Moment Tensors Of Earthquakes Research Articles

Active deformation of the Gorda plate: Constraining deformation models with new geophysical data

The Gorda plate, the southernmost fragment of the larger Juan de Fuca plate system, is an example of a nonrigidly deforming tectonic accommodation zone or buffer plate, absorbing deformation and allowing the surrounding larger plates to act in a more rigid fashion. Here we present a new structural analysis of the plate based on full-plate bathy- metric coverage, augmented by seismic reflection data and earthquake moment tensors and locations. We interpret internal deformation of the Gorda plate as an asymmetrical flexural-slip buckle with a vertical axis, utilizing reactivation of spreading-ridge fabric normal faults as strike-slip faults. Newly formed second-generation faults crosscutting the structural grain overprint the reactivated structures. The spreading fabric faults finally begin a second phase of extension as the plate approaches the subduction zone. This model, based on fault constraints, has allowed investigation of ridge-plate-subduction interac- tions, and suggests that spreading-rate variations along the Gorda Ridge may be con- trolled by internal deformation of the plate rather than the reverse, as previously hypothesized.

Three Decades of Seismic Activity at Mt. Vesuvius: 1972?2000

We analyse the seismic catalogue of the local earthquakes which occurred at Somma-Vesuvius volcano in the past three decades (1972–2000). The seismicity in this period can be described as composed of a background level, characterised by a low and rather uniform rate of energy release and by sporadic periods of increased seismic activity. Such relatively intense seismicity periods are characterised by energy rates and magnitudes progressively increasing in the critical periods. The analysis of the b value in the whole period evidences a well-defined pattern, with values of b progressively decreasing, from about 1.8 at the beginning of the considered period, to about 1.0 at present. This steady variation indicates an increasing dynamics in the volcanic system. Within this general trend it is possible to identify a substructure in the time sequence of the seismic events, formed by the alternating episodes of quiescence and activity. The analysis of the source moment tensor of the largest earthquakes shows that the processes at the seismic source are generally not consistent with simple double-couples, but that they are compatible with isotropic components, mostly indicating volumetric expansion. These components are shown to be statistically significant for most of the analysed events. Such focal mechanisms can be interpreted as the effect of explosion phenomena, possibly related to volatile exsolution from the crystallising magma. The availability of a reduced amount of high quality data necessary for the inversion of the source moment tensor, the still limited period of systematic observation of Vesuvius micro-earthquakes and, above all, the absence of eruptive events during such interval of time, cannot obviously permit the outlining of any formal premonitory signal. Nevertheless, the analysis reported in this paper indicates a progressively evolving dynamics, characterised by a generally increasing trend in the seismic activity in the volcanic system and by a significant volumetric component of recent major events, thus posing serious concern for a future evolution towards eruptive activity.

Present-Day crustal deformation field in China continent inverted from GPS observations, earthquake moment tensor and quaternary fault slip rates

Based on the bi-cubic Bessel spline function method, we have inverted the present-day crustal horizontal velocity field and deformation field in China continent by combining with 410 GPS observations, 327 seismic moment tensors of earthquakes and fault slip rates in China and its neighboring areas, and considered the geological and geophysical parameters at the same time. The results reveal that the crustal movement mostly takes on compression in SN-NE direction and extension in EW-NW direction in China continent. The continent is rotating southeast and takes on a clockwise rotation image from the west to the east, particularly in the southeast of Tibet and Chuan-dian area. Different blocks have different deformation in China continent. The south China, Ordos and Northeast block have good integrity. The deformation in the western part of China is obviously stronger than that in the eastern part. The strong strains are focused on the Himalaya, east Tibet, west Tian Shan and Chuandian block etc. The deformation at the joint of blocks is stronger than that within the blocks. The China continent deformation not only has the strike-slip faulting and extrusion feature but also has the crustal shortening and thickening feature.

