摘要:

We extend the concept of probability gain associated with a precursor (Aki, 1981) to a set of precursors which may be mutually dependent. Making use of a new formula, we derive a criterion for selecting precursors from a given data set in order to calculate the probability gain. The probabilities per unit time immediately before four large Chinese earthquakes are calculated. They are approximately 0.09, 0.09, 0.07 and 0.08 per day for 1975 Haicheng (M= 7.3), 1976 Tangshan (M= 7.8), 1976 Longling (M= 7.6), and Songpan (M‐ 7.2) earthquakes, respectively. These results are encouraging because they suggest that the investigated precursory phenomena may have included the complete information for earthquake prediction, at least for the above earthquakes. With this method, the step‐by‐step approach to prediction used in China may be quantified in terms of the probability of earthquake occurrence. The lnPversustcurve (wherePis the probability of earthquake occurrence at timet) shows that lnPdoes not increase withtlinearly but more rapidly as the time of earthquake appro

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02185

年代:1983

数据来源: WILEY

摘要:

Source time is a kinematic fault parameter corresponding to the duration of seismic source time functions and is accurately determined from phase spectra of long‐period surface waves. Source times are determined for 36 large earthquakes during the last three decades. Scaling relations among source times, seismic moments, and fault dimensions are derived. Seismic moment is proportional to the cube of source time, and fault dimension is proportional to source time. Source times for low‐angle thrust earthquakes along deep‐sea trenches are found to be longer than those of other types, such as intraplate shocks and deep shocks. They are also significantly longer than the rupture times expected from a Haskell model, suggesting that, generally, there exists an introductory stage of faulting that precedes the main stage of the rupture propag

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02191

年代:1983

数据来源: WILEY

摘要:

The analysis of the New Brunswick earthquake of January 9, 1982, has important implications for the evaluation of seismic hazards in eastern North America. Although moderate in size (mb5.7), it was well‐recorded teleseismically. Source characteristics of this earthquake have been determined from analysis of data that were digitally recorded by the Global Digital Seismograph Network. From broadband displacement and velocity records ofPwaves, we have obtained a dynamic description of the rupture process as well as conventional static properties of the source. The depth of the hypocenter is estimated to be 9 km from depth phases. The focal mechanism determined from the broadband data corresponds to predominantly thrust faulting. From the variation in the waveforms the direction of slip is inferred to be updip on a west dipping NNE striking fault plane. The steep dip of the inferred fault plane suggests that the earthquake occurred on a preexisting fault that was at one time a normal fault. From an inversion of bodywave pulse durations, the estimated rupture length is 5.5 km. Average properties of the rupture process were examined by a moment tensor analysis of long‐periodPandSHbody waves. The long‐period moment of this earthquake was 5.3 × 1024dyne cm. The static and dynamic stress drops are 41 and 65 bars, respectively, similar to those of many earthquakes with similar moment in regions that are more seismically active. The joint epicenter determination algorithm was used to locate, relative to the mainshock, the three teleseismically recorded aftershocks that occurred through March 31, 1982. The relocated hypocenters of the aftershocks are significantly different from each other and from that of the mainshock; they provide additional support for the source dimensions inferred from the waveform an

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02199

年代:1983

数据来源: WILEY

摘要:

We evaluate the relationships between the spatio‐temporal patterns and faulting mechanisms of small earthquakes and the recent temporal changes in horizontal strain observed along the ‘big bend’ portion of the San Andreas fault near Palmdale, California. Microearthquake activity along the entire big bend of the San Andreas fault increased in November 1976 concurrent with the initiation of an earthquake swarm at Juniper Hills. This activity then decreased abruptly to the northwest and southeast of Juniper Hills during the beginning of 1979. This drop in seismic activity occurred around the time that crustal dilatation was observed on the U. S. Geological Survey Palmdale trilateration network. Focal mechanisms from the study region are predominantly thrust. There are two time periods when the mechanisms are closer to strike slip than to thrust. The first period (December 1976 to February 1977) corresponds to the beginning of the Juniper Hills swarm. The second period (November 1978 to April 1979) approximately coincides with a change in trend of the strain data from uniaxial N‐S compression to dil

