Additional file 1 of Characteristics of seismicity in the southern Okinawa Trough and their relation to back-arc rifting processes

Additional file 1: Figure S1. Frequency–magnitude distributions of earthquakes in the southern Okinawa Trough. a The whole study area. b The Yaeyama Rift. c The Yonaguni Rift. d The Ishigaki Knoll and the Hatoma Knoll. Each location is indicated in the b-value map at the bottom. Red curves indicate cumulative numbers of events with a magnitude equal to or larger than each magnitude. Gray bars are event numbers for each magnitude bin. The complete magnitude was estimated to be 2.0 using the goodness-of-fit test of 95% (Wiemer and Wyss, 2000; Woessner and Wiemer 2005). Figure S2. Distribution of b value with different radii r of a 10 km and b 15 km. These two different parameters produce an almost same pattern of b value anomalies as that with r = 12 km, while samplings with r = 10 km significantly reduce the coverage. Contour lines are drawn every 0.2. Figure S3. Distribution of b value with different complete magnitude Mc of a 2.3 and b 2.5. These two different parameters also produce a similar pattern of b value anomalies as that with Mc = 2.0, while larger Mc significantly reduce the coverage. Contour lines are drawn every 0.2. Figure S4. Distribution of b value with different time period of a January 2001 to January 2010 and b February 2010 to August 2019. Although the number of seismic stations for JMA earthquake catalog increased from six to eight in February 2010 in the study area, any significant temporal change in b value distribution is not recognized. Contour lines are drawn every 0.2. Figure S5. Distribution of b value calculated by error-added locations of earthquakes. To examine the effects of location uncertainties on the b value distribution, I added a random location error of ± 5 km and ± 12 km to the longitude and the latitude of the earthquake locations from the JMA catalog and recalculated the b value distribution. The result successfully reproduces the general pattern of b value anomalies from the error-free locations (Fig. 3). Figure S6. a Number of events used to calculate b values at each grid point and b standard deviations of b values. Thin lines in a are contour lines of 50. In b, contours are drawn every 0.10.