Mamoru Kato

Upper mantle velocity structure in the western Pacific rim estimated from short-period recordings at Matsushiro Seismic Array System

We investigate lateral variation of the upper mantle P-wave velocity structure in the western Pacific rim using travel times of P phases from regional shallow events. Onsets of direct and later arrivals are manually picked up from short-period waveforms at Matsushiro Seismic Array System in central Japan by consulting the correlation among the waveforms at seven array stations. For the path along Kurile and Alaska seismic zones, observed first arrivals of shallow events (d = 5–50 km) are well explained by a standard travel time table for Japan. Observations provide no strong evidence of a distinct low-velocity zone in the uppermost mantle, and triplicated branches are adequately explained by a slight modification of model 12 by Sugiyama and Nakanishi (1989, 1999). Observed travel times of direct P phases from events along the eastern and western Philippine Sea are slightly larger at around 15° and slightly smaller at around 25°, and the upper mantle is slower and the transition zone and uppermost lower mantle faster in the eastern and western Philippine Sea than in the northwestern Pacific. Such differences could reflect a high temperature in the upper mantle, and accumulation of subducted material at the base of transition zone as a result of the past subduction.

Static Coulomb failure function and aftershocks of 1995 Kobe earthquake: A statistical test

[1]xa0The hypothesis that aftershocks are triggered by a change of the static Coulomb Failure stress is statistically tested for 1995 Kobe (Hyogo-ken nanbu) Earthquake. We focus on the focal mechanisms of the aftershocks for the first 7 months and investigate whether these mechanisms are consistent with the static CFF triggering hypothesis. Two slip distribution models for the mainshock are tested and both explain focal mechanisms of more than 3/4 of the aftershocks. From a bootstrap statistical test, however, we find that such a high consistency rate between the observation and CFF based prediction is not statistically significant. The static CFF triggering is thus not a statistically strong hypothesis for 1995 Kobe aftershocks.

Variation of teleseismic short-period waveforms at Matsushiro Seismic Array System

We investigate the origin of variation of teleseismic P phase waveforms recorded at Matsushiro Seismic Array System. Short-period, vertical component waveforms from three source regions were analyzed separately in the frequency domain. Observed variation is mainly attributed to the site effects, which are frequency dependent and strongly depend on the incoming azimuth. Such variation is not seen for the pre-event noise spectra. The secondary wavefield that is excited by the incident P phase is the likely cause of the observed waveform variation.

Anomalous propagation of Rayleigh waves along Izu‐Mariana Trench

Broad band records of Macquarie Ridge earthquake of May 1989 recorded in Japan show anomalous propagation of Rayleigh wave with frequency of 5 – 25 mHz along the back arc side of the Izu-Mariana trench. Rayleigh waves at three stations in central Japan show different polarization, that is, while one station shows a normal Rayleigh wave, other two stations, only 290 and 510 km apart, show anomalous polarization with a large deviation from the radial direction and their temporal and gradual change. Possible causes of this anomaly include anisotropy as well as laterally heterogeneous velocity structure of the upper mantle in this active marginal basin region.

Revisiting the Ishimoto–Iida Law for Strong‐Motion Seismograms: A Case Study at CEORKA Network, Japan

Abstract Frequency distribution of the maximum amplitude of seismograms at a station exhibits a power‐law relation, which is the Ishimoto–Iida law (II law). We investigate whether the II law is applicable to modern strong‐motion seismograms. Distribution of maximum amplitudes records at stations of The Committee of Earthquake Observation and Research in the Kansai Area (CEORKA) network, western Japan, exhibits properties of the power law or the II law. The observation matches well with the prediction, which is based on existing ground‐motion prediction equations for Japan. Parameters of the observed II law could be used in understanding the probabilities of future seismic hazard intuitively.

Frequency Distribution of Felt Earthquakes

Abstract Seismic intensity is a measure of how seismic ground motion is felt for a particular earthquake. Using the intensity data of the Japan Meteorological Agency (JMA) for the last 32xa0years, we report that the number of felt earthquakes exponentially decays as the function of the observed maximum intensity. The slope of this exponential relation is approximately 0.6 with some variability, and this is roughly in accord with the Gutenberg–Richter law of seismicity, and existing empirical relations among intensity, magnitude of earthquakes, and hypocentral distance. The number of felt earthquakes does not increase in proportion to a large increase of observation sites during this period.

Numerical models of convection in a rheologically stratified oceanic upper mantle: Early results

Recent seismological evidences imply that the boundary between the lithosphere and asthenosphere is a compositional boundary in the oceanic upper mantle, and a rapid increase of viscosity at this boundary is suggested. We modeled a thermal convection in the oceanic mantle numerically using the finite element method, and investigated geodynamical consequences of such a rheological layering. Early results from both quasi-steady state flows and time-dependent flows are presented in this report. We assumed a temperature- and depth-dependent viscosity law so that both the thermal effects and those of layering are taken into account. The effect of a high-viscosity layer (HVL) is small on the flow and the temperature field. Velocity gradients in the HVL are small in both directions, and the velocity field is well approximated by a one-dimensional channel flow. The HVL acts as a low-pass filter of the dynamic topography.