Research ArticleGEOPHYSICS

Mechanism of the 2015 volcanic tsunami earthquake near Torishima, Japan

See allHide authors and affiliations

Science Advances  04 Apr 2018:
Vol. 4, no. 4, eaao0219
DOI: 10.1126/sciadv.aao0219
  • Fig. 1 Observation outline.

    (A) Bathymetry map showing the locations and mechanisms of two events, the 2 May 2015 volcanic tsunami earthquake (Mw = 5.7, depth = 10 km) and the 10 May 2015 ordinary thrust earthquake (Mw = 5.9, depth = 6 km). Ten small triangles indicate stations of the ocean bottom pressure gauge array. (B) Comparison of band-pass–filtered (1 to 80 mHz) records between the 2 May event and the 10 May event. The vertical axis is not scaled so that the amplitudes can be visually compared. Amplitudes of seismic waves are similar between the two events, yet tsunamis are visible only on the 2 May records. The arrows indicate the transition from a very weak downswing to the major upswing of tsunami motion.

  • Fig. 2 Estimate of tsunami source location.

    (A) Five tsunami source locations tested in this study (stars). One is within the caldera and four are along the rim. The CMT diagrams obtained by GCMT, USGS, and JMA are shown at their epicenters. (B) Spatial distribution of the misfit of the calculated phase travel time and propagating direction to the observed values. The minimum misfit is located within the caldera. The small circle with a cross indicates the estimated location of source periphery with its uncertainty. (C) Test result showing the sensitivity of tsunami travel time to the shift of source location from north to south through C (A). Black line with error bars shows the observed travel times. (D) Test result showing the sensitivity of tsunami propagating direction to the shift of source location from east to west through C (A). Black line with error bars shows the observed propagating directions.

  • Fig. 3 Average trace of 10 tsunami records mutually time-shifted to a common zero-cross time.

    The dotted lines indicate the SD of these time-shifted records. This figure emphasizes the weak downswing before the major upswing of the tsunami motion.

  • Fig. 4 Best initial sea-surface displacement model.

    (A) Observed (black) and simulated (red) waveforms at OBP stations (see Fig. 1A for locations). The blue horizontal line represents the time window to calculate the normalized root mean square (NRMS) misfit at each station. (B) The sea-surface displacement for a model of A = 1.5 m and R = 4.1 km (see Materials and Methods). Contour lines represent the bathymetry near the Smith Caldera. (C) The cross-sectional profile of the initial sea-surface displacement. (D) Observed (black) and synthetic (red) waveforms at the Yaene Port on Hachijo Island (see fig. S5).

  • Fig. 5 Vertical ground deformations due to two phenomenological sources.

    (A) Vertical T–CLVD volume source. Top: source geometry. The dimensions are 8 km × 8 km (horizontal) and 2 km (vertical) with a top depth at δ. Arrows show the relative displacement direction. Bottom: Calculated vertical surface deformation profile for different top depths δ. The source area is shaded. The deviatoric moment is fixed at 4 × 1017 Nm so that the relative vertical displacement is 14 cm. (B) Vertical opening of a horizontal tensile fault. Top: source geometry. The horizontal fault surface is 8 km × 8 km at a depth of 0.2 km. Arrows show the dislocation direction. Bottom: calculated vertical seafloor deformation profile (red curve). The sharp edges of this profile should be truncated when the seafloor disturbance is transmitted to the sea surface (20, 32). The dislocation is fixed at 1 m. The deformation profile in an infinite medium is shown for comparison (blue curve). Note that the vertical scale is different between (A) and (B).

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/4/4/eaao0219/DC1

    fig. S1. Repeatedly occurring (approximately every 10 years) tsunami earthquakes near the Smith Caldera.

    fig. S2. Phase slowness analysis.

    fig. S3. Frequency-dependent ray trajectories and wave fronts.

    fig. S4. Search for the best initial sea-surface displacement.

    fig. S5. Waveform simulation at the Yaene Port on Hachijo Island.

    table S1. Array information.

    table S2. Result of array analysis of phase and onset arrivals.

  • Supplementary Materials

    This PDF file includes:

    • fig. S1. Repeatedly occurring (approximately every 10 years) tsunami earthquakes near the Smith Caldera.
    • fig. S2. Phase slowness analysis.
    • fig. S3. Frequency-dependent ray trajectories and wave fronts.
    • fig. S4. Search for the best initial sea-surface displacement.
    • fig. S5. Waveform simulation at the Yaene Port on Hachijo Island.
    • table S1. Array information.
    • table S2. Result of array analysis of phase and onset arrivals.

    Download PDF

    Files in this Data Supplement:

Navigate This Article