佐藤 利典

The 1994 Sanriku-oki earthquake (M-w 7.7) ruptured the entire seismogenic zone of the Pacific plate subduction beneath northeastern Japan are. We deployed 18 ocean bottom seismographs and recorded its aftershocks, which span the whole seismogenic zone, with unprecedented accuracy. We calculate hypocenters and mechanisms to improve our understanding of the plate subduction geometry and plate coupling. The rupture of the mainshock initiated at the updip end of the seismogenic zone defined as the trenchward limit of the aftershock and background seismicity. The rupture continued to slip along a shallow dipping ( < 10 degrees) interface for similar to 60 km, and then slipped across an asperity where the dip angle starts to increase at similar to 143 degrees E. Here aftershocks are few, suggesting complete release of strain. The rupture continued another 100 km or so to the downdip end beneath the coastline, where the plate dip angle reaches 30 degrees to smoothly connect to the dip angle of the Wadati-Benioff zone. Coseismic slips are larger westward of 143 degrees E, where the mantle wedge comes into contact with the subducting plate. The stress redistribution due to the mainshock seems to have caused aftershocks with normal faultings to occur offset above and below the plate boundary.

Off Sanriku region is a well known plate subduction zone, which lies on the west side of the Japan Trench. High seismic activities have been observed, but there are also low seismic activity regions. We conducted an OBS-controlled source seismic experiment across the seismic-aseismic boundary in 1996 to examine the relationship between seismic activity variation and crustal structure. A traveltime analysis using refraction and reflection waves was applied to the observed data to determine the 2D crustal structure. An amplitude analysis of reflection waves from the subducting plate boundary revealed a good correlation between acoustic-impedance (seismic reflectivity) and seismic activity; that is, high-impedance region is a low seismic region and vice versa. We propose the hypothesis that high-impedance is caused by fluid or hydrate rocks around the plate boundary, and this hypothesis explains the crustal velocity structure and the variations of observed reflection waves and seismic activities.

We determined the deep seismic structure of the seismogenic plate boundary zone of off-Sanriku, the northern part of the Japan Trench inner slope area. Travel time data from aftershocks of the 1994 Far-off Sanriku earthquake (M = 7.5) recorded both by ocean bottom seismographs and the local land seismic network enable us to image the three-dimensional seismic velocity structure of the region. Along the plate boundary, there exists a P-wave low-velocity layer of ca. 7 km s(-1) and with thickness <10 km as the topmost layer of the subducting Pacific plate. The present analysis indicates that this layer, interpreted as the subducting oceanic crust, continues without any velocity variations at least down to ca. 40 km depth. In the overriding plate, we found a zone of considerable high P-wave velocity, ca. 8 km s(-1), between the Pacific coastal line of the northeastern Japan are at the longitude of 142.5 degrees E, This manifests that the trenchward toe of the mantle wedge has significantly higher velocity than that beneath the land area of the are. The landward limit of the high P-wave velocity toe corresponds to the down-dip limit of the seismogenic plate boundary. Further to the east, the P-wave velocity is ca 6 km s(-1). The structure image obtained indicates that the nature of the overriding plate is significantly different along the plate boundary. In the eastern half of the seismogenic zone, the island are crust and subducting oceanic crust are in contact with each other, whereas the high-velocity mantle wedge makes up the hanging wall of the deeper part of the seismogenic zone. (C) 2000 Elsevier Science B.V. All rights reserved.

