Search

The P-wave velocity structure of the upper crust beneath a profile ranging from Niikappu to Samani in the southwestern foot of the Hidaka Mountains, Hokkaido, Japan was obtained through analysis of refraction and wide-angle reflection data. The mountains are characterized by high seismicity and a large gravity anomaly. The present profile crosses the source region of the 1982 Urakawa-oki earthquake (Ms 6.8). The length of the profile is 66 km striking northwest and southeast. Along the profile, 64 vertical geophones were set up and 5 shot points were chosen. For each shot, a 400–600 kg charge of dynamite was detonated. The studied area is composed of four major geological belts: Neogene sedimentary rocks, the Kamuikotan belt, the Yezo Group, and the Hidaka belt. The measurement line crosses these geological trend at an oblique angle. The structure obtained is characterized by remarkable velocity variations in the lateral direction and reflects the surface geological characteristics. A thin, high-velocity layer (HVL) was found between low-velocity materials in the central part of the profile, beneath the Kamuikotan Metamorphic Belt, at a depth ranging from 0.5 to 6 km, overthrusting toward the west on the low-velocity materials consisting of Neogene sedimentary rocks, and forming gentle folds. Outlines of the velocity structure of the Hidaka Mountains yielded by other studies have shown a large-scale overthrust structure associated with the collision of the Outer Kurile and the Outer Northern Honshu Arcs. The shallow velocity structure inferred by the present study showed a similar (although small scale) overthrust structure. The obtained structure shows that the composite tectonic force, comprising westward movement of the Outer Kurile Arc and northward movement of the Outer Northern Honshu Arc, plays an important role in the evolution of the tectonic features of the crust and upper mantle in a wide depth range beneath the Hidaka Mountains.

The Kitakami massif, which is located in the eastern part of Northern Honshu, Japan, is composed of two geological units. The northern Kitakami terrane is characterized as a Jurassic accretionary complex, while the southern Kitakami terrane consists of pre-Silurian basement and Silurian-lower Cretaceous marine sediments. The boundary region of these two units, called the Hayachine tectonic belt (HTB), is composed of mafic to ultramafic rocks. The Kitakami massif experienced intense granitic intrusions in the Cretaceous. We present a detailed crustal structure model for the eastern part of the massif derived from an extensive seismic refraction experiment conducted on a 194-km N-S line. The uppermost crust is covered with a very thin (0.5–1 km) surface layer with a velocity of 3.1–5.4 km/s. The velocity structure below this layer shows remarkable lateral variation. In the northern Kitakami terrane the P wave velocity and Vp/Vs at the top of the basement are 5.85–5.95 km/s and 1.68–1.70, respectively. The seismic attenuation in this region is high (Qp = 150–200 and Qs = 70–100). In contrast, the uppermost crust in the southern Kitakami terrane is characterized by a high P wave velocity (6.05–6.15 km/s) and Vp/Vs (1.74–1.77). The Qp and Qs also show high values of 300–400 and 150–200, respectively. Such a structural difference persists to 14-to 16-km depth, at which the P wave velocity increases to 6.45 km/s. The low velocity and high attenuation in the northern Kitakami terrane represent a highly deformed structure of the accretionary complex. The high P wave velocity and Vp/Vs in the southern Kitakami terrane indicate the relatively mafic crustal composition, which may result from the fragment of the oceanic crust incorporated by the accretion process or the uplifting in the latest Jurassic-early Cretaceous. A midcrustal interface determined from wide-angle reflections shows an abrupt southward depth decrease from 25 to 20 km under the HTB. The P wave velocity and Vp/Vs between 14- and 16-km depth and the midcrustal interface are 6.45–6.55 km/s and 1.74–1.78, respectively. The Moho depth under the northern Kitakami terrane decreases southward from 34 to 32 km. In the southern Kitakami terrane the Moho dips slightly southward. The P wave velocity and the Vp/Vs ratio in the lower crust are 6.9–7.0 km/s and 1.75–1.76, respectively. The P wave velocity in the uppermost mantle is not well resolved but is probably less than 7.7 km/s. The S wave velocity derived from relatively clear Sn is 4.35–4.40 km/s. Our results show that the HTB is a prominent structural boundary extending to the Moho. The crust of Kitakami massif was not homogenized by the Cretaceous granitic intrusions, and the original structural difference remains in the upper crust.

