2:15 PM - 2:30 PM
[SSS12-03] Seismic Quiescence Preceding the 1923 Kanto Earthquake (M7.9)
Keywords:Seismic quiescence, Kanto earthquake, Utsu catalog
The Kanto earthquake of September 1 in 1923 was the giant earthquake of magnitude 7.9, that occurred on the fault beneath the western Kanagawa and its vicinity. Regarding the precursory seismic activity of this earthquake, a doughnut pattern formed by relatively widespread activity for about 30 years and low-level activity near the epicenter (Mogi, 1980), and a decrease in the level of widespread seismic activity from about five years before the main shock (Ohnaka, 1984/85; Ogata, 1992) have been reported. In these studies, the Utsu catalog was used, whose epicenter data had been extensively compiled and recalculated by Utsu (Utsu, 1979, 1981, 1982). On the other hand, Hamada and Tsumura (2023) showed that a decrease in seismic activity may have occurred at least five stations about 3 years before the main shock, based on the records of the number of felt earthquakes in and around the Kanto area, without relying on any epicenter catalogs. Their result indicates that seismic quiescence is a universal phenomenon that can be observed regardless of detection methods, and that it may have spatial heterogeneity. In this report, therefore, we examine the details of appearance of quiescent regions preceding the Kanto earthquake based on the Utsu catalog.
The detection of quiescence and activated areas is based on the eMAP method (Yoshikawa et al., 2021). The process first counts events centered on individual epicenters that occurred during the entire period, each within a certain distance R, by dividing them into a reference period and an evaluation period. Next, the number of events during the evaluation period is converted into a Poisson probability (cumulative value) based on the average rate of occurrence during the reference period and displayed on the map. The probability for each epicenter is defined as active or quiet, if the probability is greater or less than 50%, respectively. Utsu guaranteed to have a certain accuracy for shallow, felt earthquakes of magnitude of M5.5 or greater in the Kanto and Chubu regions since 1904 (Utsu, 1981). Therefore, the analysis was carried within depths of 100 km or less, 33 to 38 north latitudes in degree, and 138 to 143 east longitudes in degree. The catalog ranks epicenters from A to D (within 1 degree) and unmarked (1 degree or more) according to the accuracy (error) of the epicenter location. The epicenters with ranks A to D were used in this study. Furthermore, for the above hypocenters, the lower limit of earthquakes that satisfy the Gutenberg-Richter law is set at M5.8.
Figure 1 shows an example of the results of the analysis (M=5.8 and larger, R=80 km). This result is in general harmony with the results of Hamada and Tsumura (2023), who indicated the possibility of quiescence, as sites such as Mt. Tsukuba, Mito, Tokyo, and Yokohama are included in the quiescence area, while Kumagaya, Kofu, and Numazu, where no quiescence was reported, are not. Temporal changes in seismic activity by region are shown in Figure 2. We can see the seismic activity has increased and decreased since around 1918, respectively.
The detection of quiescence and activated areas is based on the eMAP method (Yoshikawa et al., 2021). The process first counts events centered on individual epicenters that occurred during the entire period, each within a certain distance R, by dividing them into a reference period and an evaluation period. Next, the number of events during the evaluation period is converted into a Poisson probability (cumulative value) based on the average rate of occurrence during the reference period and displayed on the map. The probability for each epicenter is defined as active or quiet, if the probability is greater or less than 50%, respectively. Utsu guaranteed to have a certain accuracy for shallow, felt earthquakes of magnitude of M5.5 or greater in the Kanto and Chubu regions since 1904 (Utsu, 1981). Therefore, the analysis was carried within depths of 100 km or less, 33 to 38 north latitudes in degree, and 138 to 143 east longitudes in degree. The catalog ranks epicenters from A to D (within 1 degree) and unmarked (1 degree or more) according to the accuracy (error) of the epicenter location. The epicenters with ranks A to D were used in this study. Furthermore, for the above hypocenters, the lower limit of earthquakes that satisfy the Gutenberg-Richter law is set at M5.8.
Figure 1 shows an example of the results of the analysis (M=5.8 and larger, R=80 km). This result is in general harmony with the results of Hamada and Tsumura (2023), who indicated the possibility of quiescence, as sites such as Mt. Tsukuba, Mito, Tokyo, and Yokohama are included in the quiescence area, while Kumagaya, Kofu, and Numazu, where no quiescence was reported, are not. Temporal changes in seismic activity by region are shown in Figure 2. We can see the seismic activity has increased and decreased since around 1918, respectively.