The 1586 Tensho earthquake occurred around the western part of central Japan. The magnitude is estimated to be around 8, as large as the 1891 Nohbi earthquake, mainly based on the intensity distribution map. However, historical documents concerning the precise damage and detailed crustal deformation etc. for this event are fairly limited as the occurrence time was during the Warring State Period. The Shokawa fault zone, Atera fault zone and Yoro-Kuwana-Yokkaichi fault zone, typical leading active faults having high activity and fresh topographical expression in central Japan, have been considered to be major candidates for the active faults which produced this great earthquake. The mysterious Tensho earthquake was selected as the main theme of a symposium held in Japanese Society for Active Fault Studies 2010 Fall Meeting, and was discussed in the oral session, based on the recent various data obtained by the investigations for these fault zones after the 1995 Hyogo-ken Nanbu (Kobe) earthquake (M7.3).
Fault activity in the Yoro–Kuwana–Yokkaichi fault zone is considered to have been the source of the 1586 Tensho Earthquake. The Geological Survey of Japan conducted array boring and trenching surveys at Hazawa and Niwada within the ~30-km-long Yoro fault system, which is the main component of this major fault zone. This paper reports the main results of these surveys, focusing on the most recent faulting events and with new paleoseismological data. At Hazawa, the boundary between prodelta and delta-front sediments formed within the Kiso River system at around 4 ka has been displaced vertically across the fault by about 15 m, and the top of the deltaic sediments formed at around 1.7 ka by ~10 m, indicating repetitive fault activity since 4 ka. The floodplain on the up-thrown side of the fault has been terraced since about the 8th century; this probably reflects the penultimate faulting event. The last event occurred after the 8th century. At Niwada, four offlapping sedimentary units separated by angular unconformities are capped by a buried soil or peat layer containing abundant plant fossils and charcoal fragments that have provided C-14 data to constrain the timing of faulting events. At least three episodes of fault activity have produced cumulative tilting during the period of stable sea level over the last 4 ka. The last two tilting events occurred after the 15th century and around the 8th century. The most recent faulting events on the Kuwana and Yokkaichi faults occurred after the 13th century. Furthermore, the two most recent rises of relative sea-level probably reflect coseismic subsidence on the Nobi plain about 500 and 1200 years ago. It is highly likely that the Yoro–Kuwana–Yokkaichi fault zone produced both the 1586 Tensho and 745 Tenpyo earthquakes, although geological data alone is insufficient to determine the exact age of these events.
Tensho earthquake is one of the most famous large inland earthquakes during the medieval times of Japan. It occurred on Jan. 18, 1586 in Chubu district. It was almost the equal size to Nobi earthquake in 1891, and we estimated its size is around M7.8-8.0. Many major active faults in Chubu district had been assigned as the source faults of this earthquake. However, the careful examination of the historical materials, and the precise analysis of the distribution of estimated reliable seismic intensities revealed that it is impossible to cause the damage of whole Tensho earthquake by a single earthquake of M8-class. The main shock occurred in the south-western part of Nobi basin. The source faults are narrowed down to some faults near Yoro and Suzuka mountains, and Ise Bay, or southern part of Yanagase, Sekigahara, and nearby faults, when we compare the intensity distribution with expected ones for several major fault groups in Chubu district. The famous destruction of Uchigashima Family in Kiun Castle and the crushing death of Mr. and Mrs. Ukon Maeda in Kifune Castle were caused by the different earthquake, which occurred on Jan 16, in some faults near Shokawa River, and its size was around M7.0 ±0.2. In order to reveal source faults of medieval destructive earthquakes, which usually have too few historical materials, we have to be careful to the credibility of information. Those written in later years may be modified through folklore transmission. Even in the primary historical sources such as a diary of an aristocrat or a letter by a missionary, hearsay information in remote area was left with true damages they really knew. Since there are much more destructive earthquakes than those known in the catalogue we have now, the trench results should not be restricted to select the candidate of earthquakes only from the current earthquake catalogue. For example, July 31, 1585, a very strong earthquake was felt in Mikawa, Ise, and Kyoto, and no destructive damage was known in Chukyo area. This must be the first candidate for the latest activity of Atera fault.
The 1586 M~8 Tensho earthquake might have been one of the largest inland earthquakes in Japan. Historical documents and geologic evidence suggest that there might be a composite rupture process involving more than a couple of active faults. To retrospectively evaluate such a possibility of distant cascading ruptures with stress transfer hypothesis, here I compute the static Coulomb stress changes under two representative scenarios. One is the Shogawa fault, Atera fault, and Yoro-Kuwana-Yokkaichi fault zone have been the sources(Sangawa’s scenario), while the other is a movement of the norther Shogawa fault followed by the Yoro fault or Isewan fault (Matsu’ura’s scenario). To assess the imminency and instability of their earthquakes in 1586, I also take the 30-yr earthquake probabilities of the causative faults into account. Computed stress transfer in five out of six combinations in the Sangawa’s scenario suggests that Coulomb stress on the subsequent rupturing fault is loaded by the previous earthquake by up to 0.5 MPa. On the Matsu’ura’s scenario, amount of stress transfer by the northern Shogawa fault to the Yoro or Isewan fault would be smaller than 0.01 MPa. But such small stress perturbation including possible effect of dynamic shaking might have been enough to trigger the Isewan fault which had been on the verge of the next rupture. Although there are still large uncertainties to assign fault parameters, my preliminary computations imply both scenarios can be supported by the recent consensus of static stress transfer hypothesis.
Based on existing geological information and historical records, we discuss the hypothesis that the Atera fault zone is the source fault of the 1586 Tensho earthquake. Geological data, such as excavation survey results, show that the latest and penultimate faulting events of the fault zone occurred about 400 and 1,000 years ago, respectively. It is thought that these events do not cause a whole rupture of the Atera fault zone. We also examine new historical records on the damage by the Tensho earthquake around the Atera fault zone. However, there is still no reliable evidence that the Atera fault zone is the source fault of the earthquake. More precise historical records are necessary to conclude that the latest faulting event of the Atera fault zone corresponds to the 1586 Tensho earthquake.
The Miboro Fault System or Shokawa Fault Zone extends for about 67 km along the upper course of the Shogawa River in the western Hida Mountains, central Japan. A trenching survey conducted by the Geological Survey of Japan in 1990 revealed that at least the Miogo Fault, a southern portion of the Miboro Fault System, might have ruptured in or after the 11th century. There is no direct evidence to indicate that the faulting event corresponds to the Tensho earthquake of 1586. However, villages damaged by landslide or slope failure associated with the Tensho earthquake are distributed along the Miboro Fault System. The distribution of damage, together with the above excavation survey result, suggests that the latest rupture event on the Miboro Fault System caused“ a Tensho earthquake” which severely affected Hida and northern Mino Districts.