We develop a systematic approach for the analysis of 2-D refraction experiments using traveltimes that allows progressive improvement of velocity structure through a sequence from 1-D models to pseudo-2-D models, and then 2-D models. The approach consists of three steps. First, 1-D velocity models are constructed for each segment of the profile using a genetic algorithm inversion, and then pseudo-2-D models are constructed using a turning point approximation. The purpose of this step is to provide an approximate image of 2-D velocity structure, and to infer the number and general location of layer boundaries. The second step uses 1-D layered modelling, again with a pseudo-2-D conversion, to generate a rough 2-D layered structure. The third step consists of smoothing the pseudo-2-D model in order to create initial models for use in 2-D inversion and the construction of a 2-D model using a Bayesian formulation of non-linear iterative inversion. All the steps exploit traveltime data for the first arrival and do not use any trial-and-error forward modelling. The progressive approach is efficient because the results of each step are employed as the initial model for the next step, The method is applied to real data from an along-strike experiment and also to a related synthetic example so that the quality of the solution can be judged. The results indicate that the method is robust, and this is confirmed by a further synthetic example that represents a survey across a trench and dipping subduction zone. The systematic approach to the inversion of refraction data enables a complex inversion to be undertaken in 1 or 2 days, The sequential approach allows the incorporation of additional information if desired.

The eastern margin of the Japan Sea is a nascent convergent plate boundary. Previous studies proposed the existence of a seismic gap along this boundary between 39 degrees N and 40 degrees N. The trend of this gap is reported by Ohtake (Island Are 4, 156-165, 1995) to be north-northwest to south-southeast, but by Ishikawa (Gekkan Kaiyo, Suppl. 7, 102-107, 1994) and Matsuzawa (Prog. Abstr., Seismol. Sec. Jpn. 2, B92, 1995) to be north-northeast to south-southwest. During one month ocean bottom seismic observations were conducted using nine ocean bottom seismometers to investigate seismicity in and around the seismic gap area in detail. The observations revealed that the earthquake epicentral distribution had an echelon shape and could be divided into three soups. These groups have a north-northeast to south-southwest trend. This trend is consistent with the fault system in this area, which was formed by the back-are spreading in the Early to Middle Miocene. This suggests that previously formed tectonic structures affect the present seismo-tectonics and that this area has weak planes with a north-northeast to south-southwest trend. (C) 1999 Elsevier Science B.V. All rights reserved.

A seismic refraction-reflection experiment using ocean bottom seismometers with;artificial sources comprising airguns and explosions was conducted at off Kii Channel, the Nankai Trough, where interplate earthquakes occurred periodically. Using P-wave travel time analyses including non-linear travel time inversion, the depth variation of the crustal structure along the Nankai Trough was revealed. The results show that the subducting depth of the Philippine Sea plate changes at off Kii Channel. The plate is subducting at a shallow location at the west side and at a deep location at the east side; The depth of the plate boundary is about 8 km at the west side and about 9 km at the east side. This result and former studies indicate that the variation of subducting depth exists beneath Kii Channel and off Kii Channel regions. This depth change of the plate boundary might affect the process of interplate earthquake occurrence.

The Nankai trough, which runs along the western Japanese islands, is widely known for its systematic activity of large earthquakes with the Magnitudes greater than 7. According to the fault-plane studies of these earthquakes, a major seismic block boundary is supposed to run off the Kii Peninsula almost perpendicular to the trough axis. A seismic experiment was conducted across this boundary region in order to determine seismic crustal structure. P-wave travel time analyses were applied to the data from the OBSs deployed along the trough axis, including travel time nonlinear inversion. The result shows an abrupt change in thickness of the subducting Philippine Sea Plate's crust at the expected location of the seismic block boundary. The crust is thicker in the east by about two kilometers, and the interface between the overriding and subducting plates runs shallower in the east. These structural features may affect the mode of earthquake occurrence.

