1996 年 49 巻 3 号 p. 361-374
We propose a new idea for the crustal movements associated with past great Nankai earthquakes in Shikoku and the surrounding region, Southwest Japan. This study makes clear that the movements are not due to the oblique subduction of the Philippine sea plate, but due to superposition of strong seismic shaking to the uppermost crust in a compressional stress state in the E-W direction. The ground of this idea is as follows: at the time of the 1707, 1854 and 1946 Nankai earthquakes the Kochi plain subsided and the Muroto and Ashizuri peninsulas uplifted, whereas old documents show that the Kochi plain did not subside at the time of the 1605 Nankai earthquake and therefore no uplifting of the peninsulas is inferred because it is based on a set of subsidence of the Kochi plain and uplifting of the Muroto peninsula appearing at the time of the 1707, 1854 and 1946 Nankai earthquakes. This is explained by the reason why the 1605 (Keicho) earthquake was not accompanied with strong shaking of the ground owing to the tsunami earthquake. Next, because the uplifted peninsulas have anticline axes of the N-S direction, from unconsistency in stress direction it is difficult to attribute the uplifting to the subduction in the NW direction of the Philippine sea plate. On the other hand, it is easily explained that the uplifting was caused by the stress in the EW direction enhanced by strong seismic shaking. Tosa bay, which spreads between the Muroto and Ashizuri peninsulas, is characterized by depression. To be able to explain this depression is not by the elastic rebound theory, but by our idea. The undulation in the forearc zone composed of anticlines (peninsulas) and wide depressions (bays), which range alternately along the Japan island arc, can not also explained by the elastic rebound theory. In addition, we consider that the compressional strain variation in the NW direction of the ground surface observed at present in Shikoku does not result from the oblique subduction in the NW direction of the Philippine sea plate, but it is recovering the overdisplacement of the ground surface caused by the coseismic movement (2-3m at the ground surface) in the SE direction of the Muroto promontory by reverse faulting of the 1946 earthquake. Moreover, this study shows that unconsistency in directions of P axes between the 1946 Nankai earthquake and mantle earthquakes presently occurring in Shikoku and its vicinity is succesfully explained by taking account of constraint of the displacement in the direction along the Japan island arc.