Subcrustal Seismicity Research Articles

Seismotectonic analysis of the 2021 Damasi-Tyrnavos (Thessaly, Central Greece) earthquake sequence and implications on the stress field rotations

In early March 2021, three shallow earthquakes, two mainshocks with M6.3 and M6.0 and one major aftershock with M5.6 impacted both the mountainous Damasi-Tyrnavos region (northern Thessaly, Greece) and the adjacent Plio-Quaternary basin. Each major event was followed by rich aftershock activity recorded by local and regional seismographs and accelerographs. Herein, we present a comprehensive analysis of the seismic sequence, from its foreshock activity starting on 28 February, 2021 and for a period of two months using new high-resolution catalogues of relocated earthquakes and hundreds of focal mechanisms. The results indicate that the aftershocks form a zone that spans ~50 km NW-SE, while focal depths range between 5 and 15 km. More than 400 focal mechanisms, computed for events with M≥ 2.5, mainly exhibit normal faulting in a NW-SE direction, while WNW-ESE to E-W normal faulting is also evidenced, in particular after the occurrence of the last major event on 12 March. The stress-field was reconstructed on a local and broader scale by inverting focal mechanism data, revealing a rotation of the σ3 axis trend from NNE-SSW, in the Damasi-broader region, to NW-SE northwards, to the region of Kozani-Grevena that hosted an Mw = 6.6 shallow mainshock in 1995. Subcrustal seismicity, present beneath those areas, implies that large-scale tectonics and plate dynamics are likely involved in the deformation of the upper crust. Coulomb stress transfer after the 3 major events of the 2021 Damasi-Tyrnavos sequence reveals that stress-loaded areas include those where most aftershocks were triggered. The analysis provides implications to the seismic hazard of the activated area, as a major NW-SE active normal fault close to Larissa city became stress-loaded, constituting a possible candidate source for significant future earthquakes.

The TsuJal Amphibious Seismic Network: A Passive-Source Seismic Experiment in Western Mexico

The geodynamic complexity in the western Mexican margin is controlled by the multiple interactions between the Rivera, Pacific, Cocos, and North American plates, as evidenced by a high seismicity rate, most of whose hypocenters are poorly located. To mitigate this uncertainty with the aim of improving these hypocentral locations, we undertook the TsuJal Project, a passive seafloor seismic project conducted from April to November 2016. In addition to the Jalisco Seismic Network, 10 LCHEAPO 2000 Ocean Bottom Seismometers (OBSs) were deployed by the BO El Puma in a seafloor array from the Islas Marías Archipelago (Nayarit) to the offshore contact between the states of Colima and Michoacan. We located 445 earthquakes in four or more OBSs within the deployed array. Most of these earthquakes occurred in the contact region of the Rivera, Pacific, and Cocos plates, and a first analysis suggests the existence of three seismogenic zones (West, Center, and East) along the Rivera Transform fault that can be correlated with its morphological expression throughout the three seismogenic zones. The seismicity estimates that the Moho discontinuity is located at 10 km depth and supports earlier works regarding the West zone earthquake distribution. Subcrustal seismicity in the Central zone suggests that the Intra-Transform Spreading Basin domain is an ultra-low spreading ridge. A seismic swarm occurred during May and June 2016 between the eastern tip of the Paleo-Rivera Transform fault and the northern tip of the East Pacific Rise-Pacific Cocos Segment, illuminating some unidentified tectonic feature.

Earthquake Source Properties of a Lower Crust Sequence and Associated Seismicity Perturbation in the SE Carpathians, Romania, Collisional Setting

Romanian seismicity is mainly confined to the Eastern Carpathians Arc bend (ECAB), where strong subcrustal earthquakes (magnitude up to 7.9) are generated in a narrow lithospheric body descending into the mantle. The seismic activity in the overlying crust is spread over a larger area, located mostly toward the outer side of the ECAB. It is significantly smaller than subcrustal seismicity, raising controversies about possible upper mantle-crust coupling. A significant earthquake sequence took place in the foreland of the ECAB triggered on 22 November 2014 by a mainshock of magnitude 5.7 (the greatest instrumentally recorded earthquake in this region) located in the lower crust. The mainshock triggered a significant increase in the number of small-magnitude events spread over an unusually large area in the ECAB. The paper’s goal is to compute the source parameters of the earthquakes that occurred during the aforementioned sequence, by empirical application of Green’s function and spectral ratio techniques. Fault plane solutions are determined using multiple methods and seismicity evolution at regional scale is investigated. Our results highlight a still active deformation regime at the edge of the EE Craton, while the source parameters reveal a complex fracture of the mainshock and a very high-stress drop.