In order to study the Earth's structure and subduction zone tectonics, seismic data from the oceanic region are extremely important. The present seismograph distribution in the oceanic region, however, provides a very poor coverage. To improve this poor seismic coverage, a cable OBS system using a retired submarine telecommunication cable is proposed. The GeO-TOC cable runs from Ninomiya, Japan, to Guam through the Izu-Bonin forearc and the Marina Trough. The total length of the cable is 2659 km. An OBS, IZU, using the GeO-TOC cable, was successfully installed at the landward slope of the Izu-Bonin Trench in January 1997. The IZU OBS is located approximately 400 km south of Tokyo. The installation method is similar to repair work on submarine cables. The IZU OBS is equipped with three accelerometers, a hydrophone, a quartz pressure gauge, and a quartz precision thermometer with a few temperature sensors to monitor overheating of the internal electronics. After installation, the voltage increase is 90 V when the current is maintained at a constant 370 mA. Data from accelerometers are digitized by 24-bit A/D converters and sent to Ninomiya at 9600 bps for each component. Hydrophone data are sent to Ninomiya as analog signals using the AM (Amplitude Modulation) method for safety reasons. Hydrophone data are digitized at the shore station. Other slow-rate data are multiplexed and sent to the shore at 9600 bps. The instrument can be controlled by a shore computer. All data will be transmitted from Ninomiya to Tokyo and combined with other existing seismic data. (C) 1998 Elsevier Science B.V. All rights reserved.

For long-term prediction of large earthquakes at plate boundaries, it is important to develop an earthquake cycle model which can numerically simulate the processes of crustal deformation and stress accumulation at and around plate boundaries. We developed a general earthquake cycle model applicable to transform fault zones, subduction zones, and collision zones on the basis of elastic dislocation theory. In this model we regard the periodic occurrence of interplale earthquakes as a perturbation of the hypothetical steady slip motion proceeding on the whole plate boundary. Then the crustal deformation associated with the periodic occurrence of interplate earthquakes is given by the superposition of viscoelastic responses to the steady slip on the whole plate boundary, the steady back slip on the seismic zone, and the sequence of periodic seismic slips. The important points in this earthquake cycle model are as follows. Interaction between adjacent plates can be naturally represented by the increase of discontinuity in tangential displacement across the plate boundary. Viscoelastic stress relaxation of the asthenosphere strongly affects the crustal deformation during interseismic period. Since the stress relaxation time of the lithosphere-asthenosphere system is several hundreds years, we must consider the viscoelastic effects due to a series of large events occurred for the past several hundreds years to estimate crustal deformation correctly. Crustal deformation and stress accumulation due to steady slip on the whole plate boundary can not be neglected.

In order to better understand earthquake generation, tectonics at plate boundaries, and better image the Earth's deep structures, real-time geophysical measurements in the ocean are required. We therefore attempted to use decommissioned submarine cables, TPC-1 and TPC-2. An OBS was successfully linked to the TPC-1 on the landward slope of the Izu-Bonin Trench in 1997. The OBS detected co-seismic and gradual changes during a Mw 6.1 earthquake just below the station at 80 km depth on November 11, 1997. A pressure sensor co-registered a change equivalent to 50 cm sea-level change. This suggests a high possibility detecting silent earthquakes or earthquake precursors if they exist. A multi-disciplinary geophysical station has been developed for deep-sea door using TPC-2 since 1995. The station comprises eight instrument sets: broadband seismometers, geodetic measurements, hydrophone array, deepsea digital camera, CTD, etc. These activities are examples that decommissioned submarine cables can be great global resources for real-time cost-effective geophysical measurements on a deep-sea floor.