Stress field along the Ryukyu Arc and the Okinawa Trough inferred from moment tensors of shallow earthquakes

We investigated the stress field along the Ryukyu Arc and the Okinawa Trough using regional (NIED MT) and global (Harvard CMT) moment tensor catalogs. Using the combined dataset of 4 years of data from the NIED MT catalog and 24 years of data from the Harvard CMT catalog, we found an arc-parallel extensional stress province along the forearc of the Ryukyu Arc and observed a detailed extensional stress field in the Okinawa Trough. This arc-parallel extension is observed in the entire region of the Ryukyu Arc except at its northeastern end. This stress field is clearly separated from the backarc extensional stress field in the Okinawa Trough by the volcanic chain. Since it exists in a normal subduction region as well as in a region of weakly coupled subduction, the effects of oblique subduction might not be enough to explain the formation of the arc-parallel extension. Thus, we consider that the backarc opening process should play an important role in this arc-parallel extension. Along the Okinawa Trough, an arc-perpendicular extensional stress field is observed in the southeastern and central Ryukyu Arc, while in the northeastern Okinawa Trough, the direction of the extensional axis is oblique to both the arc normal direction and the direction of plate motion. Two-dimensional mechanical interaction models which consider the interactions between the subducting plate, forearc and backarc are inadequate because an arc-parallel extensional stress field is observed in the entire Ryukyu Arc, and the extension axis in the northeastern Okinawa Trough is oblique to the arc trend. The oblique backarc stress may originate beneath the backarc or on the continental side of the backarc. One of the simplest ways to explain this observation might be active rifting in the northeastern Okinawa Trough.

Stability of earthquake moment tensor inversions: effect of the double-couple constraint

We show that spurious large non-double-couple components can be obtained in inversions for the full deviatoric moment tensor for shallow crustal earthquakes due to inaccurate Earth models. The traditional “best double-couple” solution does not in general provide an optimal estimate of a double-couple mechanism, and is only reliable when the non-double-couple component of the full deviatoric solution is small. The inverse problem for the moment tensors of the 1998 Antarctic Plate and 2000 Wharton Basin strike-slip earthquakes is shown in each case to have two well-fitting minima in the misfit function of pure double-couple solutions. Such pairs of solutions are most likely to exist for earthquakes which are close either to vertical strike-slip or to dip-slip on a fault plane dipping at 45°. It is shown theoretically that these pairs of solutions arise from the combination of the pure double-couple constraint and the instability of two elements of the moment tensor. No significant non-double-couple component is found for the shallow thrusting 1996 Biak, Indonesia earthquake.

Upper-mantle anisotropy beneath the Altai–Sayan region of central Asia

The Altai–Sayan region of central Asia is actively deforming despite its being located several thousands of kilometers away from the collisional front between India and Eurasia. To add constraints to the role of the upper mantle in this type of tectonic environment, we analyzed intermediate-period, analogue records of SKS and SKKS to determine the state of mantle anisotropy in this region. We find that for all stations located in actively deforming areas, the fast direction of anisotropy correlates well with indicators of Cenozoic to the present-day deformation, and correlates with older structures only when they happen to align with younger ones. The general east-northeast trend of the fast direction of anisotropy in active regions also aligns with the principle extensional strain axis determined from earthquake moment tensors. At one station (NVS), located outside of any actively deforming region, the N–S fast direction may correspond to the long-wavelength strain field of the mantle beneath much of Siberia. These correlations, along with the size of the delay times (1–2 s), suggest that the anisotropy in the upper mantle beneath the Altai–Sayan is responding to recent tectonic activity.

Seismic hazard reappraisal from combined structural geology, geomorphology and cosmic ray exposure dating analyses: the Eastern Precordillera thrust system (NW Argentina)