Deep structure of the Hellenic lithosphere from teleseismic Rayleigh-wave tomography

SUMMARYThis research provides new constraints on the intermediate depth upper-mantle structure of the Hellenic lithosphere using a three-step Rayleigh-wave tomography. Broadband waveforms of about 1000 teleseismic events, recorded by ∼200 permanent broad-band stations between 2010 and 2018 were acquired and processed. Through a multichannel cross-correlation technique, the fundamental mode Rayleigh-wave phase-velocity dispersion curves in the period range 30–90 s were derived. The phase-velocities were inverted and a 3-D shear velocity model was obtained down to the depth of 140 km. The applied method has provided 3-D constraints on large-scale characteristics of the lithosphere and the upper mantle of the Hellenic region. Highlighted resolved features include the continental and oceanic subducting slabs in the region, the result of convergence between Adria and Africa plates with the Aegean. The boundary between the oceanic and continental subduction is suggested to exist along a trench-perpendicular line that connects NW Peloponnese with N. Euboea, bridging the Hellenic Trench with the North Aegean Trough. No clear evidence for trench-perpendicular vertical slab tearing was resolved along the western part of Hellenic Subduction Zone; however, subcrustal seismicity observed along the inferred continental–oceanic subduction boundary indicates that such an implication should not be excluded. The 3-D shear velocity model supports an N–S vertical slab tear beneath SW Anatolia that justifies deepening, increase of dip and change of dip direction of the Wadati-Benioff Zone. Low velocities found at depths <50 km beneath the island and the backarc, interrelated with recent/remnant volcanism in the Aegean and W. Anatolia, are explained by convection from a shallow asthenosphere.

Analysis the 1978–2008 crustal and sub-crustal earthquake catalog of Vrancea region

Abstract. The crustal and sub-crustal seismicity that occurred in the Vrancea area from 1978 to 2008 is investigated. Due to quarry blast contamination, the analysis of the crustal seismicity was restricted to events that occurred between 20 km and 60 km, considering sub-crustal events as all those that occurred at depths larger than 60 km. The completeness magnitude was estimated between 2.3 and 2.5 for the crustal seismicity and between 2.9 and 3.1 for the sub-crustal one. The Gutenberg-Richter b-values show clear differences between crustal and sub-crustal seismicity, with values around 0.9 and 1.3–1.4, respectively. The analysis performed in this study represents a preliminary issue for the seismic hazard assessment of surrounding regions like Moldova, for which the earthquakes generated in Vrancea represent an important threat.

Detailed geometry of the subducting Indian Plate beneath the Burma Plate and subcrustal seismicity in the Burma Plate derived from joint hypocenter relocation

With the aim of delineating the subducting Indian Plate beneath the Burma Plate, we have relocated earthquakes by employing teleseismic P-wave arrival times. We were able to obtain the detailed geometry of the subducting Indian Plate by constructing iso-depth contours for the subduction earthquakes at depths of 30–140 km. The strikes of the contours are oriented approximately N-S, and show an “S” shape in map view. The strike of the slab is N20°Eat25°N, but moving southward, the strike rotates counterclockwise to N20°Wat20°N, followed by a clockwise rotation to a strike of N10°E at 17.5°N, where slab earthquakes no longer occur. The plate boundary north of 20°N might exist near, or west, of the coast line of Myanmar. The mechanisms of subduction earthquakes are down-dip extension, and T axes are oriented parallel to the local dip of the slab. Subcrustal seismicity occurs at depths of 20–50 km in the Burma Plate. This activity starts near the 60-km-depth contour of the subduction earthquakes and becomes shallower toward the Sagaing Fault, indicating that this fault is located where the cut-off depth of the seismicity becomes shallower.

Time-clustering analysis of the 1978–2008 sub-crustal seismicity of Vrancea region

Abstract. The analysis of time-clustering behaviour of the sub-crustal seismicity (depth larger than 60 km) of the Vrancea region has been performed. The time span of the analyzed catalogue is from 1978 to 2008, and only the events with a magnitude of Mw ≥ 3 have been considered. The analysis, carried out on the full and aftershock-depleted catalogues, was performed using the Allan Factor (AF) that allows the identificatiion and quantification of correlated temporal structures in temporal point processes. Our results, whose significance was analysed by means of two methods of generation of surrogate series, reveal the presence of time-clustering behaviour in the temporal distribution of seismicity data of the full catalogue. The analysis performed on the aftershock-depleted catalogue indicates that the time-clustering is associated mainly to the aftershocks generated by the two largest events occurred on 30 August 1986 (Mw = 7.1) and 30 May 1990 (Mw = 6.9).

