9:15 AM - 9:30 AM
[SSS14-02] Epicenters and Magnitudes of Ancient Earthquakes
Keywords:historical earthquakes, seismic intensities, epicenters, magnitudes
The epicenters and magnitudes (M) of historical earthquakes in Comprehensive List of Damaging Earthquakes (List) are widely used in publications such as Chronological Scientific Tables. However, it is unclear how these values were determined. According to Usami (1966) and Utsu (2001), the values determined by Kawasumi (1951, K1951) were adopted in List 1975, but even he did not provide any specific methods for determining them, other than using seismic intensities. However, when values are changed from those of K1951, the reason is stated in the text of List and this may serve as an explanation of the method.
The 679 earthquake is the oldest with epicenter values, which were 130.4 and 32.7 in K1951. They were revised to 130.5 and 33.5 in List 1975 (square 1 in Upper), and the reason for this was given in the text as “Nihon Shoki mentions Tsukushi." The corresponding passages in Nihon Shoki are "a great earthquake shook Tsukushi Province" and "many peasant houses and villages were destroyed." This implies that the epicenter was likely located near the center of Tsukushi Province. Thus, the method was to consider the distribution of ground motion and damage, or the distribution of seismic intensity, which is their indicator, and determine the center of the distribution to be an epicenter. In List 1987, Bungo Fudoki was added as a new material, and the epicenter was changed to 130½~131 and 33~33½ (region 2). The text states "according to Bungo Fudoki, Mt. Itsuma collapsed and hot springs appeared in various places.” It can be understood that the method taken for this change was to assume that the epicenter was located within the area surrounded by this point and the point above. In List 2003, the epicenter was changed to 130.68 and 33.32 (square 3). The text states It was caused by activity in the Minou fault system. The epicenter was set at its midpoint. Following this example, the epicenters and Ms of historical earthquakes were determined from an intensity distribution, and results of active fault research if available.
We reconsidered the epicenter and M of this earthquake by incorporating both methods of intensity distribution and active fault research. According to Usami (1994), the two points were rated as having the same intensity of V or higher and are plotted with circled ≧V in Upper. In the 2004 long-term assessment of the Minou fault system (orange lines), it was determined that the most recent activity of this was the 679 earthquake. Taking both into consideration, the epicenter can be located near the Minou fault system and approximately equidistant from the two points. Considering the accuracy of the data we searched for such a point in ¼ or 1/3 units and obtained 130¾ and 33 1/3 (star). To confirm them, we calculated the intensities at the two points using the attenuation equation by Matsuzaki et al. (2006), the M and depth of the fault plane center from the assessment. The results were 5.14 and 5.02, which are consistent with ≧V.
The 684 earthquake is the next to have epicenter values. They were 134.0 and 32.5 in K1951 (square 1 in Lower), and revised to 133.5~135.0 and 32¼~33¼ in List 1987 (region 2). Of the 6 points in Usami (1994), we excluded Kumano (different from current Kumano City) and Ota (written with a different month in a class C document). Imabari, which was added to the 2010 edition, is also in a class C document, but has been included because of the correct month. After comparing both editions, 5+ of Imabari was deemed equivalent to ≧V in the 1994 edition. Tosa with intensity E was written in Nihon Shoki as “rice fields were submerged in the sea,” so it was rated (VI) as this corresponds to “upheaval or subsidence” in Supplement of Intensity Determination Table. Hongu with intensity E was written in Kumano Nendaiki as “the three Kumano shrines were severely damaged,” so it was rated 5+ corresponding to “temples and shrines: significant damage” in 1980 Intensity Scale Explanation Table. Since this is a sensory intensity, 5+ has been rewritten as V+. Two ruins were chosen from the list of earthquake archaeology results in Lists 1996 and later, and were rated V+ (liquefaction) and ≧V (landslides and collapses). The intensity distribution in Lower is similar to that of the 1946 Nankai earthquake, so we adopted its fault model by Murotani et al. (2015), and set the epicenter and depth at the model center (yellow square and star). The fault distance from each point to the model was calculated using P4CF. The intensity at each point was calculated using the distances, attenuation equation, M of Lists 1987 and later, and model center depth. We obtained 5.6 for Tosa and 5.0 to 5.5 for the others. They do not contradict Lower.