Bathymetric data from south of Hokkaido obtained during a cruise of RN Hakuho-Maru are summarized, and their correlation with earthquake occurrence is discussed. There are structural lineations on the seaward slope of the Kuril Trench, oblique to the Kuril Trench axis and parallel to the magnetic lineations in the Pacific plate. The structural lineations comprise horst-grabens generated by normal faulting. This suggests that Cretaceous tectonic structures originating at the spreading centre affect present seismo-tectonics around the trench axis. The structural-magnetic relation is compared to the case of the Japan Trench. North-east of the surveyed area, there are two major fracture zones (Nosappu Fracture Zone and Iturup Fracture Zone) that divide the oceanic plate into three segments. If the fracture zones (PZ) and the zone of paleo-mechanical weakness, represented by magnetic lineations, can control the direction of normal faults at a trench, the extent of the resulting topographic roughness on the seaward slope of the trench would be different across an FZ because of the differences in ages. By studying recent large earthquakes occurring in the south Kuril region, it is shown that several main-aftershock distributions for large earthquakes in this region are bounded by the Nosappu FZ and the Iturup FZ. Two models (Barrier model and Rebound model) are presented to interpret earthquake occurrence near the south Kuril Islands. The Barrier model explains seismic boundaries seen in several examples for earthquake occurrence in the south Kuril regions. The fracture zone forming the boundary of two segments with different magnetic lineations is also the boundary of two different normal fault systems on their ocean bottom, and the difference in sea-bottom roughness between two normal fault systems should affect the seismic coupling at a plate interface. Due to the difference of seismic coupling, earthquake occurrence is controlled by an FZ and then the FZ acts as a seismic boundary (Barrier model). Existing normal faults created by plate bending of subducting oceanic plate should rebound after its subduction (Rebound model). This rebound of normal faults may cause intraplate earthquakes with a high-angle reverse-fault mechanism such as the 1994 Shikotan Earthquake. The energy released by an intraplate earthquake generated by normal-fault rebounding is not directly related to that of interplate earthquakes such as low-angle thrust earthquakes. It is a reason why large earthquakes occurred in the same region during a relatively short period.

The 1995 Amami-Oshima-Kinkai Earthquake occurred near the Nansei-Shoto Trench where the upheaval area of the Philippines Sea plate subducts beneath the Nansei-Shoto islands. The main shock was <I>M</I><SUB>JMA</SUB> 6.6 and its largest aftershock was <I>M</I><SUB>JMA</SUB> 6.5. The aftershock distribution for these two events by Yamada <I>et al</I>. (1996) corresponds to two distinct and nearly vertical fault zones. The focal mechanisms obtained by Kikuchi (1996) are consistent to the aftershock distribution.<BR>The authors propose that the seamount found beneath the trench-continental-slope indirectly triggered this earthquake activity. If a subducting oceanic plate is normal oceanic denser than an overriding island arc, the oceanic plate should be faulted near vertically priori to the plate subduction by horizontally tensional force due to plat bending. On the other hand, an oceanic plate with seamounts or an oceanic plateau lighter than a normal oceanic plate, might resist to plate subduction due to its small density and delaying normal faultings might occur in the subducting oceanic plate. The delaying normal faultings between a subducting seamount and a preceding normal portion of the oceanic plate can compensate the subduction process. The compressional convergence margin such as the Nankai Trough, however, may not generate such normal faultings due to the nature of stress field.<BR>The low seismicity area existing across the trench axis is also seen both in this aftershock activity and ISC hypocenters. This is the same result as those in other regions. This might imply low earthquake potential for this portion of plate interface due to the existence of low density sediments and water contained in the sediments.

On October 18, 1995, an earthquake with magnitude <I>M</I>j 6.7 occurred in the central part of Ryukyu Arc at 28°1.7'N, 130°23.0' E and 38 km in depth. The next day, another earthquake with magnitude <I>M</I>j 6.6 occurred in the same area. The hypocenter parameters were determined by Japan Meteorological Agency (J.M.A.). For the purpose of investigating the aftershock activities finely, we put an array of twenty-two ocean bottom seismometers (OBSs) covering the area of aftershocks. The OBS observation started on October 28, ten days after the occurrence of the <I>M</I>j 6.7 event, and continued about four weeks.<BR>In this study, we relocated hypocenters of the aftershocks using the data of eighteen OBSs. The results give us detailed view of the hypocenter distribution as follows : <BR>Most of aftershocks occurred (1) on the fault of <I>M</I>j 6.7 event, and (2) in the focal area of <I>M</I>j 6.6 (October 19, 1995). Considering the distributions of (1) and (2), we identified one of the nodal planes of the CMT solutions as the fault plane of <I>M</I>j 6.7 event and <I>M</I>j 6.6 event. For both events, the strike is almost parallel to the trench axis. Dip angle of the <I>M</I>j 6.7 event is almost vertical. Therefore, it is obvious that this event is not a low-angle thrust-type interplate earthquake but an intraplate event.