Recent seismicity and crustal stress field in the Lucanian Apennines and surrounding areas (Southern Italy): Seismotectonic implications

We analyzed the instrumental seismicity of Southern Italy in the area including the Lucanian Apennines and Bradano foredeep, making use of the most recent seismological data base available so far. P- and S-wave arrival times, recorded by the Italian National Seismic Network (RSNC) operated by the Istituto Nazionale di Geofisica e Vulcanologia (INGV), were re-picked along with those of the SAPTEX temporary array deployed in the region in the period 2001–2004. For some events located in the upper Val d'Agri, we also used data from the Eni-Agip oil company seismic network. We examined the seismicity occurred during the period between 2001 and 2006, considering 514 events with magnitudes M ≥ 2.0. We computed the V P / V S ratio obtaining a value of 1.83 and we carried out an analysis for the one-dimensional (1D) velocity model that approximates the seismic structure of the study area. Earthquakes were relocated and, for well- recorded events, we also computed 108 fault plane solutions. Finally, using 58 solutions, the most constrained, we computed regional stress field in the study area. Earthquake distribution shows three main seismic regions: the westernmost (Lucanian Apennines) characterized by high background seismicity, mostly with shallow hypocenters, the easternmost below the Bradano foredeep and the Murge with deeper and more scattered seismicity, and finally the more isolated and sparse seismicity localized in the Sila Range and in the offshore area along the northeastern Calabrian coast. Focal mechanisms computed in this work are in large part normal and strike-slip solutions and their tensional axes ( T-axes) have a generalized NE–SW orientation. The denser station coverage allowed us to improve hypocenters determination compared to those obtained by using only RSNC data, for a better characterization of the crustal and subcrustal seismicity in the study area.

Microseismicity and active deformation of Messinia, SW Greece

By deploying a 30 3-component digital seismic array in the Messiniakos gulf and the surrounding region, we recorded for a period of 45 days the microseismic activity. With a minimum of six records per event, we located 1121 earthquakes corresponding to an average of 20 events per day. For the hypocenter location we used a local velocity model adopted to two controlled source seismic experiments. Within the array, traveltime residuals were within ± 0.2 s and the epicentral accuracy in the order of ± 2 km, while the hypocentral one is twice this value. Correlation of the seismicity with the tectonic elements indicated that most of the NW-SE oriented faults are active with strike-slip movement along this orientation and extension perpendicular to it. The neogene basins of Messini, Meligalas and Megalopolis are seismically very active and their eastern flanks are delineated by higher seismic activity than their western ones. This indicates that the basins are asymmetric with master faults defining their eastern-northeastern flanks. This hypothesis is supported by the asymmetric structure mapped at the offshore Messiniakos basin as densely spaced high resolution reflection seismic profiles have revealed. The western margins of the basins are less deformed and the seismic activity is dispersed over several minor NW-SE faults. Since the NW-SE striking faults onshore are truncated by major NE-SW oriented ones, their overall length is shortened, reducing their seismic potential and capacity to store large stresses that could produce events above Ms6.1. Offshore western Messinia, in the Ionian Sea, the size and activity of the faults is significantly larger and prone to develop events of larger magnitudes. Subcrustal seismicity indicates a deepening of the foci to the east-northeast.

Rheological control of subcrustal seismicity in the Apennines subduction (Italy)

Different rheological behavior of continental vs oceanic lithosphere can account for the shallower and lower subcrustal seismicity below the northern Apennines with respect to the deeper and more intense seismicity below the Calabrian arc in the south along the same subduction zone. We show that, for equal temperature distribution, oceanic slabs are predicted to display brittle behavior down to depths higher than those predicted for a continental slab. This opposite rheology is likely amplified by the higher strain rates and cooler slab temperatures in the south due to the faster subduction rollback and the oceanic composition of the downgoing lithosphere. Therefore the paucity of seismicity in the central‐northern Apennines might not be related to the absence of the slab, but rather to its continental nature and to lower strain rates.