Because earthquakes on large active thrust or reverse faults are not always accompanied with surface rupture, paleoseismological estimation of their associated seismic hazard is a difficult task. To improve the seismic hazard assessments in the Andean foreland of western Argentina (San Juan Province), this paper proposes a novel approach that combines structural geology, geomorphology and exposure age dating. The Eastern Precordillera of San Juan is probably one of the most active zones of thrust tectonics in the world. We concentrated on one major regional active reverse structure, the 145 km long Villic´um-Pedernal thrust, where this methodology allows one to: (1) constrain the Quaternary stress regime by inversion of geologically determined slip vectors on minor or major fault planes; (2) analyse the geometry and the geomorphic characteristics of the Villic´um-Pedernal thrust; and (3) estimate uplift and shortening rates through determination of in situ-produced 1 0 Be cosmic ray exposure (CRE) ages of abandoned and uplifted alluvial terraces. From a structural point of view, the Villic´um-Pedernal thrust can be subdivided into three thrust portions constituting major structural segments separated by oblique N40°E-trending fault branches. Along the three segments, inversion of fault slip data shows that the development of the Eastern Precordillera between 31°S and 32°S latitude is dominated by a pure compressive reverse faulting stress regime characterized by a N110° ′ 10°E-trending compressional stress axis (σ 1 ). A geomorphic study realized along the 18 km long Las Tapias fault segment combined with CRE ages shows that the minimum shortening rate calculated over the previous ∼20 kyr is at least of the order of 1 mm yr - 1 . An earthquake moment tensor sum has also been used to calculate a regional shortening rate caused by seismic deformation. This analysis of the focal solutions available for the last 23 yr shows that the seismic contribution may be three times greater than the shortening rate we determined for the Las Tapias fault (i.e. ∼3 mm yr - 1 ), suggesting that the San Juan region may have experienced a seismic crisis during the 20th century. Moreover, the ramp that controls the development of the Eastern Precordillera appears to be one of the main seismic sources in the San Juan area, particularly the 65 km long Villic´um-Las Tapias segment. A first-order evaluation of the seismic hazard parameters shows that this thrust segment can produce a maximum earthquake characterized by a moment magnitude of ∼7.3 (′0.1) and a recurrence interval of 2.4 (′1.5) kyr. This part of the Villic´um-Pedernal ramp may have ruptured during the Ms =7.4, 1944 San Juan earthquake producing very few surface ruptures and only distributed flexural slip deformation on to the Neogene foreland bedding planes between the Eastern Precordillera and Pie de Palo.

Seismic remote monitoring of stress field

We demonstrate that a real-time based moment tensor analysis using broadband seismic waveforms provides us with the information about the spatio-temporal change of stress field. Such a capability is crucial when monitoring the volcanic activity during the eruption if it causes seismic swarm. We estimated the moment tensors of earthquakes occurring during the swarm activity before the eruption of Usu volcano in March 2000 as well as those during the swarm activity of Miyakejima volcano starting in June 2000. In both cases, we detected the stress change due to the movement of magma reservoir.

Inversion for parameters of tensile earthquakes

The tensile source model generalizes the shear source model by assuming that the slip vector can be arbitrarily oriented with respect to the fault and is not constrained to lie within the fault plane. The proposed inversion for the parameters of tensile sources is based on the evaluation of the isotropic (ISO), compensated linear vector dipole (CLVD), and double‐couple (DC) components in seismic moment tensors. The most significant parameters inverted are the λ/μ ratio at the fault (denoted as the κ parameter) and the inclination α of the slip vector from the fault. The κ parameter is significant for discriminating noisy moment tensors of shear earthquakes from those of tensile earthquakes. The inclination α can be accurately determined from the DC component in the moment tensor because the DC component rapidly decreases with increasing α. For example, the inclination of 20° causes DC being ∼50–60% only. The inversion is applied to earthquakes which occurred in January 1997 in West Bohemia, Czech Republic. It is shown that some of these earthquakes display tensile faulting. The κ parameter is ∼0.1. The inclination of the slip from the fault attains values of up to 20°. This inclination is a result of tensile traction and reduced shear traction along the fault and high‐fluid pressure in the region.

Active seismic deformation in the grabens of northern Central America and its relationship to the relative motion of the North America–Caribbean plate boundary

The active seismic deformation along the grabens of northern Central America is calculated using a method which relates relative plate velocity with the seismic strain rate tensor. The latter is, in turn, obtained from seismic moment tensors.We calculate the typical or average moment tensor from available Centroid-Moment Tensors and fault-plane solutions. This average tensor is then used to obtain the sum of moment tensors of historical and modern earthquakes reported in the literature.Both historical and modern earthquakes yield an average extension rate of 8mm/yr along the grabens of northern Central America, while the relative motion between the North America and Caribbean plates expressed along the Motagua–Polochic fault system is 20mm/yr.This result suggests that part of the seismic deformation related to the plate boundary is taken up as extension along the grabens, which play the role of fault termination structures for the Motagua–Polochic system.

