The Neodani Fault extending approximately 100 km NW-SE in central Honshu is one of the well-known active faults in Japan. While under road construction at a site of Neo-Osso, Motosu City, Gifu Prefecture, part of an outcrop of the fault was discovered, where a clear vertical displacement was also identified. On the NE side of the outcrop, a horizontal boundary between a basement and terrace deposits is clearly visible at a height of 6.5 m. On the SW side of the crop, however, thick terrace deposits being only visible cause the corresponding boundary to be hardly confirmed. This study tried to explore a boundary hidden by the terrace deposits, for which a spatial autocorrelation (SPAC) method, a kind of microtremor survey methods, was applied as an exploration technique suitable for this case. Following the procedure of the SPAC method, observations of microtremors were made at four sites selected around the outcrop. For each site phase velocities of the fundamental mode Rayleigh wave were derived as a function of frequency and inverted to a subsurface structure, i.e., an S-wave velocity structure. To see a general view of the studied area, all the S-wave velocity structures derived were put together in a 2-dimensional cross-section along the outcrop and also referred to geological situations observed at ground surfaces. Consequently the depth of the boundary hidden by terrace deposits in question could be estimated at about 4.5 m at a site just near the SW side of the fault. Examining the hidden boundary estimated on the SW side together with the corresponding boundary observed on the NE side, the vertical displacement of the fault was approximately estimated at 11 m in total, which could be taken as an accumulated displacement along the fault after the formation of the late Pleistocene terrace. The vertical displacement thus estimated can be also converted to an average vertical slipping rate, which is one digit smaller than an average lateral slipping rate of approximately 2 m in every 1000 years.
Shinjima Island is a small island of ~0.2 square kilometer located northeast of the Sakurajima Volcano, Kagoshima Bay, southwest Japan. According to historical records, Shinjima Island was uplifted and emerged from bottom of the sea at the eruption of the Sakurajima Volcano in 1779. Shinijima Island is formed by horst-and-graben structures composed of numerous E-W striking normal faults due to the vocalnic activity. On the both sides of the main graben, we found two major graben-bounding faults which vertically offset a key shell bed by 15.9 m, and 23.5 m respectively. In contrast to such narrow and distinctive surface deformation, a N-S trending seismic profile displays a structure that swells up the geologic units broadly at the depths of 200m and 320 m. The profile also suggests no distinctive fault offsets but warped strata are interpreted below the depth of 50 m. To estimate volume and movement of the magma storage at the 1779 eruption and other historic and prehistoric events, we performed forward modeling to restore the surface deformation in an elastic half-space. A complex model combining a point source inflation and dike intrusions approximately reproduces the current Shinjima landform. The model suggests that the ridge-trough-ridge structure and associated active faults on Shinjima Island have been developed by such volcano-tectonic processes.
The Median Tectonic Line (MTL) active fault zone is one of the most active and longest active fault zones onshore in Japan, and is a potential source of destructive large earthquakes.The history is a key data to evaluate the long-term seismic risks of a fault zone.However, the timing of the most recent event on the MTL in west Shikoku is not well defined yet. In this paper, we studied coseismic deformation on the walls and floor of an archaeological trench across the fault scarp of the Shigenobu fault, in order to date the recent surface faulting events on the Shigenobu fault.The Late Holocene sediments including cultivated soil contain evidence for two faulting events in the past 5315 years. The recurrence interval is estimated to be less than 2200 to 2600 years. This interval is much shorter than the previous estimates of 3500 to 4000 years. The latest event occurred during the Medieval Period of the Japanese history. The penultimate event was estimated to have occurred between 5315 cal yrs BP and AD 6th century.
The Kikukawa fault zone, more than 44 km long, in the western part of Yamaguchi Prefecture, is composed of the Kikukawa fault and the Kanda-misaki-oki fault. The Kikukawa fault zone is active and inferred that it has a potential of future earthquake, but its magnitude cannot be deduced because its northwestern end is not known yet. In this study, the seafloor swath mapping was conducted to determine the northwestern extension of the Kanda-misaki-oki fault. A pull-apart basin and a fault valley caused by the activity of the Kanda-misaki-oki fault were found on the survey map and the Kanda-misaki-oki fault extends further northwest. Because these structures show no trace of marine and/or fluvial erosion, these structures have been formed after the submergence around 15,000 years ago.
The Chelungpu Fault, Taiwan, records the surface rupture by the 1999 Chichi earthquake and abundant information has been obtained on paleoseismology on this fault during the last 10 years. This study area, located on the central part of the Chelungpu Fault, is known as one of the type localities of the surface rupture, but no paleoseismological study was carried out. This paper reports the trench log across the Chelungpu Fault preserved at the Earthquake Museum of Taiwan and tries to reconstruct the faulting history at this site based on 10 bore hole data on the location of the museum. Trench log indicates the displacement of terrace deposits and underlying bedrock by the 1999 earthquake. Four subparallel faults are observed within 1.8 m width, corresponding to the location of the surface rupture. Estimated height difference of bedrock surface based on the bedrock height by bore hole data is more than 4.2 m, significantly larger than the offset by the 1999 earthquake (2-2.5 m). This fact implies that at least two earthquake events including the 921 earthquake occurred at this site. We also estimate the fault plane dips eastward at about 30 degree, based on the location and stratigraphy of one bore hole which penetrates to the terrace gravel on the footwall through the bedrock.