Dynamic modelling of stress accumulation in central Italy

We develop a viscoelastic dynamic model to calculate the stress distribution in central Italy induced by different geodynamic processes likely active in the region: Africa‐Eurasia convergence, Adriatic underthrusting/subduction and asthenospheric upwelling underneath Tuscany. Model results are compared with the hypocentral distribution of earthquakes and with the orientation of P and T axes available for the region. The first order characteristics of the seismotectonic observations in central Italy are: subcrustal seismicity, the pair extension‐compression in Tuscany and in the outer Apennines, and the flexural behaviour evident in long wavelength orientation of T axes. These features are reproduced by the combination of two processes: underthrusting of the Adriatic plate under the Apennines and asthenospheric upwelling underneath Tuscany. The comparison with available data does not give insights about slab pull as an ongoing process.

Possible Cause-Effect Relationships Between Vrancea (Romania) Earthquakes and Some Global Geophysical Phenomena

The possibility that the Earth's tides are a triggering factor of Vrancea subcrustal earthquakes is investigated in the first part of this paper. A possible correlation between Vrancea subcrustal earthquakes and geomagnetic jerks is demonstrated in the second part. The last part of the paper presents a number of results concerning a possible relationship between the regularities of strong Vrancea subcrustal seismicity and the Chandler nutation parameters. An attempt is made to integrate all of these phenomena in a more general framework that takes into account physical processes in the Earth mantle and core. A long-term prediction of the next strong Vrancea earthquake is finally attempted.

Subcrustal Seismicity beneath the Southern Part of the Chugoku Region, Japan.

Subcrustal Seismicity beneath the Southern Part of the Chugoku Region, Japan.

Evaluation of stresses in two geodynamically different areas: Stable foreland and extensional backarc

Areas which are geodynamically different have different behaviors both in their thermal regime and seismic activity. A stable area has a geotherm which can be considered as standard, extensional and compressional areas have, respectively, high and low temperature gradients. The Italian region includes different geodynamical areas and all such situations are present. We consider the Apulian platform as an example of a stable area and the Tuscany-Latium as an example of an extensional area. For both of them the present geotherms are calculated, taking into account, for the Tuscany-Latium, its thermal history. Assuming that each region is subject to a constant strain rate, the stresses are calculated as functions of depth and time. The rheological behavior is assumed to be linear viscoelastic, with viscosity dependent on temperature and elastic parameters dependent on lithology. The geothermal profile and the rheological structure of the lithosphere remarkably affect the processes of stress accumulation which control the distribution of seismic activity. The abrupt decrease of the temperature gradient at the Moho produces considerably higher stress values with respect to the case of uniform gradient, thus favoring subcrustal seismicity. In the case of a standard temperature gradient, subcrustal seismicity is predicted and a gap in seismicity, indicating a soft intracrustal layer, exists if there is a discontinuity in rheology. By contrast, in the case of a high-temperature gradient, subcrustal seismicity is not to be expected, even in the presence of a discontinuity in rheology, since subcrustal temperatures are already too high to permit a sufficient stress accumluation.

Subcrustal microearthquake seismicity and fault plane solutions beneath the Hellenic Arc

Subcrustal seismicity recorded in the southern Aegean sea during a 7‐week microearthquake study was low compared with shallow seismicity. Most intermediate depth seismicity occurred beneath the western and eastern ends of the Hellenic arc. This distribution confirms that a slab of lithosphere is being subducted at a very shallow (<15°) angle for 200 km beneath the western end (Peloponnese) but more steeply beneath the eastern end (Dodecanese). We could locate only one intermediate depth event beneath the central pan of the arc, where teleseimically located intermediate depth earthquakes also are infrequent. T axes for most of the 22 focal mechanisms of subcrustal earthquakes are roughly parallel to the local dip direction of the seismic zone. Between depths of 40 and 80 km, the mechanisms are more confused than at greater depth, perhaps because some of these earthquakes did not occur within the downgoing slab. Earthquakes deeper than 80 km, and within the subducted slab, have nearly horizontal P axes that trend NNE‐SSW in the eastern part and NNW‐SSE in the western part of the arc. These deeper mechanisms show horizontal P axes along strike, perhaps in response to the contortion of the slab or to the westward motion of Turkey, as well as lengthening downdip, probably in response to gravity acting on excess mass in the slab. Thus the short slab, both downdip and along strike, subducting beneath the Aegean is subjected to a more complex set of forces than the long slabs of the Pacific.

