Abstract
Abstract We developed a method to map and investigate temporal changes in the b-value using the signed Akaike information criterion (AIC) value and seismicity rate before the 2003 September 26–27 Northern Miyagi earthquake sequence. We studied the seismicity within a radius of 30 km around the epicenter of the largest shock, which was M 6.4, by three approaches: an analysis of all the events listed in the catalog; an analysis of only background events, but on data extracted by two different methods. We found that the change in the b-value of the clustered activities and background activities are different. The b-value of all events decreased from 1.2 to 0.7 and that of background event decreased even more. We empirically modeled the precursory temporal variation of the b-value and searched similar patterns among 971 areas chosen sequentially from inland Japan. The precursory pattern was both rare and significant; the AIC value was the second most significant factor among the more than 3000 tests conducted on 971 areas, each with four time windows. If the catalog was not declustered, the AIC of the precursory pattern became lower.
Highlights
We developed a method to map and investigate temporal changes in the b-value using the signed Akaike information criterion (AIC) value and seismicity rate before the 2003 September 26–27 Northern Miyagi earthquake sequence
We studied the seismicity within a radius of 30 km around the epicenter of the largest shock, which was M 6.4, by three approaches: an analysis of all the events listed in the catalog; an analysis of only background events, but on data extracted by two different methods
Introduction some doubt their significance of precursory seismic patterns before large earthquake, there have been a number of reports on such patterns, including precursory swarm or preshocks (Sekiya, 1976), accelerating seismic moment release (Varnes, 1989; Sykes and Jaume, 1990), precursory seismic quiescence (Inouye, 1965), decreases in b-value (Imoto, 1991), donut pattern (Mogi, 1969), horizontal migration (Kagan and Knopoff, 1976), downward migration (Mogi, 1988), and bursts of aftershocks (KelisBorok et al, 1980)
Summary
Some doubt their significance of precursory seismic patterns before large earthquake, there have been a number of reports on such patterns, including precursory swarm or preshocks (Sekiya, 1976), accelerating seismic moment release (Varnes, 1989; Sykes and Jaume, 1990), precursory seismic quiescence (Inouye, 1965), decreases in b-value (Imoto, 1991), donut pattern (Mogi, 1969), horizontal migration (Kagan and Knopoff, 1976), downward migration (Mogi, 1988), and bursts of aftershocks (KelisBorok et al, 1980). Rundle et al (2002) obtained some positive results on the possibility of earthquake prediction based on seismicity pattern. Such patterns should be investigated given the impact of earthquake prediction to society. If the b-value is constant for most of area and period where no earthquake followed, we may be able to predict any forthcoming earthquake by searching similar patterns prospectively. To test the uniqueness of the observed precursory pattern and study the feasibility of the earthquake prediction based on b-value change, we applied the obtained empirical model to various areas taken from throughout inland Japan. Our aim was to determine how often we would find a similar pattern in areas where no large earthquake had occurred in the preceding 1 year
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
