[SCG62-P04] Estimation of spatiotemporal distribution of interplate slip after the 2003 Tokachi-oki earthquake incorporating viscoelastic relaxation
Keywords:The 2003 Tokachi-oki earthquake, GNSS, Postseimsic deformation, Afterslip, Viscoelastic relaxation
The 2003 Mw 8.0 Tokachi-oki earthquake is an interplate earthquake along the Kurile trench. Its postseismic deformation has been observed by GNSS [e.g., Miyazaki et al. 2004]. Estimation of spatiotemporal afterslip is a key to clarify the healing process after large earthquake. Because the postseismic deformation should be caused by both viscoelastic relaxation and afterslip, it is important to incorporate both effects for the modeling. In this study, we estimated a spatiotemporal interplate slip for about 7.5 years following the 2003 event as well as the coseismic slip of the 2003 and M 6-7 class earthquakes simultaneously. We included a viscoelastic response of interplate slip in the estimation of the slip.
For the data analysis, we corrected the effect of the 1993 Hokkaido-Nansei-oki earthquake for the observed GNSS data in Hokkaido by using the model of Ueda et al. [2003]. The secular velocity before the 2003 event was estimated from the corrected data and removed from the postseismic data. And then, we removed a seasonal variation and displacements of the M6-7 events in the postseismic period. Finally, we down-sampled the residual time series with an interval of 1-6 months. We used about 7.5 years long GNSS data until the 2011 Tohoku-oki earthquake.
For the modeling of postseismic deformation, we constructed a model consisting of the coseismic slip of the 2003 and the following M6-7 class events, interplate slip including afterslip following these events and viscoelastic relaxation. We assumed the two-layers viscoelastic structure estimated by Itoh and Nishimura [2016] to estimate interplate slip distribution.
A preliminary result shows large postseismic slip occurred in the up-dip and down-dip extensions of the coseismic slip region and implies an interplate coupling had not been recovered to that before the 2003 event at the time of the 2011 event.
For the data analysis, we corrected the effect of the 1993 Hokkaido-Nansei-oki earthquake for the observed GNSS data in Hokkaido by using the model of Ueda et al. [2003]. The secular velocity before the 2003 event was estimated from the corrected data and removed from the postseismic data. And then, we removed a seasonal variation and displacements of the M6-7 events in the postseismic period. Finally, we down-sampled the residual time series with an interval of 1-6 months. We used about 7.5 years long GNSS data until the 2011 Tohoku-oki earthquake.
For the modeling of postseismic deformation, we constructed a model consisting of the coseismic slip of the 2003 and the following M6-7 class events, interplate slip including afterslip following these events and viscoelastic relaxation. We assumed the two-layers viscoelastic structure estimated by Itoh and Nishimura [2016] to estimate interplate slip distribution.
A preliminary result shows large postseismic slip occurred in the up-dip and down-dip extensions of the coseismic slip region and implies an interplate coupling had not been recovered to that before the 2003 event at the time of the 2011 event.