［Abstract］
　The primary research approaches of Lithospheric Structure Research Group are integrated geophysical-geological studies for investigating plate dynamics, including active/passive seismic studies, swath-bathymetry mapping, magnetic/gravity profiling, geodetic observation. From recent results of the Nankai seismogenic zone, we have obtained structural features in controlling rupture processes ofmega thrust earthquakes.

［Subducted seamount, highly reflective layer at the down-dip end of the seismogenic zone off Muroto］
　Figure 1. Trench-arc lithospheric transect in southwestern Japan obtained by an onshore-offshore integrated seismic survey. Seismic activity (after Nakamura et al, 1997) along the profile is plotted on the model. The arrows of the top of the model represent areas of the low resistivity layer obtained by MT study (Yamaguchi et al., 1999) and the locally no co-seismic slip zone obtained by geodetic data (Sagiya and Thacher, 1999). Ages of accretionary prisms are simplified after Taira et al. (1996). MTL: median tectonic line. This transect indicates a subducted seamount and anomalously low P-wave velocities beneath it. Both the locally no co-seismic slip zone during the 1944 Tonankai and the 1946 Nankai earthquake and the low resistivity area show good correspondence to the location of our observed highly reflective layer. A trapped water between the plate is considered to be a possible interpretation of the highly reflective/low resistivity layer.


［Nankai subduction structure］
　Figure 2. Crustal structure models placed from the east. The area denoted by "A" indicates the young accretionary sediments. Coseismic rupture zones of the 1944 Tonankai and 1946 Nankai earthquake [Ando, 1975; Ando, 1982] are shown at the top of the figure (a),(b). Interseismic locked and transition zone estimated by geothermal data, and critical boundary temperature (150, 350, 450 degree) [Hydman et al., 1995] are also indicated.
(a)Off the Kii Peninsula. (b)Off the Cape Muroto. This section shows the largest young accretionary sediments among three profiles. (c)Off the cape Ashizuri. The model indicates the smallest young accretionary sediments among three profiles.

［Seismogenic zone drilling, possible drill sites］
　Figure 3. Red broken lines indicate depth of subducting oceanic crust below sea floor, which are obtained from a series of seismic surveys. Water depth of 2500 m is shown by red line. Possible riser drill sites, water depth < 2500 m and basement depth < 5000 m, are highlighted by pink areas. In the western Nankai: the initial break starting off the cape Shiono extended over the entire locked zone with short rise time, then the rupture jumped around the subducted seamount. A small spot of a possible drill area is recognized to the east of the subducted seamount. This area is located on the edge of the seismically determined rupture zone (Kanamori 1972) and in a sub-fault showing 2-3 m slip by tsunami data (Tanioka et al., 1999).
In the central Nankai: a co-seismic rupture zone and an aftershock area are located at landward of the possible drill area, however tsunami data show that 1-2 m slip was extended to the possible drill site. In the eastern Nankai: the possible drill site is farther seaward of the slip zone during the 1944 Tonankai earthquake. by an onshore-offshore integrated seismic survey. Seismic activity (after Nakamura et al, 1997) along the profile is plotted on the model. The arrows of the top of the model represent areas of the low resistivity layer obtained by MT study (Yamaguchi et al., 1999) and the locally no co-seismic slip zone obtained by geodetic data (Sagiya and Thacher, 1999). Ages of accretionary prisms are simplified after Taira et al. (1996). MTL: median tectonic line. This transect indicates a subducted seamount and anomalously low P-wave velocities beneath it. Both the locally no co-seismic slip zone during the 1944 Tonankai and the 1946 Nankai earthquake and the low resistivity area show good correspondence to the location of our observed highly reflective layer. A trapped water between the plate is considered to be a possible interpretation of the highly reflective/low resistivity layer.