An earthquake of <I>M</I><SUB>JMA</SUB> 6.7 occurred at 19 : 37 on October 18, 1995 at about 50 km southeast of Kikai-jima in the central part of Ryukyu-islands, Japan. About 16 hours after this event, another large earthquake of <I>M</I><SUB>JMA</SUB> 6.6 occurred in the same area. We relocate hypocenters of two large events and their aftershocks using JMA data by considering the station corrections, which are obtained by joint using the temporary observation data of ocean bottom seismometers (OBS) and JMA data. The relocated hypocentral distribution obtained by joint using of OBS and JMA data show the possibility that two large events occurred on the adjacent faults. The events during four days after the <I>M</I><SUB>JMA</SUB> 6.7 event occurred in the region of about 70 km in length in NNE-SSW direction and the epicentral area is divided into two sub-regions, NNE and SSW sub-regions. The aftershock activity of the <I>M</I><SUB>JMA</SUB> 6.7 event is predominant in the SSW sub-region until 7 : 00 a.m. on the next day. After 7 : 00 a.m., the activity increases in the NNE sub-region and the <I>M</I><SUB>JMA</SUB> 6.6 event at 11 : 41 a.m. on October 19 and its aftershocks occurred in this region. On the other hand, the activity in SSW sub-region becomes low in this period. Such the alternative rise and fall activities in two sub-regions are thought to be caused by the interraction of stresses on the different faults. Therefore, the present activity is concluded to be constituted by two foreshockmainshock-aftershock sequences occurred on the adjacent faults in a short time interval.

We have done a numerical simulation of tectonic loading at transform plate boundaries with a lithosphere-asthenosphere coupling model subject to steady relative plate motion. In this model a seismic fault is represented by a locked patch with a finite length on an infinitely long vertical dislocation surface cutting the entire lithosphere. Through the numerical simulation we have obtained the following results. Stress accumulation on the fault is partly due to viscous drag at the base of the lithosphere (base loading) and partly due to dislocation pile-ups at horizontal edges of the fault (edge loading). If the viscosity of the asthenosphere is less than 10(18) Pa s, the base loading is not effective, since its rate exponentially decays very soon. If the Viscosity of the asthenosphere is greater than 10(20) Pas, both types of loading are effective, and then the stress accumulation rate in the early stage of loading is formally written as d tau/dt = V-pl(a + b/L), where V-pl is a rate of relative plate motion and L is a fault length. When the fault length L is small, the effect of edge loading (the second term) is dominant, and the stress accumulation rate is in proportion to the inverse of L. When the fault length L is large, the effect of base loading (the first term) becomes dominant, and the stress accumulation rate is independent of L. It is well-known that the general L-cubed dependence of seismic moment M-0 breaks for large interplate earthquakes. This break in the moment-length relation can be ascribed to difference in loading mechanism between small and large earthquakes. From the results of numerical computation, if interplate earthquakes have the same stress drop regardless of their fault lengths, we may derive a L-squared dependence of Mo for large interplate earthquakes and a linear L-dependence of M-0 for very large interplate earthquakes.

We conducted two seismic surveys in the Yap region to investigate tectonic activity in this area. One was conducted in the northern Yap Trench for ten days using ocean bottom seismometers, the other was conducted on the Yap Islands for eight months using a small array. From these observations we,detected many earthquakes beneath the Yap region and found a characteristic pattern of seismicity. Many earthquakes occurred in the inner trench slope, no seismicity was observed in the axial region of the trench, and a few earthquakes occurred in the outer trench slope. This pattern is similar to the typical pattern in active subduction zones. This result indicates that the Yap Trench is still actively subducting. We did not detect earthquakes deeper than 40 km during the observations.

