G11B-0482:
Revisiting the Interplate Coupling Beneath Northeast Japan Before the 2011 Tohoku Earthquake (M9.0) Based on Terrestrial and Seafloor Geodetic Observations
G11B-0482:
Revisiting the Interplate Coupling Beneath Northeast Japan Before the 2011 Tohoku Earthquake (M9.0) Based on Terrestrial and Seafloor Geodetic Observations
Monday, 15 December 2014
Abstract:
Large coseismic slip along the Japan Trench during the 2011 Tohoku Earthquake (M9.0) highlighted the necessity of the re-examination of the interplate coupling on the boundary between the subducting Pacific and overriding continental plates beneath the northeastern Japanese Islands. Interplate coupling along the shallowest portion of the megathrust before the Tohoku Earthquake must have been persistent, but spatial resolution of the inversion analysis based only on terrestrial geodetic data is generally not high enough to constrain the coupling state in the far offshore area.Meanwhile, seafloor geodetic observation has been developed and applied off the Pacific coast of Tohoku district in this decade, and the secular displacement rates before the 2011 Tohoku Earthquake are reported by Japan Coast Guard and Tohoku University. These seafloor geodetic observation data are invaluable to estimate the interplate coupling beneath the northeastern Japan, but have been difficult to be included in the inversion analyses, because the measurements contain large uncertainties and are more sparsely sampled both in temporal and spatial domains than the terrestrial observations. To overcome this difficulty, we have taken into account the correlations between the observed displacement rates at terrestrial GPS stations in the inversion analyisis. We assumed that the covariance between the displacement rates at two different GPS sites depends on the distance, and configured the covariance between the different components, such as EW, NS and UD, by applying the result of raw GPS data processing.
We performed numerical test to examine the advantage of involving the covariance matrix, and concluded that the covariance between the observations should be taken into account in the inversion analysis. We appied the inversion to the actual displacement field data obtained before the 2011 Tohoku Earthquake. The results show that evident temporal change of the slip deficit had occurred after an M7.2 earthquake near off coast of Miyagi Prefecture in 2005 and continued as a slow slip event in far off Fukushima and Ibaraki Prefectures regions in 2008. However, almost full coupling was estimated along the shallowest portion of the plate interface where extremely large (> 50 m) coseismic slip occurrred during the M9.0 Tohoku Earthquake.