摘要:

Although many studies have shown that faults and fractures are self‐similar over a large range of scales, none have tested the fault structure for self‐similarity in three dimensions. In this study, earthquake hypocentral locations in central and southern California were used to illuminate three‐dimensional (3‐D) fault structures, for which we measured the fractal capacity dimension,D0(3‐D). Hypocentral distributions from the Joshua Tree, Big Bear, and Upland aftershock sequences, as well as background seismicity at Parkfield were found to be fractal, whereD0(3‐D) increased with increasing event density, asymptotically approaching a stable value. The Joshua Tree data set stabilized atD0(3‐D) = 1.92 ± 0.02, the Parkfield data set asymptotically approachedD0(3‐D) = 1.82, and the Big Bear data set approachedD0(3‐D) = 2.01. As a test of the effects of location errors upon the measured value ofD0(3‐D), the Upland aftershock data were located with both the southern California Hadley and Kanamori (1977) (H‐K) velocity model, and the more accurate Hauksson and Jones (1991) (H‐J) velocity model. Events located with the H‐K model asymptotically approachedD0(3‐D) = 2.07, and events located with the H‐J model approachedD0(3‐D) = 1.79, suggesting that improved hypocentral locations may decrease the measured fractal dimension. One interpretation of our results ofD0(3‐D) ≤ 2.0 for all of the hypocentral data is that earthquakes only occur on the “percolation backbone” of a fault network, i.e., the active part of the network that accommodates finite strain deformation (Sahimi et al., 1993). We show that a percolation model that allows for healing of previously broken bond

ISSN:0148-0227    

DOI:10.1029/94JB02463

年代:1995

数据来源: WILEY

摘要:

We catalog and relocate Aleutian arc seismicity. Between 1957 and 1991, two great earthquakes ruptured the same 250‐km‐long portion of the central Aleutian arc: the 1957 Aleutian Islands earthquake and the 1986 Andreanof Islands earthquake. Because accurate estimates of the moment distribution of the 1957 earthquake are not available, the spatial distribution of aftershocks for each of these events is compared and tested against models describing the modes of occurrence of great subduction zone earthquakes. Earthquake relocations are based onPwave arrival times published in the International Seismological Summary, the Bureau Central International Seismologique, and the International Seismological Centre bulletins and include corrections for the near‐source velocity structure associated with the down‐going slab. Magnitude estimates are extracted from bulletins and prior to 1964 are estimated by us from microfilmed records. Our catalog is complete above magnitude 5.5. Aftershocks associated with the 1957 and 1986 earthquakes appear to occur in different areas. East of the main shock epicenters, aftershock locations are anticorrelated. West of the main shock epicenter, aftershocks of the 1986 earthquake tended to concentrate along the updip edge of aftershock clusters associated with the 1957 earthquake. If we assume aftershocks rim the distribution of seismic moment release associated with each event, these observations imply that the moment distribution of the 1986 earthquake was different from that of the 1957 earthquake. This suggests that we should use caution in identifying mechanically strong portions of a fault, asperities, by simply mapping the moment distribution of a single great earthquake. A fundamental tenet of the asperity model, that rupture always occurs on the strongest portions of the fault with weaker portions rupturing either aseismically or dynamically as a result of rupture on a strong fault patch, may in the case of the central Aleutian arc not be correct. Thus observing the moment distribution from a single great earthquake may tell us little about what the distribution of moment release will look like during the next ear

ISSN:0148-0227    

DOI:10.1029/94JB02641

年代:1995

数据来源: WILEY

摘要:

Scattering by heterogeneities is an important factor controling shapes of seismograms. Analyzing horizontal component seismograms in the band 1.0–8 Hz of local earthquakes in western Japan, we found that coda energy concentration just after theSwave arrival is more than expected for any multiple isotropic scattering model in uniform infinite medium. This behavior was seen even for deep earthquakes and became stronger with increase of frequency. To try to interpret this observation, nonisotropy is introduced into the multiple scattering model. The observed envelopes are compared with the model in order to estimate the amount of nonisotropy in scattering. Coda wave envelopes are synthesized using the Monte Carlo method for the nonisotropic scattering media. Two types of angular dependent scattering are considered as models of the nonisotropy with relation to the perturbation of seismic wave velocity in the lithosphere: perturbed media with exponential and Gaussian autocorrelation functions. It is found that forward scattering is dominant and the strength of sideways scattering is at least 10 times larger than that of backward scattering. The amount of the nonisotropic scattering means that correlation distance is estimated to be larger than several hundred meters in the case of Gaussian random medi

ISSN:0148-0227    

DOI:10.1029/94JB02064

年代:1995

数据来源: WILEY

摘要:

Relative source time functions (RSTF) have been estimated for four underground nuclear explosions and seven earthquakes in Central Asia using broadbandPwaveforms of nearby smaller events as empirical Green's functions (EGF). RSTFs of the four explosions (mb= 5.3 to 6.5) are each dominated by a simple pulse with a source duration of 0.4 to 0.8 s. RSTFs for two of the explosions show a significant secondary pulse with a pulse width similar to that of the first pulse. We conclude that the secondary phases are most likely associated with the spall slapdown phenomenon. Seismic moment releases by the spall phases are less than one third of those by the first explosion pulses. Elastic radii of the explosions are estimated to be about 0.25 to 0.5 km, and stress drops of the explosions range from 13 to 52 MPa. In contrast, RSTFs of earthquakes studied (mb= 5.5 to 6.6) typically comprise multiple source pulses with a total source duration from a few to several tens of seconds, indicating that the complex source process involves a fault dimension of several tens of kilometers. Stress drops of the earthquakes are much smaller than those of the nuclear explosions, ranging from 0.5 to 3.5 MPa. Our study demonstrates the power of the EGF method for retrieving RSTFs and reveals that differences in RSTFs and source parameters can be used to distinguish large nuclear explosions from moderate to large earthquakes (mb≥ 5.5