We conducted three weeks of seismic observations at the Rodriguez Triple Junction (RTJ) in the Indian Ocean using 18 ocean-bottom seismometers over an area of 90km x 90km. We identified six teleseismic events and obtained significant anomalies in the relative travel-time residuals of the teleseismic P-waves. The residuals are positive at the RTJ, the northern part of the Southeast Indian Ridge (SEIR), and the eastern side of the Central Indian Ridge (CIR), and are negative at and around the Southwest Indian Ridge (SWIR). It is suggested that there is a relatively hotter mantle under the triple junction and the northern part of the SEIR segment, along-axis variations in mantle temperature along the SEIR segment, and cooler mantle under the SWIR segment. The CIR segment has an asymmetrical distribution of relative residuals.

An OBSH (ocean bottom seismometer with hydrophone) was deployed just 1 m from a hot thermal vent on the summit region of a seamount located 18 km east of the south Mariana Trough axis to record continuous data on hydrothermal activity. Although the seismicity beneath the seamount was very weak, numerous pulse shape (pressure) events showing sudden pressure decrease were observed on the hydrophone channel. The characteristic period of the pressure events varied from 20 seconds to 1 minute. After the submersible Shinkai 6500 relocated the OBSH 10 m south of the vent, where no thermal vents existed, the number and amplitude of pressure events drastically decreased. This suggests that pressure events were generated by the hot water upwelling from the thermal vent. The pressure events appear to be related to the ocean tide. This experiment indicates that observations using OBSHs are an effective means of monitoring hydrothermal activity.

A one month microearthquake survey was conducted at Iheya Knolls in the middle Okinawa Trough in 1992 to reveal microearthquake characteristics of back‐arc rifting. We deployed five ocean bottom seismometers and observed over 1700 events during 30 days. Hypocenters were determined for 48 earthquakes with good resolution. From this survey, we conclude that the survey area with 100km × 100km has high seismicity, many earthquakes occur in the western part of Iheya Knolls, and few occur in the central part. The earthquakes are distributed in an en echelon shape and throughout the entire depth of the crust. There is no evidence for the existence of transform faults, suggesting that the middle Okinawa Trough is in a stage of back‐arc rifting. Copyright 1994 by the American Geophysical Union.

Plate subduction zones are typically characterized by the patterns of surface topography and gravity anomalies consisting of island‐arc high, trench low, and outer‐rise gentle high. These patterns are stable on the time scale of 106–107yr. At some subduction zones, regardless of its age, steady uplift of marine terraces formed by eustatic sea‐level changes during the last 105 yr can be also observed. The stable patterns of topography and gravity anomalies and the steady uplift of marine terraces seem to contradict each other. We constructed a kinematic model which could explain the evolution process of island are‐trench systems and demonstrated that this puzzle could be solved by considering the effect of accretion at plate boundaries. In our model the lithosphere‐asthenosphere system is represented by a stratified visco‐elastic half‐space under gravity, which consists of a high‐viscosity surface layer and a low‐viscosity substratum. The rheological properties of both layers are assumed to be a Maxwell fluid in shear and an elastic solid in bulk. Interaction between oceanic and continental plates is represented by the steady increase of discontinuity in tangential displacement across the plate boundary. The other essential factors considered in our model are accretion of oceanic sediments at plate boundaries, erosion on land, and sedimentation on inner trench wall. We computed the evolution process of island arc‐trench systems by using the kinematic model and obtained the following results: the island are‐trench systems grow at nearly constant rates in the early stage of plate subduction. Therefore, at young subduction zones, we can generally expect the steady uplift of marine terraces. The stable patterns of topography and gravity anomalies are formed within several million years after the initiation of plate subduction. When the accretion process continuously proceeds at plate boundaries, these stable patterns gradually migrate seaward as a whole. In such a case, we can expect the steady uplift of marine terraces even at old subduction zones. We demonstrate that our model can consistently explain the observed surface topography, gravity anomalies, and uplift rates of marine terraces at subduction zones. Copyright © 1993, Wiley Blackwell. All rights reserved