The Beppu-Shimabara graben, in Kyushu, Japan, has active seismic regions and active volcanoes. Studies using triangulation, gravity surveys and mechanism solutions of large earthquakes show that the graben is in a state of extensive stress associated with rifting. A temporary seismic study was carried out from February 20 to March 8 in 1992 in the region, during which time many seismic signals were observed. The microearthquake seismograms recorded inside the graben have a later phase dominant on the horizontal component following the direct S-wave. Two-dimensional ray-tracing was applied to investigate the cause of the later phase. The phase was identified as an S-reflected S phase (SxS-phase) from a discontinuity with a depth of 20 km. The reflection points are located inside the Beppu-Shimabara graben that is in a state of extensive stress associated with rifting.

An extensive seismic refraction experiment with the use of explosive sources was conducted in the Kitakami region, northern Honshu, Japan, on November 1, 1990. The experiment area is divided into two geological units by the Hayachine Tectonic Belt (HTB). The southern terrane consists of pre-Silurian basements and Silurian-lower Cretaceous marine sediments, while the northern one is characterized by a Jurassic accretionary complex. Both of the units were intruded by Cretaceous granitic rocks. An almost N-S seismic refraction profile of 194-km length was extended from Kuji City, Iwate Prefecture to Ishinomaki City, Miyagi Prefecture, on which 4 shots with a charge size of 450-700 kg were fired

The key subjects of the 1986 Joint Seismological Research in Western Nagano Prefecture (JSR'86) were to determine hypocenters of aftershocks of the 1984 Western Nagano Prefecture Earthquake and the shallow crustal structure precisely. Therefore, we deployed a temporary seismic network of closely spaced stations with a high quality in the southern foot of Mt. Ontake, the epicentral area of the 1984 earthquake. As one of the key subjects of JSR'86, a three-dimensional velocity modeling was executed using the natural earthquake data observed during the experiment. In addition to this inversion of observed data of earthquakes, we conducted an explosion seismic experiment to obtain information on the shallow crustal structure more precisely to improve hypocenter determination as well as to find characteristics of the structure. The two-dimensional profile was derived from observed data of seismic waves by two shots at 43 temporal stations. The length of the profile was 20 km, crossing the Otaki River which may be concordant with the structural boundary as pointed out geologically and topographically. The derived time-term distribution represents the two-dimensional surface-layer thickness, and has correlations to the distribution of the Bouguer gravity anomaly and the hypocenter distribution in the area. The rim of high consistency of contour lines on the Bouguer gravity anomaly map and the time-term distribution map may coincide with the boundary of the seismically active area.

Focal mechanisms around Ontake Volcano, central Japan, were investigated by using the data set obtained by a seismic observation project of the 1986 Joint Seismological Research in Western Nagano Prefecture (JSR'86). The data set included P-wave first motions of aftershocks of the 1984 Western Nagano Prefecture Earthquake (M=6.8) mainly. A total of 556 events was selected for the focal mechanism analysis. The P-wave data from the surrounding university routine networks were also utilized additionally. Four hundred two events were successfully analyzed. Based on the types of focal mechanisms and on the spatial distribution of the events, we classify these events into the following 4 groups: (1) events of strike-slip fault type with a common nodal plane parallel to the ENE-WSW striking nodal plane in the fracture zone of the 1984 main shock; (2) events of reverse fault type in the area of the preexisting earthquake swarm to the northeast of the main shock, in a possible relation with an earthquake swarm activated by the main shock; (3) events of reverse fault type that occurred near the largest aftershock (M=6.2), but are not the continued aftershock sequence of the largest aftershock; (4) events of strike-slip fault type at depths to the easternmost part where no activity had been observed before the operation of JSR'86. The plural types of focal mechanisms obtained and the spatial distribution of foci suggest a complex seismo-tectonic structure that consists of at least 2 kinds of fracturings around Ontake Volcano.

