[J06-3] The spectrum of fault-zone deformation processes (from slow slip to earthquake) III

[J06-3-04] Continuous S-wave signals following 2014 Mw 6.8 SSE in the Hikurangi subduction margin offshore New Zealand

^○Yuriko Iwasaki¹, Kimihiro Mochizuki¹, Motoko Ishise¹, Erin Todd², Susan Schwartz², Stuart Henrys³, Martha Savage⁴, Anne Sheehan⁵, Yoshihiro Ito⁶, Laura Wallace^3,7, Spahr Webb⁸, Tomoaki Yamada¹, Masanao Shinohara¹ (1.The University of Tokyo, Tokyo, Japan, 2.The University of California Santa Cruz, Santa Cruz, California, 3.GNS Science, Lower Hutt, New Zealand, 4.Victoria University of Wellington, Wellington, New Zealand, 5.University of Colorado–Boulder, Boulder, USA, 6.Kyoto University, Kyoto, Japan, 7.University of Texas Institute for Geophysics, Austin, USA, 8.Columbia University, Palisades, USA)

From May, 2014, for about a year, marine seismic and geodetic experiment was conducted at the Hikurangi subduction margin. During this experiment, a Mw 6.8 SSE occurred from September through October, 2014, directly beneath the ocean bottom seismometer (OBS) network (Wallace et al., 2016). In this study, we used continuous waveform data recorded by these OBSs, and applied an S-wave splitting analysis (Ando et al., 1983) and a polarization analysis for monitoring S-wave signals. These methods have been successfully applied to waveform data from onshore seismic networks (Bostock and Christensen 2012; Ishise and Nishida 2015).

As a result, we detected the continuous arrival of S-wave signals that appeared to have started in the later half of the SSE. This continuous signal was identified as tremor and its source location was determined by the envelope cross-correlation method (Todd et al., 2017, in prep). Our result, however, suggests that these signals occur continuously rather than as sporadic individual events, and that they last for more than two weeks. Polarization directions change at the same time and remained stable through the two week duration. Our analysis requires less OBSs than other methods for monitoring such S-wave signal, which may enable us to detect as yet unidentified signals in the Hikurangi margin where seismic attenuation has been known to be large. Distribution of the OBSs detecting such continuous signals suggests that they were generated only around the subducted seamount adjacent to the slow slip area. This is consistent with the location of individual tremors identified with envelope cross-correlation methods (Todd et al., 2017, in prep). The slow slip along the plate interface circumvented the subducted seamount (Wallace et al., 2016). By considering our result with the slip distribution, we can put more constraints on relationship between frictional properties along the plate interface and its topographic features.