Strain pattern of the Southern Tyrrhenian slab from moment tensors of deep earthquakes: implications on the down-dip velocity

Seismic strain is analysed along the slab face of the Southern Tyrrhenian subduction zone using focal mechanisms of deep earthquakes which occurred in the period 1960-1998. Results show that the slab is mostly affected by down-dip shortening which strongly increases below 250 km depth. Extensional strain is mainly confined to the direction perpendicular to the slab face. The dominance of down-dip shortening and the minor along strike inplane extension implies thickening of the slab below 250 km depth. Assuming constant seismic efficiency along the slab, this strain pattern also implies a decrease of the down-dip velocity below 250 km depth. We also locate lower magnitude intermediate-depth and deep earthquakes using arrival times since 1985 available from the Italian seismic national network. These data show that the slab reaches the deeper part of the upper mantle, as suggested by the occurrence of a few ??600 km depth earthquakes, and that a large portion of the Tyrrhenian slab, between 100 and 250 km depth beneath the offshore of the Calabrian arc, is aseismic. Only a short part of the Tyrrhenian slab is seismically continuous from the top to the bottom. The lack of seismicity may indicate either that aseismic subduction is occurring or that the slab is detached from its upper part. Although the data are still inconclusive, they suggest that an aseismic subduction is the most appropriate interpretation, considering recent tomographic images of the slab and the results of this study, which agree well with the presence of a neutral down-dip stress zone, as also observed worldwide in deep slabs.

Seismic moments of major earthquakes and the rate of shortening across the Tien Shan

Three Twentieth Century earthquakes that Richter assigned M ≥ 8 and two of comparable magnitude in the Nineteenth Century imply rapid deformation within the Tien Shan. Seismic moment tensors of major earthquakes in this century suggest an average shortening rate of 7 (± 2) mm yr−1 across the Tien Shan. In the western part, however, where three of the five largest earthquakes occurred, the calculated rate is consistent with the value of ∼20 mm yr−1 measured using GPS in that area by Abdrakhmatov et al. As Avouac et al. suggested, the high rate in the western part apparently is a manifestation of counter‐clockwise rotation of the Tarim Basin relative to Eurasia about an axis near the east end of the Tien Shan.

Investigation of the submarine active tectonism in the Gulf of Gökova, southwest Anatolia–southeast Aegean Sea, by multi-channel seismic reflection data

Submarine active tectonism in the Gulf of Gökova located at the southwest Anatolia–southeast Aegean Sea region was investigated by means of multi-channel seismic reflection data. The Gökova basin is filled by the latest Miocene–Pliocene–Quaternary sediments with maximum thickness of about 2.5 km. The Lycian Nappes, which predominantly cover extreme southwestern Anatolia, constitutes the basement rocks for the Gökova province. The gulf was mainly opened by a buried major listric normal fault, so-called Datça Fault, which has not been previously discussed in the literature. The north-dipping, mainly E–W-trending Datça Fault is located at the southern part of the gulf, whereas its associated antithetic faults are located at the north. The onset time of the opening of the gulf is possibly in the latest Miocene–Pliocene. In terms of local rather than regional effects, the activity of the Datça Fault has decelerated, possibly since the Pleistocene. The Datça Fault might have gained its curved fault plane as it evolved, beginning as planar and/or using antecedent planes of the Lycian Nappes in the area. As the extension progressed, i.e., as the hanging wall block slipped further north, gravity may have impeded rather than helped the faulting. On the other hand, continuing extension in the area may have initiated a second phase of faulting, i.e., WNW–ESE-oriented subgrabens in the gulf and major E–W normal faulting in the northeast margin. A bathymetric low in the mid-gulf area and a horst–graben system in the eastern part of the gulf are observable from the bathymetric data and are well correlated to the seismic data. Although the main orientation of the gulf is E–W, more recent WNW–ESE structures are remarkable in the mid-gulf and in the eastern part of the gulf. The latest WNW–ESE structures are also in agreement with the results of GPS and SLR studies as well as plate motion modelling by total moment tensor of earthquakes in the western Anatolia–Aegean Sea region, particularly in southwestern Anatolia. The amount of total N–S extension within the gulf is estimated as at least 5.5 km since the latest Miocene–Pliocene with overall constant extension rate of at least 1.1 mm/y where the estimated extension factor is about β=1.3.

