Spatial heterogeneity of the structure and stress field in Hyuga-nada region, southwest Japan, deduced from onshore and offshore seismic observations

In Hyuga-nada region, the Philippine Sea (PHS) plate is subducting beneath the Eurasian (EU) plate (the southwest Japan arc) along the Nankai trough at a rate of about 5 cm per year. The seismic activity in the boundary between the PHS and the Eurasian (EU) plates varies spatially along the Nankai trough. Especially the region from off coast of Shikoku to the Bungo channel and Hyuga-nada has large variation of seismicity. Although usual microearthquake activity is active in Hyuga-nada, it is inactive near Shikoku. On the other hand, although the great earthquake (M>8) has occurred repeatedly in near Shikoku at intervals of about 100 years, in Hyuga-nada, smaller earthquakes (M7 class) has occurred at intervals of about dozens of years, and so plate coupling varies dozens of kilometers specially. Big earthquakes (M7 class) have occurred in the north region from latitude 31.6 degrees north, but it has not occurred in the south region from latitude 31.6 degrees north. The largest earthquake ever recorded in Hyuga-nada region is the 1968 Hyuga-nada earthquake (Mw 7.5). And microseismicity varies spatially. It is important to understand seismic activity, stress field, and structure in such region in order to understand seismic cycle. We performed extraordinary seismic observation in and around Hyuga-nada region. More than 20 pop-up type OBSs were deployed above hypocentral region of Hyuga-nada using Nagasaki-maru and several data loggers were deployed in order to compensate a regular seismic network on land. We detected earthquakes more than 2 times of JMA. Seismic activity in source region of the 1961 Hyuga-nada Earthquake (M7.0) is low, but around its source region, seismic activity is very high. In order to obtain a 3D seismic velocity structure and precise hypocenter distribution and focal mechanisms around the Hyuga-nada region, we used Double-Difference (DD) Tomography method developed by Zhang and Thurber (2003). We could detect the structure of subduction of Kyushu-Palau Ridge at low seismicity area. We estimated the stress filed using a stress tensor inversion method by polarity of first arrivals from earthquakes [Horiuchi et al. (1995)], and we found that there is a good correlation between the slip distribution at large earthquakes and the angle between maximum principal axis and the plate boundary in northern part of Hyuga-nada region [Uehira et al. (2007)]. Because the shear stress of plate boundary is large on the subducted Kyushu-Palau Ridge, we suspected that it might be caused the strong interplate coupling. We also found a subducted seamount in the southwest margin of source region of the 1968 Hyuga-nada earthquake (Mw 7.5). This may acts as a barrier.