Focal depths and moment tensor representations of shallow earthquakes associated with the Great Sumba Earthquake

Abstract

Focal depths in the range 8 to 24 km below the seafloor were determined for seven of the larger shocks associated with the Great Sumba earthquake of August 19, 1977. Depth control was achieved by modeling digitized long‐period P waves recorded by the World‐Wide Standard Seismic Network. If the base of the rupture at the time of the mainshock were no greater than the depth of the deepest aftershock, the stress drop would be nearly 500 bars. Amplitude and first motion data from the P waves were inverted for point source moment tensors. The inversion of Rayleigh and Love wave spectra computed from digital seismograms recorded by the Seismic Research Observatories (SRO) and the Abbreviated Seismic Research Observatories (ASRO) broadened the source study to include earthquakes with moments as small as 1024 dyn cm: too small for either traditional first motion solutions or P wave modeling. The inversion procedure was applied to data in the passband from 30 to 120 s. The propagation phase for the Love and the Rayleigh waves was calibrated with signals from two earthquakes for which the source phase was inferred from the P wave solutions. Agreement with the corresponding P wave solutions is nearly the same whether or not the smaller eigenvalues of the inner product matrix are retained in the inversions. The double couple components generally account for more than two thirds of the total moment. Normal faulting along an east to northeasterly strike characterizes the mechanisms in the principal zone of aftershock activity. Strike slip mechanisms consistent with horizontal compression normal to the arc prevail in the delayed aftershock zone located 200 km northwest of the mainshock. Of the earthquakes subjected to this surface wave analysis the foreshock stood out as having felt a significantly higher apparent stress, perhaps as much as one half an order of magnitude higher than the others. These results suggest that accurate source parameters for moderate to small magnitude earthquakes can be determined from surface waves recorded by the SRO and ASRO. In particular, the mechanisms of small magnitude foreshocks could be obtained in this way.

Faulting caused by earthquakes beneath the outer slope of the Japan trench.

“…In contrast, SILVER and JORDAN (1983) obtained a centroid depth shallower than 20 km from the moment spectra. Consistent with this centroid depth, FITCH et al (1981) found that the depth range of major aftershocks is 8-24 km and LYNNES andLAY (1988) obtained the source extent not deeper than 35 km from the analysis of long-period body waves.…”

“…The incorrect depths may lead to incorrect calculation of the take-off angles to local stations and thus result in incorrect fault plane solutions (SENO and KROEGER, 1983). Therefore, previous studies of trench-outer rise events (HERMANN, 1976;CHEN and FORSYTH, 1978;FITCH et al, 1981;STEIN et al, 1982;FORSYTH, 1982;CHINN and ISACKS, 1983;WARD, 1983;CHRISTENSEN and RUFF, 1987) have emphasized that these events must be studied individually using waveform analysis. In this study, we determined focal depths and fault plane solutions by comparison of synthetic and observed seismograms of WWSSN long-and short-period P waves.…”

On the resolution of the isotropic component in moment tensor inversion

“…This misestimation of the depth and source time function is systematically reflected as a strong correlation between the elements of the isotropic component, and among these elements M,, appears to be particularly poorly solved in surface-wave inversions, especially at shallow depths. This formalizes the trade-offs observed empirically between the isotropic part, the hypocentral depth and the source time function (Fitch et al 1981;Stein & Wiens 1986). It also indicates that, when an additional constraint must be applied in order to reduce the number of unknowns from six to five, choosing a constraint on the isotropic part, and particularly the condition M,, = -M x x -My,, is among the best possible choices.…”

Moment tensor inversion for focal mechanism of the Beibuwan earthquakes

“…Moment tensor inversion has become an important and active domain in the physics of seismic source. In order to obtain some information, such as the source time function describing source rupture process, to determine the moment tensor describing the focal mechanism, and to estimate the scalar moment measuring the earthquake magnitude, seismologists have applied the moment tensor inversion to various data, such as the free oscillation (Gilbert, Dziewonski, 1975), surface wave (McCowan, 1976;Mendiguren, 1977;Aki, Patton, 1978;Kanamori, Given, 1981;Romanowicz, 1982;Lay, et al, 1982), body wave (Stump, Johnson, 1977;Ward, 1980;Fitch, et al, 1981;Dziewonski, et al, 1981) and near-source recording (Stump, Johnson, 1984;Ni, et al, 1991, Wu, etal, 1994 ~.…”