A dense temporary seismograph network was set up in the aftershock area of the 1984 Western Nagano Prefecture Earthquake of M 6.8 by the group for the 1986 Joint Seismological Research in Western Nagano Prefecture. Seismograms of about 1,500 events were obtained during the observation period of 52 days from September 1 to October 22, 1986, and 530 events in and around the aftershock area were located precisely. Extent of the aftershock area is about 20 km in the east-west and about 7 km in the north-south.

The detailed three-dimensional P and S wave velocity structure in the focal region of the 1984 Western Nagano Prefecture Earthquake occurring close to an active volcano, Mt. Ontake, central Japan, is derived from a tomographic inversion of travel time data obtained by the 1986 Joint Seismological Research in Western Nagano Prefecture. The data set includes 7, 693 P-wave and 6, 070 S-wave arrival times observed at 49 stations from 212 local earthquakes and 2 explosions. The velocities in the shallow portion of upper crust are determined at each grid point with its spacing of 1-2 km and good resolutions are obtained from the Earth's surface to a depth of 4 km. There exist strong lateral heterogeneities especially from the surface to a depth of 1 km. The seismic velocity map obtained in the height of 1 km above sea level well corresponds to the maps of surface geology and Bouguer gravity anomaly. This correspondence, however, is not so clear in the deeper layers. And, the greater the depth, the less heterogeneous the velocity structure. The velocity distribution on the fault seems to correspond to the distribution of dislocation and rupture front estimated from an analysis of strong motion and geodetic data: the low velocity region has the large amount of dislocation and the retarded rupture front.

In order to compare crustal velocity structures of East Queen Maud Land with those of Enderby Land, large explosion seismic experiments are planned along 24°E, for a 400km profile from Breid Bay through Asuka Station toward inland of the Nansen Ice Field. We are now developing an easily deployable (expendable) seismograph from the helicopter (Antarctic penetrator) to install 40-50 seismic stations along the profile with the elevation range from 0 to 3000m in a limited operational schedule. The penetrator must have memories to store the digital seismic records. The recorded data have to be transmitted by radio-telemetry to the central data acquisition system on the helicopter. This report briefly summarizes scientific objectives of the planned explosion experiments, and the status of the development of the Antarctic penetrator.

Newly developed idea of the nascent plate boundary along the eastern Japan Sea has become to be discussed in the last five years. The boundary extends southwards through central Japan, where the Itoigawa-Shizuoka Tectonic Line (ISTL), one of the great tectonic lines in Japan, traverses to Suruga Bay. In the south of central Japan, the Philippine Sea plate is subducting under Japanese Islands, which lie on the North American plate and the Eurasia plate, and, therefore, a triple junction is formed there, if we acceptthe new idea of the plate boundary. Central Japan is, then, considered to be situated under the complex tectonic stress fields due to the plate interactions, and has the structures resulting from ongoing or past geodynamical process.Many seismic probings have been undertaken in central Japan and the derived structure will open informations on the geodynamic process and verifications of the nascent boundary. New resolution methods in seismic probing of the crust and the upper mantle can derive a three-dimensional velocity structure. The derived structure in central Japan, including ISTL, shows a low velocity body beneath the Hida mountains at the east of ISTL. However, the difference between the North American plate and the Eurasia plate has not beenresolved over the whole area along ISTL from the P-wave velocity of viewpoint. Many explosion profiles have been conducted in central Japan, some of which cross ISTL. The result at the north of ISTL shows a distinctive reverse fault dipping east near the Matsumoto Basin, though the fault location is slightly different eastwards from the geologically estimated tectonic line. The dip of the fault is concordant to that derived from the newidea of subducting the Eurasia plate under the North American plate. We cannot find other seismic profiles showing the similar structure in this area. In the southern part of ISTL, the northsouth fault is derived, which may be deformed by the plate interaction between thePhilippine Sea plate and the Eurasia plate.Unraveling a highly and widely resolved structure will be required to step forward to verify the new plate boundary hypothesis, and to put it practice, it is necessary to develop observation systems as well as the processing techniques. We address a request ofdeep seismic sounding which has controlled sources on a long-range profile with seismic waves penetrating deep into the crust and the upper mantle on a scale to discusss the new plate boundary hypothesis. Almost simultaneously, seismic tomography using natural sources, having a resolution of a three-dimensional image, will provide important information for better understanding of the geodynamical process in central Japan.