ISSN:0148-0227    

DOI:10.1029/94JB02465

年代:1995

数据来源: WILEY

摘要:

TheM7.4 Landers earthquake triggered widespread seismicity in the western United States. Because the transient dynamic stresses induced at regional distances by the Landers surface waves are much larger than the expected static stresses, the magnitude and the characteristics of the dynamic stresses may bear upon the earthquake triggering mechanism. The Landers earthquake was recorded on the UPSAR (U.S. Geological Survey Parkfield Small Aperture Array) array, a group of 14 triaxial accelerometers located within a 1‐square‐km region 10 km southwest of the town of Parkfield, California, 412 km northwest of the Landers epicenter. No triggered earthquakes were observed at Parkfield. Multiple filter analysis shows that the displacements, obtained by double integration, are dominated by the fundamental mode Love and Rayleigh modes, with some higher‐mode contributions for periods shorter than 10 s. Most of the surface waves propagated along the great circle path from Landers, but a late arriving surface wave appears to have been scattered from the Sierra Nevada Mountains. We used a standard geodetic inversion procedure to determine the surface strain and stress tensors as functions of time from the observed displacements. Peak dynamic strains and stresses at Earth's surface are about 7 μstrain and 0.035 MPa, respectively, and they have a flat amplitude spectrum between 2‐s and 15‐s period. These stresses agree well with stresses predicted from a simple equation using the ground velocity spectrum observed at a single station. Peak stresses ranged from about 0.035 MPa at the surface to about 0.12 MPa between 2 and 14 km depth, with the sharp increase of stress away from the surface resulting from the rapid increase of rigidity with depth and from the influence of mode shapes. Because of the free‐surface boundary conditions, the horizontal components of the stress tensor tend to dominate in the top 5–6 km of the crust, which might cause triggered seismicity to have strike‐slip or normal mechanisms. Comparison of dynamic stresses induced by the Landers, Loma Prieta, and Petrolia earthquakes at a variety of sites indicates that the Landers stresses were not spectacularly larger than those induced by the other sources. Landers dynamic stresses were comparable to Coalinga static stresses at Parkfield. The effective strain caused by Landers at Parkfield, where no earthquakes were triggered, are the same amplitude as those at some sites in Nevada where earthquakes were triggered. Comparing various authors' observations of dynamic stresses, there is no obvious characteristic of these stresses that correlates with the trig

ISSN:0148-0227    

DOI:10.1029/94JB02477

年代:1995

数据来源: WILEY

摘要:

Three high‐resolution seismic reflection lines were acquired in the northern part of the San Jacinto graben. The graben, a pull‐apart basin formed by a dilational right step of the San Jacinto fault zone, has been previously interpreted as a simple rhombochasm. The reflection survey located at least one significant and previously unidentified intrabasin fault, referred to here as the Farm Road strand. This fault lies approximately halfway between the Claremont and Casa Loma strands of the San Jacinto fault zone. At the north end of the basin, the southwestern boundary of the graben is interpreted to be the newly identified Farm Road strand and not the Casa Loma strand as was previously thought. The identification of this intrabasin fault allows us to infer that the San Jacinto basin comprises coalescing subbasins and is not a simple pull‐apart basin with an unusually large length:width ratio. The distances between the en echelon Casa Loma, Farm Road, and Claremont strands are between 1 and 2 km. This close spacing would likely permit an earthquake rupture to jump between strands and thus propagate through the San Jacinto

ISSN:0148-0227    

DOI:10.1029/94JB02469

年代:1995

数据来源: WILEY

摘要:

Applications of reflection tomography for the determination of complex geologic structures calls for the generalization of this method so that it can take triplications and other multiple arrivals into account. In this way, we propose a new formulation of travel time inversion. It relies on the choice of an adequate parametric representation of travel time information: the parameters we have chosen for this representation are the receiver location and the ray parameter at the receiver, some quantities directly measured from seismic data. The forward problem involved in the solution of this new inverse problem consists in shooting rays from a receiver according to the measured values of the ray parameter at the receiver. We can thus predict for a given model the emergence point of the reflected ray (i.e., the shot location) and the associated reflection arrival time. The least squares formulation of the inverse problem consists in minimizing an objective function that measures the mismatch between predicted and actual shot locations on one side and predicted and actual reflection arrival times on the other side, for the considered receiver locations and the associated measured ray parameters. However, inversion of noise corrupted kinematic data calls for a realistic definition of the uncertainties associated with the data. In particular, those uncertainties should take into account the sensitivity of reflection arrival times and shot locations to an error in the measurement of the ray parameter at the receiver. The objective function to minimize being chosen, the solution of the inverse problem is performed by a Gauss‐Newton method, the Jacobian of the forward modeling operator being computed by the adjoint state technique. It is interesting to note that no two‐point ray tracing is required in our method which is therefore cheaper than classical reflection tomography. The effectiveness of this approach is illustrated on a difficult synthetic example with large lateral velocity variations and strongly noise corrupted d

ISSN:0148-0227    

DOI:10.1029/94JB02461

年代:1995

数据来源: WILEY