The question of how future earthquake hazards should be estimated in general remains open. Thirty or forty years ago it was widely believed that “characteristic earthquakes” occurred more or less periodically, and that this could be used as the basis for making hazard estimates. However, statistical tests have refuted the characteristic earthquake
D609 (Kagan et al., 2012) and hazard maps made using the characteristic earthquake hypothesis fail to agree with subsequent seismicity (Stein et al., 2012), so it appears that new methods are required. As for the past tsunami history in Tohoku, a paleo-tsunami study by Minoura et al. (2001) found that in addition to the very large tsunami in 869 (during the “Jogan era,” in Japanese history), the Sendai Plain had also experienced two other similar tsunamis in the past 3000 years. The size of these three tsunamis was generally comparable to
leaves of the 2011 event (Mori et al., 2012). Paleo-tsunami studies further to the north have not been made, due mainly to the lack of preserved tsunami deposits, so the spatial extent of the Jogan event and the two earlier events found by Minoura et al. (2001) remains unclear. The need for new approaches to hazard estimation, the uncertainties regarding the past events, and the possibility of SMF-generated secondary tsunamis are all challenges that must be addressed in future studies.