The 1986 fissure eruption of Izu-Oshima volcano was preceded and followed by strong volcanic tremor activities. The post-eruption tremor activity consists of an alternating sequence of a tremor episode lasting about 10 min and the subsequent quiescence of about 40-50 min. The authors observed this post-eruption activity by a 24-channel seismic array deployed on the caldera floor. The semblance analysis of the recorded data revealed that high frequency (>5 Hz) components of the tremor were radiated from two separated sources beneath the caldera. The tremor first occurred at one source and then migrated to the other source in a transition time of about a few minutes during each tremor episode. This transition suggests that the two sources are dynamically linked with each other in a magma-filled crack or conduit.

Two explosion seismic experiments were conducted in 1980 by the 21st Japanese Antarctic Research Expedition (JARE-21) to reveal the velocity structure beneath the Ongul Islands, East Antarctica. In one experiment eleven stations were established along a line of 5.2 km long at intervals of about 0.5 km on the outcrops on East and West Ongul Islands and two explosions of 100 and 80 kg were fired at the bottom of the sea. The other smaller-scale experiment was made on East Ongul Island to reveal the structure shallower than that obtained by the above-mentioned experiment. Five shots were fired and the seismic waves were observed at eight stations along two lines of 0.7 and 0.9 km long respectively. In both experiments direct waves were clearly recorded at all the stations as first arrivals. The mean apparent velocity and the mean average velocity of P-waves are 6.19±0.11 and 5.74±0.88 km/s, respectively. The obtained velocity in the ice-free area of the Ongul Islands is nearly the same as the velocities under the ice sheet in East Antarctica hitherto obtained by the research expeditions of the United States, USSR and Japan. The sedimentary layer with a velocity less than 5.5 km/s does not exist. Clear later arrivals with large amplitude were also detected at all the stations that ranged from 0.1 to 5.2 km from the shot point. The mean apparent velocity and the mean average velocity were 2.84±0.03 and 2.94±0.37 km/s, respectively. These later arrivals are presumably Rayleigh waves.

第20次日本南極地域観測隊は, 1979年1月に, 南極・昭和基地の近傍の宗谷海岸で人工地震観測を実施した。これは, 日本隊としては初めて地下構造を求めることを目的とした大規模な実験の一環であり, 次の諸テストを主な目的とした。1)南極における人工地震観測システムの確立, 2)ダイナマイト埋設用雪氷ボーリング機械, 3)低温下で使えるダイナマイト。オングル海峡の海面下110mの水中で1000kg, および内陸の62mの掘削孔下で560kgの火薬を爆発させ, 実験を行い, 所期の目的を達した。今回の測線は80kmで, 大陸内に10点の観測点を設置した。得られた地下構造で特徴的なことは, P波速度にして4～5km/sの堆積層と思われる層が500～1000mの厚さに, 氷床直下に存在することで, これは東南極地域で, これまでには見出されていない。, The upper crustal structure of the Soya Coast, Antarctica was estimated by a seismic refraction experiment which was carried out in January 1979. A surface thin layer with a P-wave velocity of 4-5km/s was derived from the traveltime analysis for the first time in East Antarctica by explosion seismology.

Three big seismic explosion experiments were carried out by the 21st and 22nd Japanese Antarctic Research Expeditions in the vicinity of Syowa Station and in the northern Mizuho Plateau in East Antarctica. Twenty-seven temporal seismic observation stations were set up along a 300-km long profile from Syowa to Mizuho Stations. Explosives were fired near Syowa Station, near Mizuho Station and at the middle point of the profile. The apparent P-wave velocity in the upper crust varies from 6.0 to 6.4 km/s, suggesting a gradual increase of the P-wave velocity with depth. Apparent velocities of 6.9 km/s for P^* and 7.9 km/s for P_n were observed from the biggest shot near Syowa Station. From the analyses of travel times and an amplitude study by synthetic seismograms, the crustal structure of the northern Mizuho Plateau was determined. The depth of the Conrad and the Moho discontinuities was determined as about 30 km and 40 km, respectively. The P-wave velocity increases from 6.0 km/s on the surface to 6.4 km/s at a depth of 13 km. At the Conrad, there is no sharp velocity discontinuity. The thickness of the transition zone is 2.4 km. Comparing our results with those in Dronning Maud Land (in the vicinity of Novolazarevskaya Station of USSR), East Antarctica, velocity values are nearly the same, but the thickness of lower crust in the northern Mizuho Plateau is about a half of that in Dronning Maud Land.