We plan to conduct the above analyses for earthquakes before 1185.
The 679 earthquake is the oldest with epicenter values, which were 130.4 and 32.7 in K1951. They were revised to 130.5 and 33.5 in List 1975 (square 1 in Upper), and the reason for this was given in the text as “Nihon Shoki mentions Tsukushi." The corresponding passages in Nihon Shoki are "a great earthquake shook Tsukushi Province" and "many peasant houses and villages were destroyed." This implies that the epicenter was likely located near the center of Tsukushi Province. Thus, the method was to consider the distribution of ground motion and damage, or the distribution of seismic intensity, which is their indicator, and determine the center of the distribution to be an epicenter. In List 1987, Bungo Fudoki was added as a new material, and the epicenter was changed to 130½~131 and 33~33½ (region 2). The text states "according to Bungo Fudoki, Mt. Itsuma collapsed and hot springs appeared in various places.” It can be understood that the method taken for this change was to assume that the epicenter was located within the area surrounded by this point and the point above. In List 2003, the epicenter was changed to 130.68 and 33.32 (square 3). The text states It was caused by activity in the Minou fault system. The epicenter was set at its midpoint. Following this example, the epicenters and Ms of historical earthquakes were determined from an intensity distribution, and results of active fault research if available.
We reconsidered the epicenter and M of this earthquake by incorporating both methods of intensity distribution and active fault research. According to Usami (1994), the two points were rated as having the same intensity of V or higher and are plotted with circled ≧V in Upper. In the 2004 long-term assessment of the Minou fault system (orange lines), it was determined that the most recent activity of this was the 679 earthquake. Taking both into consideration, the epicenter can be located near the Minou fault system and approximately equidistant from the two points. Considering the accuracy of the data we searched for such a point in ¼ or 1/3 units and obtained 130¾ and 33 1/3 (star). To confirm them, we calculated the intensities at the two points using the attenuation equation by Matsuzaki et al. (2006), the M and depth of the fault plane center from the assessment. The results were 5.14 and 5.02, which are consistent with ≧V.
The 684 earthquake is the next to have epicenter values. They were 134.0 and 32.5 in K1951 (square 1 in Lower), and revised to 133.5~135.0 and 32¼~33¼ in List 1987 (region 2). Of the 6 points in Usami (1994), we excluded Kumano (different from current Kumano City) and Ota (written with a different month in a class C document). Imabari, which was added to the 2010 edition, is also in a class C document, but has been included because of the correct month. After comparing both editions, 5+ of Imabari was deemed equivalent to ≧V in the 1994 edition. Tosa with intensity E was written in Nihon Shoki as “rice fields were submerged in the sea,” so it was rated (VI) as this corresponds to “upheaval or subsidence” in Supplement of Intensity Determination Table. Hongu with intensity E was written in Kumano Nendaiki as “the three Kumano shrines were severely damaged,” so it was rated 5+ corresponding to “temples and shrines: significant damage” in 1980 Intensity Scale Explanation Table. Since this is a sensory intensity, 5+ has been rewritten as V+. Two ruins were chosen from the list of earthquake archaeology results in Lists 1996 and later, and were rated V+ (liquefaction) and ≧V (landslides and collapses). The intensity distribution in Lower is similar to that of the 1946 Nankai earthquake, so we adopted its fault model by Murotani et al. (2015), and set the epicenter and depth at the model center (yellow square and star). The fault distance from each point to the model was calculated using P4CF. The intensity at each point was calculated using the distances, attenuation equation, M of Lists 1987 and later, and model center depth. We obtained 5.6 for Tosa and 5.0 to 5.5 for the others. They do not contradict Lower.
We plan to conduct the above analyses for earthquakes before 1185.