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02213

年代:1983

数据来源: WILEY

摘要:

The three‐dimensional crust and upper mantle structure at the Nevada Test Site (NTS) is derived by combining teleseismicPwave travel time residuals withPnsource time terms. The NTS time terms and relative teleseismic residuals are calculated by treating the explosions as a network of ‘receivers’ which record ‘shots’ located at the surrounding stations. Utilization of thePntime terms allows for better crustal resolution than is possible from teleseismic information alone. Average relative teleseismicPwave residuals show a consistent progression of positive (late arrivals) to negative residuals from east to west across the NTS. However,Pntime terms beneath Rainier Mesa are at least 0.3 and 0.5 s less than those beneath Pahute Mesa and Yucca Flat, respectively, indicating the presence of high‐velocity crustal material or crustal thinning beneath Rainier Mesa. The time terms at Pahute Mesa are surprisingly uniform, and the largest time terms and residuals are observed in the northwest and southern parts of Yucca Flat. ThePntime terms show a slight correlation with the working‐point velocity at the shot point for Pahute Mesa and Yucca Flat, indicating that part of the observed lateral variations are caused by shallow effects of the upper crust. Three‐dimensional inversion of the travel time residuals suggests that Yucca Flat is characterized by low‐velocity anomalies confined to the upper crust, Rainer Mesa by very high velocities in the upper and middle crust, and Pahute Mesa by a high‐velocity anomaly extending through the crust and into the upper mantle. Relatively low velocities are observed in the lower crust beneath the Timber Mountain caldera south of Pahute Mesa with no expression in the upper mantle. These observed differences in velocity beneath the Tertiary Silent Canyon and Timber Mountain calderas may be related to their magma volume and mode of enrichment from a mantle

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02220

年代:1983

数据来源: WILEY

摘要:

Analytical results are presented for Rayleigh waves excited by a sudden change in the rate of growth of a subsurface zone of rupture. The curved rupture front advances across an inclined plane. The rupture can be brittle or cohesive tractions can act at its front. The analysis consists of two parts: First, ray theory is used to calculate wavefront approximations to the waves emitted when the rupture front speed suddenly changes. Secondly, a representation integral for the Rayleigh wave, where the integration is performed over a surface enclosing the rupture front, is constructed by using the emitted waves in combination with art appropriate Green's tensor. This integral is evaluated asymptotically. Synthetic accelerographs are constructed which illustrate how the rupture process, and the geometry of the rupture front and the fault plane affect the excitation of Rayleigh waves.

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02233

年代:1983

数据来源: WILEY

摘要:

The principal purpose of this paper is to examine whether movement of matter in a stratified viscoelastic earth can enhance the excitation of the Chandler wobble. We have constructed analytically a two‐layer model consisting of an elastic lithosphere overlying a Maxwell viscoelastic mantle in order to calculate explicitly the temporal evolution of the inertia tensor from earthquake faulting. It is found that the ability of the viscoelastic flows to produce an additional increment of the inertia tensor depends critically onR, the ratio of the inner radius to the outer radius. Only for R ≳ 0.7 is there a substantial increase. For the earth (R≃ 0.98) an amplification of between four and six is obtained from the models considered. However, the time scales over which this enhancement of the moment of inertia takes place is about 10 Maxwell times, thus making this model unlikely to excite the Chandler wobble for values of the mantle viscosity derived from postglacial uplifts. This phenomenon of strain amplification occurs by virtue of ‘transient membrane mechanics,’ as the ratio between the horizontal wavelength of the deformation field is large compared with the elastic shell. By means of a similarity argument we have found that for angular orders less than 40 to 45 ‘transient membrane mechanics’ can be important in global postseismic deformation in the earth. This mode of transient deformation arises from the three‐dimensional nature of the problem and hence cannot be modeled with conventional fl