Deep earthquakes beneath central Taiwan: Mantle shearing in an arc‐continent collision

The island of Taiwan owes its existence to the collision of an island arc on the Philippine Sea Plate with the Eurasian continental shelf. Most of the earthquakes in the mantle beneath Taiwan clearly are related to one of these plates subducting beneath the other, but beneath the central part of the island there is a zone of subcrustal seismicity that is not obviously related to any subduction zone. We investigated the uncertainties in the locations of these anomalous earthquakes by rereading P and S wave arrival times and computing information density functions for several of these events. The seismic zone itself is then defined through a joint probability function incorporating the entire data set. Individual information density functions show that several events are likely to be located at depths exceeding 60 km, with two events exceeding 80 km depth. The joint probability function shows a narrow zone of seismicity, about 20 km wide along strike, plunging to the east. Combining these results with evidence from surface geology, focal mechanisms and subsurface tomography, we suggest that these events are caused by an eastward displacement of the Eurasian mantle north of 24°N which juxtaposes it against colder subducted crust to the south. Moreover, the heat produced by this type of shearing in the mantle must be insignificant for brittle failure to exist at these depths.

Seismic and volcanic activities and aseismic movements as plate motion components in the Aegean area

Slip rates varying between 0.22 and 2.18 cm yr −1 are determined from seismicity in several segments of the Aegean fore-arc and back-arc regions. The rate along the main zone of plate interaction is about 0.5 cm yr −1 while aseismic slip rate, equal to 2.5 or 0.5 cm yr −1 depending on the plate motion rate assumed, indicates that at least 50% of the plate motion is accommodated by creep. These values, with respect to the rate of volcanic eruption (= 0.13 eruptions per year/1000 km of arc length) reinforce the suggestion that active volcanism in subduction zones is related to aseismic subduction. The eruption rate is, however, anomalously low with respect to the degree of plate decoupling (= 0.83 or 0.5) implying that volcanism is not necessarily related to the degree of decoupling and suggesting that the high aseismic slip component of the plate motion is taken up by subcrustal seismicity. A tentative hypothesis is made that the Hellenic subduction zone is of intermediate type between the Marianas and Chile subduction types.

Distribution of Late Cenozoic volcanic vents in the Cascade range: Volcanic arc segmentation and regional tectonic considerations

Spatial, temporal, and compositional distributions of approximately 4000 volcanic vents formed since 16 Ma in Washington, Oregon, northern California, and northwestern Nevada illustrate the evolution of volcanism related to subduction of the Juan de Fuca plate system and extension of the Basin and Range province. Vent data were obtained from published map compilations and include monogenetic and small polygenetic volcanoes in addition to major composite centers. On the basis of the distribution of 2821 vents formed since 5 Ma, the Cascade Range is divided into five segments, with vents of the High Lava Plains along the northern margin of the Basin and Range province in Oregon forming a sixth segment. Some aspects of the Cascade Range segmentation can be related to gross structural features of the subducting Juan de Fuca plate. The orientation of the volcanic front of segments one and two changes from NW in northern Washington to NE in southern Washington, paralleling the strike of the subducting Juan de Fuca plate. Segments one and two are separated by a 90‐km volcanic gap between Mount Rainier and Glacier Peak that is landward of the portion of the subducting plate having the least average dip to a depth of 60 km. A narrow, N‐S trending belt of predominantly andesitic vents in Oregon constitutes a third segment, which is landward of the seismically quiet portion of the subduction zone. The narrowness of this segment may indicate steep dip of the subducting plate beneath the Cascade arc in Oregon. Vents are sparse between segment four (containing the Mount Shasta and Medicine Lake centers) and segment five (containing Lassen Peak), where the Juan de Fuca and Gorda North plates are characterized by differing age, amounts of subcrustal seismicity, and probably geometry. From the relation between seismicity at depth of 60 km and the position of the volcanic front of vents formed since 5 Ma, transitions between subducting‐plate segments of varying geometry likely occur near boundaries between independently defined volcanic segments in northern Oregon and northern California. In the Basin and Range province east of the Cascade arc, volcanism migrated into the region adjacent to the Cascade Range during the interval 5–10 Ma. Since 5 Ma, the impingement of the two provinces is characterized by cessation of basin‐range volcanism in southern Oregon, continuation of basaltic volcanism in northeastern California where the impingement process may not yet be complete, and contraction of the area of mafic volcanism around Mount Shasta, Medicine Lake, and Lassen Peak. In central Oregon where the northern margin of basin‐range volcanism (the High Lava Plains) intersects the Cascade arc, impingement of basin‐range extensional volcanism approximately coincides in time and space with the development of the High Cascade graben between Three Sisters and Mount Jefferson.