Studies on the crustal and upper-mantle structure in Antarctica have been one of the major contributions to Antarctic research since the International Geophysical Year of 1957–1958. Many refraction surveys with small charge size have been conducted in Antarctica, but long-range experiments were also made in 5 regions on the margin of the Antarctic continent. In 1979–1981, the scientific program of the Japanese Antarctic Research Expedition was focused on the earth sciences, and in particular, an explosion seismic experiment along a long survey line was the major item during these years. An experiment along a 300 km-long line with three shots and 27 observation stations was successfully made in the northern Mizuho Plateau, East Antarctica. From the analysis of travel times and the amplitude study of synthetic seismograms, the crustal structure of the northern Mizuho Plateau was determined. The depths of the Conrad and the Moho discontinuties were determined as 31 km and 42 km, respectively. The P-wave velocity and depth relation was determined as 6.0 km/s on the surface of the bedrock, 6.3 km/s at a depth of 2 km, 6.8 km/s at the Conrad and 7.9 km/s at the Moho. The velocity in the crust increases gradually. The crustal structure obtained is representative of East Antarctica.

The upper crustal structure of the Soya Coast, Antarctica was estimated by a seismic refraction experiment which was carried out in January 1979. A surface thin layer with a P-wave velocity of 4-5km/s was derived from the traveltime analysis for the first time in East Antarctica by explosion seismology.

Various kinds of geophysical surveys have been carried out in the Mizuho Plateau, East Antarctica by the Japanese Antarctic Research Expeditions (JARE). The correlation between the high-level gravity anomaly and the bedrock elevation is examined along a route where both data are sufficient to permit deriving a crustal model from gravity, radio-echo sounding and explosion seismic data. The bedrock elevation usually correlates well with the high-level gravity anomaly. However, along the traverse route S-H-Z from the Syowa to Mizuho stations, the bedrock elevation has a weak negative correlation with the high-level gravity anomaly. Such a weak negative correlation is attributed to the deeper part of the crust. The crustal structure between the Syowa and Mizuho stations is modeled from the gravimetric data and the radio-echo sounding of bedrock elevations, so as to fit the P-wave velocity structure derived from the data of explosion seismic experiments. Then the structure is extended from Syowa Station seaward across Lutzow-Holm Bay and from Mizuho Station southeastwards inland, where only gravimetric data are available. Thus, a crustal section about 600 km long is obtained on a margin of East Antarctica. The depth of the Moho increases by about 7 km from Syowa Station to the point Y200 (71° 46′S, 48° 56′E), about 500 km from the coast. A graben-like structure is obtained along the line across Lutzow-Holm Bay. This suggests that both sides of the bay are bounded by faults.

Explosion seismic experiments, gravity measurements and aeromagnetic surveys were made in the northern Mizuho Plateau including the Ongul Islands, East Antarctica, from 1979 to 1982 by the Japanese Antarctic Research Expeditions. The objective of these field operations was to determine the crustal structure along the 300 km-long oversnow traverse route between Syowa and Mizuho Stations. Three big shots were fired; at sea near Syowa Station, in an ice hole near Mizuho Station and in an ice hole between both stations. Twenty-seven temporal seismic stations were set up along the route. Gravity measurements were carried out at 30 points along this route. Aeromagnetic surveys over the area were made four times. In the seismic experiments, clear refracted waves from the Conrad (estimated depth 30 km) and the Moho (estimated depth 40 km) discontinuities were recorded. No layer with a velocity of less than 6 km/s was found in the Ongul Islands nor beneath the ice sheet in the surveyed area. The P-wave velocity in the upper layer varies with depth from 6.0 km/s on the surface to 6.4 km/s at a depth of 13 km. Comparing the observed record section with synthetic seismograms, it was derived that the Conrad was not associated with a sharp velocity discontinuity, but a linear velocity increase of 0.55 km/s in a transition zone of 2.4 km thick. Velocities of P∗ and Pn were determined as 6.95 km/s and 7.93 km/s assuming a flat layered structure. Bouguer gravity anomalies could not be calculated along the whole profile because of a lack of data on bedrock topography, so reduced gravity anomalies were calculated. These anomalies indicate no abrupt changes of the bedrock topography.