How well constrained are well‐constrained T, B, and P axes in moment tensor catalogs?

The T, B and P axes of earthquake moment tensors (MT) are often used to evaluate regional stress directions and other tectonic parameters; we here undertake three comparisons to assess the uncertainty in the orientations of these axes. These are (1) a direct comparison of common MT in the Harvard, U.S. Geological Survey, or Earthquake Research Institute (ERI) catalogs; (2) a comparison of MT slip vectors and plate motion vectors in several tectonically straightforward regions; and (3) an analysis of the axial variability in the Harvard and ERI catalogs implied by the reported uncertainties in individual MT components. All three comparisons indicate that there is considerable variability within the catalog concerning the axial orientation of MT, but all suggest that axis orientations of the majority of Harvard MT have uncertainties of 15° or less. For compensated linear vector dipole (CLVD) components among the three catalogs, the correlation is very low. For the Harvard catalog, three statistics are especially useful for selecting better constrained MT; these are (1) the relative error Erel, which is the ratio of the scalar moments of the reported error tensor and of the MT itself; (2)ƒCLVD, a measure of the strength of the CLVD component; and (3) nfree, the number of MT elements not fixed at zero in the inversion. For selecting better constrained MT, the appropriate statistical cutoffs chosen depend on the problem of interest, the data available, arid personal preference. However, for analysis of shallow earthquakes we have used Erel ≤ 0.15, ƒCLVD ≤ 0.20, and nfree = 6. While this eliminates 53% of the catalog, our calculations suggest that nearly all the remaining events have T, B, and P axes with azimuth and inclination angle uncertainties of 5°–10° or less.

Contemporary kinematics of the western United States determined from earthquake moment tensors, very long baseline interferometry, and GPS observations

Using moment tensors from earthquakes between 1850 and 1995, we determine the horizontal velocity gradient tensor field associated with the seismic deformation in the western U.S. plate boundary zone. The velocity vectors obtained from the integration of the seismic strain rates across the entire plate boundary lie within 5° of the NUVEL‐IA Pacific‐North American plate motion direction. The magnitude of the earthquake‐related velocity is 62% of the NUVEL‐1A total Pacific‐North American plate motion. If earthquake occurrence is a random, Poisson‐like process, then the large formal errors associated with the model velocity estimates are an indication that the 144‐year interval is too short to define the long‐term seismic moment release rate to within an uncertainty less than the deficit between the observed seismic moment release rate and the total long‐term moment release rate associated with both the seismic and aseismic deformation. To spatially resolve the deficit of the earthquake moment release in the last 144 years, we have also estimated the total long‐term horizontal velocity gradient tensor field using seismic strain rate tensors and recent geodetic velocities, both with and without the NUVEL‐1A Pacific‐North American rigid plate motion constraint. We find a systematic difference between the Southern California Earthquake Center (SCEC) geodetic velocity map and NUVEL‐1A North American reference frames with a pole at (38.8°N, 124.2°W, 0.33° m.y.−1). This difference is significant at the 99% confidence level and indicates that 4–7 mm/yr of NE‐SW convergence is absorbed in offshore areas of southern California if both the SCEC velocity map and NUVEL‐1A correctly describe the relative motion between the Pacific and North America plates. Deficit strain rates calculated from comparison of observed seismic strain rates over the last 144 years with long‐term model strain rate tensors are largest in areas east of San Francisco Bay, in the Transverse Ranges, and along the southern San Andreas fault and San Jacinto fault. The deficit in each 100 by 50 km grid area within these regions is equivalent to an accumulation of strain in the last 144 years that would be released by an Mw>7 earthquake. Elsewhere the accumulated strain in the last 144 years is small, except in northern Mojave Desert, along the Garlock fault, and in the region southwest of the southern San Andreas‐San Jacinto fault zone. Within each 100 by 50 km grid area in these regions the accumulated strain is equivalent to that released by an Mw=6.5–7.0 earthquake.