Three‐dimensional P and S wave velocity structures beneath Taiwan: Deep structure beneath an arc‐continent collision

The island of Taiwan is a byproduct of one of the few active collisions between a continent, represented by the Eurasian continental shelf, and an island arc, represented by the northern extension of the Luzon arc on the Philippine Sea plate. To understand better the evolution and current tectonics of this collision, we selected 1260 well‐recorded earthquakes from an initial set of 50,000 located by the Taiwan Telemetered Seismograph Network for the determination of one‐ and three‐dimensional P and S wave velocity structures beneath the island. The results for both structures and earthquake relocations reveal that the island is divided tectonically into three distinct zones. In the east, the Philippine oceanic plate is underthrusting the Eurasian plate east of a north‐south boundary that is well defined by both seismic activity and a region of high velocity. In the south, the Eurasian continental plate is underthrusting the Philippine Sea plate south of an east‐west boundary at about 23°N that is sharply defined by both subcrustal seismicity and a zone of relatively low velocities. The dip of the subducted continent is shallow until it reaches the Luzon island arc 50 km east of the main island. Structure under the main part of the island north of 230N reveals a shallow dipping zone of low velocities in the west above 25 km depth that narrows and steepens below that depth to at least 50 km beneath the central range. The dip of this low‐velocity region is outlined by a narrow zone of seismicity that extends to depths of 100 km. This seismic zone lies in a velocity “saddle” and marks an apparent offset in the central low‐velocity region at 24°N. Evidence for the subduction of the Eurasian continent beneath Taiwan therefore exists everywhere beneath the island, and the low‐velocity regions in the mantle support, but do not require, the subduction of some 6–16 km of lower continental crust to depths of at least 50 km. The distinct change in both the velocity structure and the seismic activity at 23°N and the offset of the low‐velocity regions at 24°N argue for abrupt changes in the nature of subduction across these latitudes. These changes may be caused by an interaction of the Luzon island arc with the subducted continental shelf that mimics a similar interaction evident at the surface. Finally, the velocity structure of the subducting Philippine sea plate suggests that the earthquakes beneath 70 km depth are not occurring within the highest‐velocity regions of the plate, but are probably located near the upper edge of the plate. There is also evidence from earthquake locations and velocity structure that suggests that the subducting plate is segmented, and that subduction is presently occurring as far south as 22°N.

Regional variations of amplitudes of ScSp phases observed in the Japanese Islands

ScSp phases are examined with special reference to their amplitude variations over seismograph networks in the Japanese Islands. ScSp of relatively large amplitude are observed along the Pacific side of the Japanese Islands. Amplitude ratios of the ScSp with respect to ScS are explained in terms of ScS-to-P conversion at the upper boundary of a descending Pacific plate which dips at about 30° and whose seismic wave velocity is between 5 and 10 per cent higher than that of the surrounding mantle. Large-amplitude ScSp are observed at stations nearest the Kurile or Japan trench. The amplitude ratios of ScSp/ScS are found to decrease rapidly with distance inland from the trench. An estimate is made of the effects of sharpness of the conversion interface, the incident angle of ScS upon it, and lateral variations of anelasticity within the asthenosphere overlying the descending Pacific plate, upon the ScSp/ScS ratios. We show that the sharpness of the conversion interface is the dominant factor. This result implies that the nature of the ScS-to-P conversion interface varies with distance from the trench or equivalently with depth. The presence of a low-velocity zone in the vicinity of the conversion interface is inferred from the relative polarity between ScSp and ScS observed at the stations on the outermost side of north-eastern Japan. ScSp arrivals from shallower conversions at depths less than about 80 km, observed in central and south-western Japan, are also examined to elucidate the structure of the Philippine Sea plate descending beneath this region. A discussion is made on the correlations between dipping subcrustal seismicity and the ScS-to-P conversion interfaces. Except for the Chugoku region, the ScS-to-P conversion interfaces correlate well with the dipping subcrustal seismicity. Our ScSp observations indicate the presence of a dipping discontinuity in the uppermost mantle beneath the Chugoku region, where no subcrustal earthquakes have been detected.