XVI INQUA Congress

Paper No. 14
Presentation Time: 1:30 PM-4:30 PM

HOLOCENE SUBDUCTED PLATE MOTION ALONG THE SAGAMI TROUGH, CENTRAL JAPAN, REVEALED FROM PALEOSEISMOLOGICAL STUDY


SHISHIKURA, Msasnobu, Active Fault Research Center, Geol Survey of Japan/AIST, SiteC7, 1-1-1Higashi, Tsukuba, 305-8567, Japan, m.shishikura@aist.go.jp

Sagami Trough extending NW-SE off South Kanto is convergent plate boundary where the Philippine Sea Plate subducts beneath the North American Plate. Two major historical earthquakes of the 1703 Genroku Kanto Earthquake (M 8.2) and the 1923 Taisho Kanto Earthquake (M 7.9) occurred along the trough. As a result of the coseismic crustal movements associated with these earthquakes and repeated pre-historic earthquakes, Holocene emerged shoreline indicators (e.g. emerged wave-cut bench, fossilized mollusk etc.) are distributed on about 10 or more levels along the coast of the Miura Peninsula and the Boso Peninsula. The height distributions of lower two levels of them in the Miura Peninsula indicate more than 1 m of uplift during both the 1703 and 1923 events. In the Boso Peninsula, although geodetic data show up to 2 m of uplift during the 1923 event, historical documents and the height distribution of emerged shoreline indicators suggest that the coseismic crustal movement during the 1703 event had resulted steep northward tilting accompanied with uplift of more than 5 m in the southernmost area and subsidence of ca. 1 m in the central area. According to above results, it is inferred that the fault source model of the 1703 event is composed of dual fault system (fault A and B). The fault A is same as the model of the 1923 event which has already been estimated by Ando (1974). The uplift of the Miura Peninsula (both the 1703 and 1923 events) and the Boso Peninsula (only the 1923 event) can be explained by about 6.7 m slip of the fault A. The fault B is low angle dip thrust located off the southeast of the Boso Peninsula. The unique 1703 coseismic crustal movement in the Boso Peninsula has been derived from about 12 m slip of the fault B. 14C ages of the older paleo-shorelines above two recent levels indicate that the characteristic earthquake generated from the fault A has been occurred repeatedly about every 400 years. Basing on the feature of emerged shoreline topography, one of several events is the 1703 type earthquake accompanied with the slip of fault B. Recurrence interval of this type is estimated to be 2000-2700 years. Therefore, the slip rate of these faults are estimated to be 16.8 mm/year (fault A) and 5.2 mm/year (fault B) respectively. These value are greatly different each other and smaller than the recent back-slip rate (30 mm/year) estimated from GPS data.