Spatial distribution of horizontal seismic strain in the Apennines from historical earthquakes

Horizontal principal seismic strain rate axes have been calculated within a regular mesh of triangles covering the Italian peninsula in a time interval of 700 years. I have used both the method of Kostrov (1974), that requires knowledge of the seismic moment tensor of earthquakes, and the modified version provided by England and Molnar (1997) that makes use of length and kinematics of the activated faults. Seismic moment tensor of historical earthquakes can be inferred from recent literature, while length of faults has been obtained from the observation that strain drop is almost constant for large Apenninic earthquakes. Spatial strain distribution from historical earthquakes shows that the Apennines can be divided into three homogeneous structural arcs (Northern Apenninic, Southern Apenninic and Calabrian arcs) within which strain is roughly constant. Although NE-SW extension is the main deformation process along the two Apenninic arcs it involves a velocity more than five times greater in the Southern Apennines. Along the Calabrian arc, I tested the effect on the strain field of the contemporaneous ~WNW-ESE and ~NNE-SSW extension due to the longitudinal dilatation of the arc during its still ESE migration.

Three-dimensional seismic structure and moment tensors of non-double-couple earthquakes at the Hengill-Grensdalur volcanic complex, Iceland

SUMMARY The volcanic and geothermal areas of Iceland are rich sources of non-double-couple (non-DC) earthquakes. A state-of-the-art digital seismometer network deployed at the Hengill‐Grensdalur volcanic complex in 1991 recorded 4000 small earthquakes. We used the best recorded of these to determine 3-D V P and V P /V S structure tomographically and accurate earthquake moment tensors. The V P field is dominated by high seismic wave speed bodies interpreted as solidified intrusions. A widespread negative (’4 per cent) V P /V S anomaly in the upper 4 km correlates with the geothermal field, but is too strong to be caused solely by the eVect of temperature upon liquid water or the presence of vapour, and requires in addition mineralogical or lithological diVerences between the geothermal reservoir and its surroundings. These may be caused by geothermal alteration. Well-constrained moment tensors were obtained for 70 of the best-recorded events by applying linear programming methods to P- and S-wave polarities and amplitude ratios. About 25 per cent of the mechanisms are, within observational error, consistent with DC mechanisms consistent with shear faulting. The other 75 per cent have significantly non-DC mechanisms. Many have substantial explosive components, one has a substantial implosive component, and the deviatoric component of many is strongly non-DC. Many of the non-DC mechanisms are consistent, within observational error, with simultaneous tensile and shear faulting. However, the mechanisms occupy a continuum in source-type parameter space and probably at least one additional source process is occurring. This may be fluid flow into newly formed cracks, causing partial compensation of the volumetric component. Studying non-shear earthquakes such as these has great potential for improving our understanding of geothermal processes and earthquake source processes in general.

Geodetic strain of Greece in the interval 1892–1992

A first‐order triangulation of Greece was carried out in the 1890s. Reoccupation, using Global Positioning System receivers, of 46 of the 93 original markers yielded estimates of the deformation of the region over the intervening interval. Broad regions have similar geodetic strain over the 100‐year time span. Strain north of the Gulf of Korinthos is predominantly north‐south extension, though with a significant east‐west component. The central Peloponnisos is relatively stable, whereas the gulfs of the southern Peloponnisos are all characterized by uniaxial east‐west extension. The seismic expression of strain for the entire region, calculated from the seismic moment tensors of earthquakes of MS≥5.8 during the past 100 years, accounts for only 20–50% of the geodetically determined strain. At a scale of 50–100 km, the fraction of the strain that is expressed seismically varies much more than this range. In particular, whereas seismic strain in the eastern Gulf of Korinthos over the past 100 years is commensurate with the geodetic strain, there is rapid extension across the western Gulf of Korinthos (∼0.3 μstrain yr−1), with negligible seismic strain for the 100 year period prior to 1992. The Egion earthquake of June 1995 in the western Gulf of Korinthos released only a small proportion (≤20%) of the elastic strain that had accumulated in that region. The observed distribution of displacements can be explained by the relative rotation of two plates with a broad accommodation zone between them, but it is equally consistent with the deformation that would be expected of a sheet of fluid moving toward a low‐pressure boundary at the Hellenic Trench. A simple calculation implies that if the region does behave as a fluid, then its effective viscosity is ∼1022–1023 Pa s. Such viscosities are consistent with the deformation of a lithosphere obeying a rheological law similar to that obtained for olivine in the laboratory.

