Preliminary investigation of seismicity rate change immediately before mainshock and fault maturity for damaging intraplate earthquakes in Japan

Abstract

Since the magnitude 7.3 Hyogo-ken-nambu earthquake of 1995, several other damaging earthquakes have struck Japan, including the 2000 Tottori-ken-seibu and 2004 Niigata-ken-chuetsu earthquakes. The latter earthquakes did not have dominant surface ruptures comparable to their subsurface fault lengths, suggesting that seismic hazard assessments should include not only large earthquakes produced by mapped active faults but also moderately sized earthquakes with few surface clues. This short report aims to (1) calculate pre-seismic indexes for the following recent intraplate earthquakes in Japan: 1995 Hyogo-ken-nambu,1997 Kagoshim-ken-hokuseibu,1997 Yamaguchi-ken-hokubu,1998 Iwate-Shizukuishi,2000 Tottori-ken-seibu,2003 Miyagi-ken-hokubu,2004 Niigata-ken-chuetsu, and 2005 Fukuoka-ken-seiho-oki earthquakes; and (2) discuss these indexes based on a fault evolution model (Wesnousky,1999). Earthquake data for 1984-2005 were obtained from the Annual Earthquake Report of the Japan Meteorological Business Support Center and the High Sensitivity Seismograph Network Japan (Hi-net), which is operated by the National Research Institute for Earth Science and Disaster Prevention. Earthquakes of magnitude> 2.0 and depth< 20 km were selected using both 0.2 × 0.2 and 1.0× 1.0 degree rectangular grids with the hypocenter located at the center. The calculated statistical indexes include the cumulative number of earthquakes, magnitude-frequency relationship, Gutenberg-Richter parameters (b and a values), AS-function, and the LTA (long-term average) -function as a function of time. The preliminary results show (a) positive relationships for magnitude, the step width of abrupt seismicity rate changes, and the time until the main shock, although only four data sets were available for regression; (b) a large change in b-value amplitude as a function of time for earthquakes of magnitude 7.0 or larger, with the Hyogo-ken-nambu earthquake having the maximum amplitude and a dominant surface rupture; and (c) indication of seismic quiescences by the AS-function and LTA-function. Although these results were qualitatively interpretable from the fault evolution model, no qualitative relationship was shown from which to predict the time until the main shock and its magnitude. To improve the results, further objective quantitative measurements of continuous/discontinuous fault traces are necessary for incorporation into the fault evolution model.