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02240

年代:1983

数据来源: WILEY

摘要:

Observations of very long period elastic waves are attributed to ‘silent’ earthquakes, events which are presumed to occur at depth in a nonbrittle regime. The fault at depth is modeled as a thin viscoelastic gouge layer embedded in an elastic medium. The source of silent earthquakes is envisaged as a prolonged, slow slippage of the fault faces, which yields a radiation spectrum strongly depleted in the high‐frequency components and dominated by the near‐field term. The results of the model are compared with the dominant frequencies, mean duration, and amplitude of a long‐period perturbation recorded at Trieste in the years preceding the 1976 Friuli (Italy) earthquake. The role that silent earthquakes may play in the overall earthquake mechanism (in particular, as precursory phenomena) is

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02251

年代:1983

数据来源: WILEY

摘要:

Group velocities of fundamental mode Rayleigh waves of 109 wave paths within Eurasia have been measured by an improved frequency‐time analysis technique. Stochastic inversion theory is applied to the mixed‐path measurements to extract pure path group velocities of 10°×10° grid elements that form the subdivisions of Eurasia. Then a three‐dimensional crust and upper mantle structure of the Eurasian continent to a depth of 300 km is constructed. The average crustal thickness of Eurasia is found to be about 40 km. Abnormally thick crust from 50 to 65 km is obtained in areas of Tien Shan, the Altai mountains, Afghanistan, Pamir, Tibet, and Burma. Among them, Tibet has the least dense crust and highest elevation.This implies that at least part of the elevation of Tibet is due to isostatic compensation. The upper mantle seismic velocity of the tectonically active regions of Asia is considerably lower than that of the European and Siberian cratons. If high velocity implies high rigidity, this explains that large‐scale tectonic deformations of Eurasia occur in regions with weak upper mantle that deforms more plastically. Regions with stronger upper mantle remain relatively undeformed. This is especially evident for the Tertiary tectonic deformation as a consequence of the collision of India wi

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02261

年代:1983

数据来源: WILEY

摘要:

We have measured 37 differential travel times for the phase pairs SS‐S, sSS‐sS, ScS2‐ScS, ScS2‐S, sScS2‐sScS, and sScS2‐sS. Each pair included a surface bounce point in the central Indian Ocean between 4°N and 21°S and 69° and 85°E. Residuals were calculated relative to both the J‐B travel time tables and the PEM‐C earth model and were corrected for the elevation of the bounce points. The mean J‐B residual associated with bounce points in the Central Indian Basin is −2.8±0.7 s (fast). This contrasts with +4.2 s (slow) reported for the western Pacific by Sipkin and Jordan, and is similar to the −4.0 s found for old continental nucleii. After correcting for known differences between continental and oceanic crustal structure and for differences in measurement technique, we find that travel times in the mantle beneath the Central Indian Basin are midway in character between those beneath pre‐Cambrian shields and those beneath the western Pacific ocean basin. The fastest differential travel times occur along a line passing close to the center of the large Indian Ocean geoidal low. There is a clear increase in travel time with distance from the center of the geoid low, suggesting that the source of the geoid anomaly may also be largely responsible for the difference between the Central Indian and Pacific basins. The Chagos‐Laccadive Ridge seems to coincide with a transition from the fast differential travel times of the Central Indian Basin to slower differential travel times associated with bounce points beneath young seafloor near active spreading centers. There are also variations within the Central Indian Basin that are not related to surface geological features or to the age of the seafloor. These variations are periodic with a wavelength of about 640 km and with a NNE trend, which agrees approximately with the trend that would be expected for linear, convec

ISSN:0148-0227    

DOI:10.1029/JB088iB03p02273

年代:1983

数据来源: WILEY