Seismic waves generated by two explosions of dynamite, 5 tons each, in the Sea of Japan off the coast of northern Honshu were observed at more than 100 temporary and permanent seismological stations in the Hokkaido, Honshu, Sado, and Oki islands. A purpose of these measurements was to extend our investigation of lateral variation in Pn velocity which has been found around northeastern Japan in the previous explosion experiments. In fact, a lateral variation in Pn velocity by about 5% was confirmed in regions of the uppermost mantle below the Sea of Japan and the Honshu island, although the boundary where the velocity change takes place was not determined.The measurements have also revealed an indication that the upper mantle just beneath the Moho interface under the area in the southeastern half of the Sea of Japan is anisotropic with respect to P-wave velocity. The velocity variation in the anisotropy is approximately 0.4km/sec (i.e., 5%) about a mean velocity of 7.94km/sec. The direction of the maximum velocity is 141°E of north which corresponds roughly to a direction perpendicular to the general trend of northern Honshu as well as to magnetic lineations.

In 1979-81, ten shots were fired in drill holes at the Prince Olav Coast and the Mizuho Plateau, Antarctica, for explosion seismic experiments by JARE-20, -21 and -22. The yield ranged from 10 to 1400kg. An empirical relationship between yield and maximum amplitude of seismic wave was derived for shots in drill holes in ice sheet. This will help to design an adequate charge size for explosion seismic experiments in Antarctica. The maximum amplitudes obtained are compared with those generated from explosion in sea water in Antarctica and in solid rock in Japan. For yields smaller than 500 kg, the logarithm plots of the amplitudes are in a linear relation with the logarithm of yields, and the amplitudes are about 1/5 to 1/8 of those generated in solid rock. For the yields larger than 500 kg, the wave amplitudes in the ice sheet become larger than those estimated from the above relationship. The amplitude of the biggest shot of 1.4t detonated in the ice sheet is nearly the same as that in solid rock in Japan. This is because small shots were fired in shallow drill holes where the ice is very porous, its density is only about 500 kg/m^3. The biggest shot was fired at a depth of 140 m in the ice with a density of 800 kg/m^3 or greater. The maximum amplitudes of seismic waves generated from a 1-t explosion in sea water are about 8 to 10 times larger than those from equivalent explosions in the ice sheet. The relation is similar to the results obtained for shots in sea water and solid rock in and around Japan.

A refraction experiment with a 100-kg charge was carried out in the ice sheet in the Prince Olav Coast, East Antarctica by the 21st Japanese Antarctic Research Expedition (JARE-21) in 1980. Prominent Rayleigh waves were clearly recorded in a distance range of 2 to 8 km from the explosion point. The maximum period of Rayleigh waves was 0.7 s with a phase velocity of 1.8 km/s. A phase velocity of surface waves is considered to be most sensitive to the P- and S-wave velocity structures of ice layers to a depth of one-fourth of the wave length. Since the longest wave length of Rayleigh waves in the present experiment is estimated to be 1300m for the maximum period of 0.7s (1.8 km/s in phase velocity), P- and S-wave velocity structures of ice layers to a depth of 300 to 400 m may be inferred from the dispersion curves. Assuming velocity structures of ice sheet in the Mizuho Plateau based on the previous works of JARE, theoretical dispersion curves were calculated for three different velocity structures and they were compared with the observed ones. For example, the calculated phase velocities at the period of 0.4s were 1.85 km/s for the highest velocity model, 1.7 km/s for the middle and 1.6 km/s for the lowest one, and the observed phase velocity at the period of 0.4s was 1.75 km/s. The velocity structure of P- and S-waves in the upper part of ice sheet estimated from observed dispersion curves of Rayleigh waves is three percents higher than those of the averaged one in the Mizuho Plateau. The P- and S-wave velocities from 100 to 1000 m in depth range from 3.7 to 4.0 km/s, and from 1.8 to 2.0 km/s.