Contribution of seismic tomography in moment-tensor inversions using teleseismic surface-wave spectra

Abstract The knowledge of lateral heterogeneities is crucial for path corrections in moment tensor inversions using surface waves. After some attempts to use regionalized Earth models for very long-period surface-wave moment-tensor inversions, recent tomographic Earth models offer the possibility to make short-period path corrections and therefore retrieve more reliable moment tensors for teleseismic earthquakes. First we try to evaluate the precision required for path corrections in comparison with source effects. Some selected Earth models are tested to evaluate how their results compare to those using multiple-frequency filtering techniques. Some real cases illustrate the sensitivity of moment-tensor solutions to the different path corrections, and it appears clearly that regionalized Earth models and tomographic models deduced from long-period data alone (greater than 150 sec) cannot lead to trustworthy broadband moment-tensor inversions. Recent tomographic models using phase velocities at much shorter periods (40 to 200 sec) offer a precision comparable to that of the multiple-frequency filtering technique. Both methods lead to acceptable source mechanisms, using a small number of stations, in more than two cases out of three. The use of recent global tomographic models based upon shorter-period surface waves might thus be a useful alternative to heavy multiple-frequency filtering techniques to automate source studies, especially for rapid determinations using a small number of stations.

Velocity gradients in the northern Indian Ocean inferred from earthquake moment tensors and relative plate velocities

The region in the northern Indian Ocean between the Central Indian Ridge, the Sumatra Trench, and 15°S to 9°N exhibits high seismicity, along with basement and sediment deformation. This area represents a diffuse plate boundary between the distinct and separate portions of the Indian and Australian plates. We invert the spatial distribution of horizontal strain rates in this diffuse plate boundary zone to obtain the horizontal velocity field associated with the moment release from 61 earthquakes that have occurred over the past 80 years. In the inversion, the three‐dimensional rotation vector function W is determined at all corner points (i.e., knot points) of a grid such that the predicted strain rate values match the observed strain rates within the grid areas in a least squares sense. The rotation vector of Australia relative to India, defined by the earthquake data alone, is 10.1°S±9.1°, 81.5°E±5.4°, 0.1°±0.06°/m.y. (1 sigma error), which is slightly different in location from the expected vector estimated using seafloor magnetic anomalies and transform fault trends (3.2°S, 75.1°E, 0.288°/m.y.) [DeMets et al, 1994a, b]. By minimizing the rate of work within a continuous viscous medium that accommodates Australia's motion relative to India [DeMets et al, 1994a], we determine the complete horizontal velocity gradient tensor within the zone of diffuse deformation. The viscosity tensor within all of the deforming areas can be specified. When all the regions are free to strain in any direction and have the same viscosity (uniform isotropic viscosity), the predicted strain pattern is either pure extension or pure compression, which is in contrast to the style of earthquake deformation east of Chagos where both thrust and strike‐slip mechanisms are observed. However, when the isotropic viscosity of regions within the Wharton Basin is increased relative to the isotropic viscosity levels of areas along the Ninetyeast Ridge, the plate motion is accommodated by combined thrust and strike‐slip styles of strain along the Ninetyeast Ridge and within the eastern Central Indian Basin. This predicted strain that accommodates plate motion is in agreement, both in direction and style, with the earthquake‐related strains in these regions. Although most of the observed style and direction of strain can be explained by lithospheric strength differences between the Wharton Basin and Ninetyeast Ridge areas alone, the presence of strike‐slip faulting within the Wharton Basin requires a preexisting fabric that favors strike‐slip faulting there as a strain mechanism that accommodates, in part, the motion of Australia relative to India. To first order, the direction of principal strain rates, the style of strain, and the spatial distribution of strain rates from the earthquakes are in accord with the accommodation of the expected plate motion, whereas across the plate boundary zone the rates from 80 years of seismicity average out to at most 60% of the total strain rate.