcock et al. We explore spatial variations in locking depth in Section 4.2. South of the San Bernardino Mountains, the total slip on the Indio SAF and SJF is somewhat larger than expected, 38 mm yr-1 as opposed to 35 mm yr-1, perhaps because our fault-model simplifications lead to the concentration of slip on these two faults. This value is comparable to the uncertainty in the GPS data, with 56 and 90 per cent of our residuals smaller than 2 mm yr-1 and 4 mm yr-1, respectively. Hence, we include a priori information about strain localization at known geological structures and steep velocity gradients across faults (cf. - Well-constrained fault - Moderately constrained fault - Inferred fault This manuscript benefited from detailed and constructive reviews by Rick Bennett and Peter Clarke. Crook R.J. Allen C.R. It also explains why the same earthquake can shake one area differently than another area. That places fault movement within the Quaternary Period, which covers the last 2.6 million years. (a) Binned and summed moment tensors on a 0.1 0.1 grid (every other data point shown) interpreted as strain rate . Concealed fault zones or fault trend zones formed in the cap rocks of sedimentary basin, which is influenced by the regional or local stress field, and activities in the basement rift system. The mean 1s uncertainty on the individual horizontal-component velocities is 1 mm yr-1. Bonkowski M.S. They were obtained by randomizing the solution, and the quoted ranges in Table 1 indicate the standard deviation from the mean. If we assume that all earthquakes have 5 meters (5000 millimeters) of slip, we will have earthquakes on average every 150 years: 5000 millimeters divided by 33 millimeters per year equals 150 years. More than 250 structures throughout the United States have been outfitted with seismic. Sieh et al. However, between = 0 and 1 there is a clear improvement in model fit to stresses, while the fit to velocities only deteriorates slightly. This result implies that interseismic loading and seismicity appear to be correlated over the lengthscales and timescales we have studied in our model. Fig. The scale for slip rates [different for (a) and (b)] and residual velocities is indicated along with the mean residual velocity vector length, v, and the component-wise mean misfit (in brackets). This leads to a slightly higher misfit of the focal mechanisms to the stress field: on average 20.5 in rake, compared with 19.5 for a model with no smoothing. Fig. The misfit to the GPS velocities is slightly larger for this joint model, with v> 2.3 mm yr-1 (2v= 3666, compared with the = 0 result (#x3008;v> 2.1, 2v= 3082). Below are charts from an L3 Harris letter. This movement may occur rapidly, in the form of an earthquake - or may occur slowly, in the form of creep. 1) takes up about 18 mm yr-1 of right-lateral slip, with the rest distributed on the western faults, including the Mojave segment of the SAF. 2003), fault reorientation over geological timescales, and viscoelastic effects (e.g. Shaded circles are plotted at the centre of each dislocation patch and scale with the slip rate. During the inversion, we allow block L to readjust the reference frame by treating the long-term block motion, L, as a free parameter (see Section 2.3.1 and Tables A1 and A2). Data source: USGS. This corresponds to a rotation pole and rate of 143.04E/-66.58N 0.02 Myr-1 in geographic coordinates, where x, y, and z are axes at 0E/0N, 90 E/0N, and the geographic North pole, 90N, respectively. Faults may range in length from a few millimeters to thousands of kilometers. Residual GPS velocities v and predicted fault slip rates for a joint inversion of GPS and stress data, = 1 (compare with Fig. 2000; Schroeder et al. Decreasing formal uncertainties therefore do not necessarily mean a better solution, but we prefer the damped inversion as it has a smaller model norm. Fine lines denote the fault trace of the Landers 1992 event (243.5E/34.3N, see Figs 2 and 11). We therefore have nGPS= 533 velocity observations, with ?GPS=nGPS 2 horizontal components (no vertical motions are included in the SCEC model). Is one available in GIS format? Sieh & Jahns 1984; Rockwell et al. 5) and joint ( = 1, Fig. (1996) and our block model, Fay & Humphreys found higher slip rates along the SAF Indio segment than along the SJF. Our method is different in that we include stress data for the first time and use a different parametrization. Misfits for this model are 2v= 3110, 2= 17 402, and , compared with for the simpler geometry as shown in Fig. Faults can extend deep into the earth and may or may not extend up to the earth's surface. Bigger earthquakes have more and larger aftershocks. 5) and joint inversion ( = 1, Fig. Why are there so many earthquakes and faults in the Western United States? Shading of the extensional vectors scales with the mean horizontal strain rate, , positive values indicating extension. Shaw J. Suppe J. Huftile G.J. 1994), PAC-NAM, and have larger amplitudes (Table A2). Flesch et al. Compare the relative in the Table (amplitudes are in ? Our choice of fault locations was primarily guided by mapped surface traces along the major strands of the SAF system (after Jennings 1975). Restricting the inversion to GPS data (= 0), we find that there is a broad minimum of 2 for locking depths of 9 km. 8, which shows 2v and 2t as functions of the stress weighting parameter . High may be associated with fault zones that have a greater proportion of velocity-strengthening material (Marone et al., 1991; Perfettini & Avouac, . 