Rupture and tsunami generation process of the 2015 Mw 5.9 Bonin earthquake revealed by in-situ pressure gauge array observations and an integrated simulation of seismic and tsunami waves

Tuesday, 15 December 2020
Tatsuya Kubota1, Tatsuhiko Saito1, Yoshio Fukao2, Hiroko Sugioka3, Aki Ito2, Takashi Tonegawa2, Hajime Shiobara4 and Mikiya Yamashita5, (1)National Research Institute for Earth Science and Disaster Resilience, Tsukuba, Japan, (2)JAMSTEC Japan Agency for Marine-Earth Science and Technology, Kanagawa, Japan, (3)Kobe University, Graduate School of Science, Hyogo, Japan, (4)The University of Tokyo, Tokyo, Japan, (5)National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan
Absolute pressure gauges (APGs) installed at seafloor have widely used to observe various phenomena such as tsunamis and ocean acoustic waves (e.g., Fukao et al. 2018; Saito & Kubota 2020). On September 1 2015, an Mw 5.9 interplate earthquake occurred near the Bonin Trench (Fukao et al. 2020, submitted to S001 Session). An array of eight pressuremeters with the APG (and one broadband ocean bottom seismometer) installed at the region overlapped with the focal area (in-situ observation) made it possible to observe both seismic waves, tsunamis, and permanent crustal deformation during earthquake rupture and tsunami generation. We analyzed the APG to reveal the rupture and tsunami generation process of the earthquake.

We removed the ocean tides and applied the lowpass filter (cutoff of 0.033 Hz). Large impulsive signals (~0.02 Hz) were observed just after the focal time and then the low frequency waves (< ~0.005 Hz) and pressure offset changes were recorded. We simulated the pressure changes including seismic waves, tsunamis, and crustal deformation (Saito 2019), to find that the impulsive signals were generated by the dynamic pressure changes due to seismic motion, and the subsequent low frequency signals and offset changes were by tsunamis and permanent crustal deformation, respectively.

We then constructed the finite fault model. We found the pressure offsets and tsunamis (characterized by slow propagation velocity) contributed to constrain the fault size, horizontal location and seismic moment. The center of the main rupture was located at ~15 km east and ~10 km northwest from the USGS and GCMT centroids, respectively. The estimated fault model had the seismic moment of 9.5 × 10 18 Nm (Mw 5.9), consistent with the GCMT and USGS solutions. We also obtained the stress drop of 0.51 MPa. The dynamic pressure changes contributed to constrain the rupture duration Tr. Although the lowpass filter (0.033 Hz) made it difficult to resolve the duration Tr in detail, Tr > ~30 s could not explain the observed dynamic pressure changes at all. We also investigated the ocean acoustic wave signals (> ~0.33 Hz) to find Tr ~5–6 s explained the observed signals well. These results suggest that this earthquake was an ordinary earthquake rather than a tsunami earthquake as pointed by Fukao et al. (2020).