4:30 PM - 4:45 PM
[SSS05-11] Estimation of Average Displacement Rate Vector Using GNSS Data in Yaeyama Islands
Keywords:Average displacement rate, interplate locking, Ryukyu Trench, GNSS, back-slip, Yaeyama earthquake
The Southern Ryukyu region including Yaeyama Islands is located in the southwestern part of the Ryukyu Trench, where the Philippine Sea Plate is subducting beneath the Eurasian Plate. It is well known that this region has been recurrently damaged by tsunami, e.g. the 1771 Yaeyama earthquake. Nakamura (2009) estimated the source region of the 1771 earthquake at the shallow part of plate boundary. On the other hand, on the deeper part of the plate interface at depths of 30 – 60 km, SSEs with Mw ~6.6 recur with intervals of ~6 months (Heki and Kataoka, 2008; Nishimura, 2014; Tu and Heki, 2017). To understand crustal deformation in this region, GNSS observations have been conducted since 2010 by Kyoto University in addition to the permanent GEONET stations. Although these GNSS data have revealed the detail source processof the SSEs (e.g. Kano et al, 2018), few studies discussed interplate locking along the plate boundery. Therefore, we estimate average displacement rate in Yaeyama Islands based on GNSS data and discuss the interplate locking along the southern Ryukyu Trench.
We used daily coordinate data in the IGS14 coordinate system from 2010 to 2018 in a 18 GNSS stations consisting of 10 GEONET and 8 GNSS stations by Kyoto University. The average displacement rate was estimated by removing deformation due to SSEs, steps by earthquakes and antenna replacement from timeseries. Following Tu and Heki (2017), the time constant for each SSE was determined by grid search using the N20W components at 0751 (Hateruma Island), the nearest station to the trench. This direction is almost parallel to that of the plate convergence. We used the onset of each event determined by Tu and Heki (2017) or Kano et al. (2018) for the SSEs before the 2016 and visually defined for those after 2017. We considered the steps of earthquakes larger than M6.0. Then, the average displacement rate, steps of earthquakes and antenna replacements, amplitudes of SSEs were estimated using the least squares method.
As a result, Yaeyama Islands are entirely moving in a southeast direction at a rate of 5 to 8 cm/yr with respect to the IGS14 coordinate system. This steady motion is thought to be the sum of the rigid block motion, consisting of translation and rotation component, and the effect of the plate subduction. The translation component of rigid block motion was removed by subtracting the rate of reference site (0498, Miyako Island). The rotation component of the southern Ryukyu block relative to the Amurian plate was calculated from the Euler-pole and angular rate estimated in Nishimura et al. (2004).
We estimated coupling area based on the back-slip model. A fault model was approximated by a single rectangle assuming a semi-infinite homogeneous elastic body (Okada, 1992). For the fault locations, we assumed two areas: one in the shallow part near the 1771 Yaeyama earthquake and the other in the deep SSE area. We referred to Nakamura (2009) for the fault parameters such as fault length and fault width to the values for the shallow part and to Heki and Kataoka (2008) for those of the deeper part. The fault width, rake angle, and back-slip were adjusted by trial and error to reduce the difference between the observed and calculated rates. We successfully reproduced the observations at most of the stations although there were some negligible misfits at stations in Yonaguni and Ishigaki Island. The back-slip rate are 9 cm/yr at the shallow part and 12 cm/yr at the deeper part. By comparing the plate subduction rate of ~9 cm/yr, our results suggest the existence of the interplate locking on the shallow part of the southern Ryukyu subduction zone.
We used daily coordinate data in the IGS14 coordinate system from 2010 to 2018 in a 18 GNSS stations consisting of 10 GEONET and 8 GNSS stations by Kyoto University. The average displacement rate was estimated by removing deformation due to SSEs, steps by earthquakes and antenna replacement from timeseries. Following Tu and Heki (2017), the time constant for each SSE was determined by grid search using the N20W components at 0751 (Hateruma Island), the nearest station to the trench. This direction is almost parallel to that of the plate convergence. We used the onset of each event determined by Tu and Heki (2017) or Kano et al. (2018) for the SSEs before the 2016 and visually defined for those after 2017. We considered the steps of earthquakes larger than M6.0. Then, the average displacement rate, steps of earthquakes and antenna replacements, amplitudes of SSEs were estimated using the least squares method.
As a result, Yaeyama Islands are entirely moving in a southeast direction at a rate of 5 to 8 cm/yr with respect to the IGS14 coordinate system. This steady motion is thought to be the sum of the rigid block motion, consisting of translation and rotation component, and the effect of the plate subduction. The translation component of rigid block motion was removed by subtracting the rate of reference site (0498, Miyako Island). The rotation component of the southern Ryukyu block relative to the Amurian plate was calculated from the Euler-pole and angular rate estimated in Nishimura et al. (2004).
We estimated coupling area based on the back-slip model. A fault model was approximated by a single rectangle assuming a semi-infinite homogeneous elastic body (Okada, 1992). For the fault locations, we assumed two areas: one in the shallow part near the 1771 Yaeyama earthquake and the other in the deep SSE area. We referred to Nakamura (2009) for the fault parameters such as fault length and fault width to the values for the shallow part and to Heki and Kataoka (2008) for those of the deeper part. The fault width, rake angle, and back-slip were adjusted by trial and error to reduce the difference between the observed and calculated rates. We successfully reproduced the observations at most of the stations although there were some negligible misfits at stations in Yonaguni and Ishigaki Island. The back-slip rate are 9 cm/yr at the shallow part and 12 cm/yr at the deeper part. By comparing the plate subduction rate of ~9 cm/yr, our results suggest the existence of the interplate locking on the shallow part of the southern Ryukyu subduction zone.