1986; Savage & Lisowski 1998), to name a few. A concealed fault zone is characterized by strong concealment and is associated with dominant fault zone. The typical along-strike length and down-dip width of each patch were 6 km 15 km; further lateral refinement (or coarsening) did not affect the results significantly. Tags . Illustration of the Savage & Burford (1973) block modelling method. We typically use locking depths as inferred by hand from the depth of seismicity in the Hauksson (2000) catalogue, but now explore the variation of model misfit as a function of dl. Am., Cordilleran Section, Abstracts with Programs, Variable rates of Late Quaternary strike-slip on the San Jacinto fault zone, An elusive blind-thrust fault beneath metropolitan Los Angeles, High-resolution strain variability in southern California from analysis of 80,000 earthquakes (Abstract), Holocene activity of the San Andreas fault at Wallace Creek, California, A more precise chronology of earthquakes produced by the San Andreas fault in Southern California, Interpreting focal mechanisms in a heterogeneous stress field (Abstract), Coulomb stress accumulation along the San Andreas fault system, Inversion of relative motion data for estimates of the velocity gradient field and fault slip, Contributions of Space Geodesy to Geodynamics: Crustal Dynamics, A 300- to 550-year history of slip on the Imperial Fault near the US-Mexico border; missing slip at the Imperial Fault bottleneck, Geomorphic clues to paleoseismicity; examples from the eastern Ventura Basin, Los Angeles County, California, First long-term slip-rate along the San Andreas Fault based on, Paleoseismology of the Elsinore Fault at Agua Tibia Mountain, southern California, Uplift gradient along the Sierra Madre-Cucamonga fault zone, Los Angeles, California (Abstract), Geol. This movement may occur rapidly, in the form of an earthquake - or may occur slowly, in the form of creep. 5, 2=v2= 3082 (VR= 91.5 per cent, ), which is substantially smaller than the misfit we obtain for rigid-block motions without any strain accumulation, namely . A fault is a fracture or zone of fractures between two blocks of rock. Stein S.. Deng J. Gurnis M. Kanamori H. Hauksson E.. Dolan J.F. The California Emergency Management Agency will then issue an advisory based on scientists' recommendations. The rigid-body rotation we determined for L from the SCEC data away from known faults before the inversion based on GPS sites NEED, 0809, and 0801 is Lr= (-0.007, 0.005, -0.02) Myr-1 in a Cartesian system. There are three main types of fault which can cause earthquakes: normal, reverse (thrust) and strike-slip. There are large uncertainties in and off-diagonal entries in C if we do not damp the solution, indicating significant trade-offs between individual i Euler vectors. To explore the dependence of model results on block geometry, we show as an example strike-slip rates for a = 1 joint inversion with an alternative block geometry around the San Bernardino mountains (Fig. This section describes how earthquakes happen and how they are measured. Further exploration of the model's successes and, more interestingly, its failures seems promising. Comparison of the stress inversion in Fig. Here, 1 and 3 denote the largest and the smallest eigenvalue of t, respectively, with tension taken positive. If you continue to use this site we will assume that you are happy with it. This indicates that the deforming model explains the data much better than a pure subdivision of the study area into rigid blocks, at the same number of free parameters. However, we defer a refinement of the fault geometry to subsequent work and discuss the robustness of our results with the help of an example of an alternative geometry in Section 4.5. (4); scale stress data to the amplitudes predicted initially by the block model; solve eq. The location below the earths surface where the earthquake starts is called the hypocenter, and the location directly above it on the surface of the earth is called the epicenter. consistent meanings. (8) by least-squares like eq. A Quaternary fault is one that has been recognized at the surface and that has moved in the past 1,600,000 years (1.6 million years). 1997), respectively; (7) Weldon & Sieh (1985); (8) and (9) Dokka & Travis (1990); (10) Sieh & Jahns (1984); (11) McGill & Sieh (1993); (12) Combination of Deep springs: 1 mm yr-1 normal (Lee et al. For the model in Fig. Using this method, much of the velocity field can be explained regionally, such as by introducing an Aegean subdivision to the Anatolian microplate (McClusky et al. What did government do for increase trade with other countries? Bonafede et al. We have also excluded some data from the dense GPS networks around Parkfield and Anza for a more uniform spatial coverage, and have removed those GPS stations that either we or Shen et al. Then the original earthquake is considered a foreshock. An extension of this approach, in which one solves for individual Euler vectors for each block, was applied to the eastern California shear zone by McClusky et al. Jennings 1975), there are large ambiguities involved in determining if faults are presently active or not. Most faults of this category show evidence of displacement some- time during the past 1.6 million years; possible exceptions are faults which displace rocks of undifferenti- ated Plio-Pleistocene age. We will assume that the stress inversion results of Fig. What are the physical hazards in the workplace? The new model, referred to as the third Uniform California Earthquake Rupture Forecast, or "UCERF" (http://www.WGCEP.org/UCERF3), provides authoritative estimates of the magnitude, locat, This poster summarizes a few of the more significant facts about the series of large earthquakes that struck the New Madrid seismic zone of southeastern Missouri, northeastern Arkansas, and adjacent parts of Tennessee and Kentucky from December 1811 to February 1812. Horizontal components of scaled stresses from an inversion including focal mechanisms after Landers (filled sticks, compare with Fig. Block models such as that of Meade et al. mouse-over each fault to get a pop-up window An online map of faults (Quaternary Fault and Fold Database of the United States) that includes California is in the Faults section of the Earthquake Hazards Program website. Soc. We use crustal velocities as provided by the SCEC Crustal Motion Map, version 3 (Shen et al.