The three-dimensional (3-D) P-wave velocity structure beneath Central Japan has been investigated in detail by an inversion method. 7490 P-wave arrival times from 120 shallow and intermediate depth earthquakes that have occurred in this region are used to estimate velocity anomalies in 3-D subdivided blocks and hypocentral perturbations, simultaneously. The results reveal complex 3-D structures, with low-velocity zones in the wedge portion of the upper mantle above the high-velocity Philippine Sea and Pacific plates subducting beneath this region. Prominent low-velocity bodies exist just beneath active volcanoes, particularly in the Hida mountain range. Low-velocity bodies are spatially correlated with the low-Q zones estimated from seismic intensity data. One low-velocity body coincides with an anisotropic body detected from the study of shear-wave splitting. Dome-shaped low-velocity masses seem to represent partially melted mantle diapirs. No clear evidence on velocity contrast has been identified across the Fossa Magna, which is a tectonic boundary between Northeast and Southwest Japan.

Six explosions ranging from 300kg to 600kg and a quarry blast of 7.5 tons in and around the Shizuoka district were observed within distances less than 150km from the shot points for refraction surveys of the crust in this province. As the geological structures in the Shizuoka district are highly complex, the shooting and the recording had to be made in various geological formations, which gave rise to difficult conditions that all seismic rays consisted of complicated paths. Accordingly, some of the results obtained by the experiments so far made near this region were also incorporated into this study. Combination of our experiments and those made by other researchers, in spite of short distances of observation lines, enables us to propose a four layered crustal model on the uppermost mantle. The main features of the model obtained here are as follows: (1) The east side of the Median Tectonic Line is depressed by about 5.7km and covered with thick surface layers. (2) The upper boundary of the granitic layer is expected to be upward convex beneath the western Izu Peninsula at a depth of about 2km. (3) The basaltic layer has a thickness of about 15 km. (4) The Moho is estimated as 33km deep beneath Suruga Bay.

The P-wave velocity structure of a part of the Prince Olav Coast, East Antarctica was revealed by a seismic refraction experiment conducted by the 21st Japanese Antarctic Research Expedition in July 1980. Ten seismic stations were set along a line with a length of about 10 km at intervals of about 1 km. The station consists of a seismometer and a direct analogue data-recorder with an electric panel-heater in an insulated box. The observed travel times of the first and later arrivals from two explosions at both ends of the line were analyzed. The P-wave velocity of the basement complex is 6.02 km/s which is a typical value in East Antarctica. This basement complex is covered by the ice sheet with the thickness of about 700-1000 m. No layer with a velocity less than 6 km/s is detected between the ice sheet and the basement complex.

wo big explosions of 7 and 5 tons were fired in the Pacific Ocean east of the Japan trench. This experiment was carried out for the purpose of confirming the results of the SEIHA-1 experiment reported in the previous paper in which we had suggested a lateral variation of the Pn velocity around northeastern Japan. The observation was made at about 50 stations on land and by an ocean-bottom seismograph. Records of good quality were obtained over a distance of 800km. The travel times suggest that the Pn velocity on the land side of the "aseismic front" is about 7.6km/sec while that on the sea side of it is about 8.1km/sec. It is also suggested that accuracy of hypocentral determination for earthquakes around the Japan trench may be remarkably improved by considering the lateral heterogeneity in the upper mantle revealed here.

Two types of noise recorded in seismic explosion experiments around Syowa Station in Antarctica are discussed. One is wind-induced noise which is detected directly by a seismometer. This type of noise increases with a wind velocity, and when the wind velocity is over 15 m/s, seismic observations are not in success. With the object of noise reduction, seismometers buried in a hole at depths of 3, 5 and 10 m were examined, but the noise existed at 10 m. Seismometers should be buried deeper than 10 m. Another is electrostatic noise which frequently deranges a clock signal or destroys an electric circuit in the worst case. This noise is caused by drifting electrified snow, not by a seismometer itself. It was excited on a signal cable stretched on the snow surface from the seismometer to a recorder. To reduce this noise, the cable should be as short as possible and be laid in parallel with the direction of wind.