In southwest Japan, where the Philippine Sea Plate is descending beneath the Eurasian Plate at the Nankai Trough, we can observe cyclic crustal movement related to the periodic occurrence of interplate earthquakes with the time interval of 102yr, steady uplift of the marine terraces formed by eustatic sea‐level changes for the last 105yr, and gradual evolution of the island arc‐trench system through the last 4 X 106yr. We demonstrate that these phenomena with very different characteristic time‐scales can be consistently explained by a single‐plate subduction model. In our model, the lithosphere‐asthenosphere system is represented by a stratified viscoelastic half‐space under gravity, consisting of a high‐viscosity surface layer and a low‐viscosity substratum, and interaction between oceanic and continental plates by steady slip motion over the whole plate boundary and its perturbation associated with the periodic occurrence of earthquakes. The effects of accretion of oceanic sediments at plate boundaries, erosion on land, and sedimentation on inner trench walls are also considered in the model. From comparison of theoretical results with observed data we obtained the following conclusions valid for young subduction zones: observed deformation cycles cannot be explained by a simple rebound model in which the effect of steady‐plate subduction is ignored. The steady‐plate subduction brings about steady uplift of marine terraces. The present patterns of surface topography and gravity anomalies are held nearly stable by the balance of erosion (sedimentation) rates and substantial growth rates. Copyright © 1992, Wiley Blackwell. All rights reserved

A kinematic model for the earthquake cycle at convergent plate boundaries has been constructed on the basis of dislocation theory. We model the lithosphere‐asthenosphere system by a stratified semi‐infinite medium under gravity, consisting of an elastic surface layer, an intervening layer with Maxwell viscoelasticity, and an elastic substratum. The steady motion of plate convergence is naturally represented by uniform slip at a constant rate on the upper boundary of the descending oceanic plate. The occurrence of interplate earthquakes is regarded as a perturbation of the steady state plate motion, and represented by a periodic sequence of step slips on a finite seismic zone of the plate interface. Based on dislocation theory we can show that the steady slip on the interface deeper than the lithosphere‐asthenosphere boundary does not contribute to surface deformation in the steady state. Accordingly the surface deformation associated with the earthquake cycle is given by the superposition of viscoelastic responses to the steady slip on the interface shallower than the lithosphere‐asthenosphere boundary, the steady backslip on the seismic zone, and the periodic sequence of seismic slips. We have computed cyclic patterns of vertical displacements at the free surface for two representative cases, taking account of gravity effects. The patterns of vertical displacements differ notably in the latter stage of the cycle depending on whether or not the seismic zone extends through the entire thickness of the lithosphere. After the completion of one earthquake cycle, our model yields a certain amount of permanent deformation resulting from the steady plate convergence. The pattern of the permanent deformation, that is characterized by steep uplift on the continental side, sharp subsidence at the plate boundary, and gentle uplift on the oceanic side, is irrespective of the cyclic process of stress accumulation and release repeated on the seismic zone. The earthquake cycle model developed here provides a possible explanation for the formation of earthquake‐origin marine terraces. Copyright © 1989, Wiley Blackwell. All rights reserved

A subduction model that interprets the whole pattern of topography and gravity anomalies across the island arc-trench system has been constructed on the basis of dislocation theory. Models the lithosphere-asthenosphere system by a gravitating, stratified half-space, which consists of an elastic surface layer, an intervening layer with Maxwell viscoelasticity, and an elastic substratum. Numerical results show that the pattern of vertical motion at subduction zones is characterized by steep uplift on the continental side, sharp subsidence at the plate boundary, and gentle uplift on the oceanic side. This pattern bears a striking resemblance to the general pattern of topography across island arc-trench systems. The whole pattern of topography and gravity anomalies across the S Kurile arc is well matched by the subduction model with a lithosphere thickness of 50 km and a dip angle of 30o at depths.-from Authors