Recognition of a locked state in plate subduction from microearthquake seismicity

• Published in: Pure and Applied Geophysics 1999-08, Vol.155 (2-4), p.669-687
• Main Authors: MATSUMURA, S, KATO, N
• Format: Article
• Language: English
• Subjects: Computer simulation; Earth sciences; Earth, ocean, space; Earthquake construction; Earthquakes; Earthquakes, seismology; Exact sciences and technology; Geophysics; Internal geophysics; Mathematical models; Plate tectonics; Recognition; Seismic activity; Seismic phenomena; Seismic waves; Seismicity; Seismology; Shear stress; Stress concentration; Stresses
• ISSN: 0033-4553; 1420-9136
• DOI: 10.1007/s000240050282

--A tectonic state of a locked subduction is considered to be a possible source of a future interplate earthquake. Discriminating an actually locked state to verify its extent is therefore essential in constructing an accurate prospect against the forthcoming earthquake. Micorearthquake seismicity is an effective tool for such an analysis because it is considered to be a faithful indicator of the stress state, and is expected to exhibit a characteristic pattern in the area where the locked state in the subduction appears with a certain stress concentration. Focusing on the microearthquake seismicity around the Tokai district in central Japan, where a large interplate earthquake is feared to occur, we tried to identify such an area of locked subduction on the Philippine Sea plate, possibly related to the future earthquake. We investigated the microearthquake seismicity from various perspectives. First, the hypocenter distribution was analyzed to identify the extent of the locked area. The characteristic profile of the distribution was presumed to represent a stress concentrated area induced from the mechanical contact between both plates. The second approach is to interpret stress patterns reflected in focal mechanisms. The locked state was recognized and verified by a comparison of the P-axis distribution pattern with that expected from a model imaging a partially locked subduction. The third approach is to monitor the temporal change of the seismic wave spectrum. Analyzing predominant frequencies of P and S waves and monitoring their changes for a period of 10 years, we found a trend of gradual increase common to both waves. This means an increase of stress drop in microfracturings, and in its turn implies accumulation of stress around the focus area. The rate of the stress change converted from the frequency change was compared with the result derived from a numerical simulation. The simulation, performed on the basis of a constitutive friction law for a stick sliding on the plate interface, computed a changing rate of the maximum shear stress around the locked zone and showed its spatial variation along the subduction axis. Thus the simulated result indicated a certain compatibility with the observed one. Although ambiguities and uncertainties still exist in the study, all the results derived here seem to indicate an identical conclusion that the plate subduction is actually locked in this region at present.[PUBLICATION ABSTRACT]