A refraction experiment with a 100kg charge was carried out on the ice sheet of the Prince Olav Coast, East Antarctica by the 21st Japanese Antarctic Research Expedition (JARE) in 1980. Prominent Rayleigh waves were clearly recorded in a distance range of 2 to 8km from the explosion point. The maximum period of Rayleigh waves was 0.7s with a phase velocity of 1.8km/s. A phase velocity of surface waves is considered to be most sensitive to the P- and S-wave velocity structures of ice layers to a depth of one-fourth of the wave length. Since the longest wave length of Rayleigh waves in the present experiment is estimated to be 1300m for the maximum period of 0.7s (1.8km/s in phase velocity), P-aand S-waves velocity structures of ice layers to a depth of 300 to 400m may be inferred from the dispersion curves. Assuming velocity structures of ice sheet in the Mizuho Plateau based on the previous works of JARE, theoretical dispersion curves were calculated for three different velocity structures and they were compared with the observed ones. For example, the calculated phase velocities at the period of 0.4s were 1.85km/s for the highest velocity model, 1.7km/s for the middle and 1.6km/s for the lowest one and the observed phase velocity at the period of 0.4s was 1.75km/s. The velocity structure of P-aand S-waves in the upper part of ice sheet estimated from observed dispersion curves of Rayleigh waves is three percents higher than those of the averaged one in the Mizuho Plateau. The P-aand S-wave velocities from 100 to 1000m in depth range from 3.7 to 4.0km/s and from 1.8 to 2.0km/s.

Two explosion seismic experiments were conducted for the investigation on the velocity structure of the snow-free area of East Antarctica by the 21st Japanese Antarctic Research Expedition (JARE-21). In one of the experiments made along a line of 5.2km, two explosions of 100 and 80kg in charge amount were fired in the sea and eleven stations were established at about 0.5km intervals on the rocks outcropped in East and West Ongul Islands. The other smaller charge experiment was made in East Ongul Island to reveal shallower structure of snow-free area. Five shots were fired and generated seismic waves were observed at eight stations along two lines of 0.7 and 0.9km long. In both of the experiments direct waves were clearly recorded at all the stations as the first arrivals. The mean apparent velocity and the mean interval velocity (distance from shot is divided by travel time) of the waves are 6.19±0.11 and 5.74±0.88km/s, respectively. The obtained velocity of the snow-free area in the vicinity of the Ongul Islands is nearly the same as the velocities under the ice sheet of all areas in East Antarctica previously obtained by the research expeditions of the United States, USSR and Japan. A velocity increase with depth was not obtained in the area. The sedimentary layer with low velocity is very thin or does not exist, as the intercept time of the 6km/s-layer is nearly zero. Clear later arrival phases with large amplitude were also detected at all the stations of 0.1 to 5.2km from the shot point. The mean apparent velocity and the mean interval velocity are 2.84±0.03 and 2.94±0.37km/s, respectively. As the velocity is about a half of P-wave and smaller than that of S-wave derived from the usual V_p/V_s, for example 1.73,the phase seems to be Rayleigh wave.

Ten shots were fired in drill holes in Mizuho Plateau, East Antarctica for explosion seismic experiments in 1979-1981. The yield charge sizes are in a range of 10 to 1400kg. An emperical relationship between the yield charge and the maximum amplitude of generated seismic waves was derived for the shots in drill holes in the snow field, which will aid to design an adequate charge size for future explosion seismic experiments in Antarctica. The maximum amplitudes obtained above were compared with those generated in sea water in Lutzow-Holm Bay, Antarctica and those generated in the rock in the Izu Peninsula, Japan. When the explosives are less than 500kg, the amplitudes are about 1/10 of the amplitudes generated in the sea, and about 1/5 of amplitudes generated in the rock. When the charge is large, the difference in amplitude between the shot in the sea and that in the rock is small. The amplitude of the largest shot of 1400kg is nearly the same as that in the rock. This is because small shots were fired in shallow drill holes of 5 to 30m deep where the ice density is about 0.5g/(cm)^3 and ice is very porous, whereas the largest shot was fired at a depth of 140m where the density is greater than 0.7g/(cm)^3. The maximum amplitude of the seismic wave by a 1000-kg explosion at sea is about five times as large as that in the drill hole, which is similar to the relation for shots in the sea and in the rock obtained in Japan.