1:15 PM - 1:30 PM
[S03-08] A decade of GNSS/Acoustic measurements on the back-arc spreading in the southwestern end of the Okinawa Trough
The sustained seafloor geodetic measurement is crucial, even yet sparse in time and space, to constrain the undersea crustal deformation near the plate boundaries versus time. We present an integrated analysis of a decade of GNSS/Acoustic data collected at the site 60 km to the east of the northeast coast of Taiwan near the opening axis of the Okinawa trough back-arc basin. As a result, we obtained the horizontal and vertical positioning time series of the benchmark by the eighteen measurements from 2009 to 2019. The positioning time series of the benchmark shows a southeastward movement at a rate of 44±3 mm/yr, with respect to the Yangtze plate, following the potential offset in 2012. The horizontal motion can be explained by the clockwise block rotation of the Yonaguni block united with the northern part of the Central Range. In addition, the vertical displacement of the transponder array shows the rapid subsidence of 20±7 mm/yr from 2012 to 2019. The fast subsidence rate and the negative free-air gravity anomaly in this region indicate that the crustal thinning is compensated mainly by the surface deformation instead of the upward migration of the Moho. Considering the offset in 2012 due to the change of transponder array, the horizontal positioning time series of our site is best explained by two linear lines with a slope change in August 2013. The timing of the velocity change roughly coincides with the change of the periphery seismicity rate and the dike intrusion 150 km away from the site. The results show the potential of our seafloor geodesy site in accessing the temporal variation in the deformation rate near the spreading center of the Okinawa Trough, which is not accessible by on-land GNSS stations.
============Figure caption=================
Time series of the OILN benchmark position relative to the stable part of Yangtze plate. (a) Same with Figures 3a and 3b. North-south (solid circles) and east-west (open circles) components. The coordinates for the revived data before 2012 are shifted assuming an offset at 2012. The original north-south and east-west coordinates are shown by black and gray crosses. Each time series is fitted by linear lines with a break at August 7, 2013 (vertical line) which is determined by AIC which is shown by the imposed diagram. The data for the middle eight campaigns are common with Chen et al. 2018 but solved by our new procedure. (b) Cumulative seismicity larger than M2.0 detected within the corresponding colored rectangular shown in Figures 6c and 6d. The red dashed line fits the seismicity during the period from January 2009 to July 2013. (c) Horizontal velocity vectors and seismicity larger than M3.0 during the period from Jan 1, 2009 to August 7, 2013. D1 and D2 show the location of dike intrusion modeled by Nakamura & Kinjo (2018) which occurred in the end of 2002 and April 2013, respectively. (d) Horizontal velocity vectors and seismicity during the period from August 8, 2013 to September 30, 2019.
============Figure caption=================
Time series of the OILN benchmark position relative to the stable part of Yangtze plate. (a) Same with Figures 3a and 3b. North-south (solid circles) and east-west (open circles) components. The coordinates for the revived data before 2012 are shifted assuming an offset at 2012. The original north-south and east-west coordinates are shown by black and gray crosses. Each time series is fitted by linear lines with a break at August 7, 2013 (vertical line) which is determined by AIC which is shown by the imposed diagram. The data for the middle eight campaigns are common with Chen et al. 2018 but solved by our new procedure. (b) Cumulative seismicity larger than M2.0 detected within the corresponding colored rectangular shown in Figures 6c and 6d. The red dashed line fits the seismicity during the period from January 2009 to July 2013. (c) Horizontal velocity vectors and seismicity larger than M3.0 during the period from Jan 1, 2009 to August 7, 2013. D1 and D2 show the location of dike intrusion modeled by Nakamura & Kinjo (2018) which occurred in the end of 2002 and April 2013, respectively. (d) Horizontal velocity vectors and seismicity during the period from August 8, 2013 to September 30, 2019.