Large Earthquakes in the Tokyo area

An interim report on a Swiss Re-sponsored study, 2006
[Printable article (1.9 Mb)]

M. Bertogg 1, M. Guatteri1, S. Tschudi1
(1)Swiss Re


The first official probabilistic earthquake hazard assessment for Japan was released in March 2005 in a milestone study initiated by the Japanese government. A group of scientists from the AIST¹, NIED², GSI³, JMA4 (Japan) and the USGS5 (USA), sponsored by Swiss Re, embraced the challenge and took it one step further. The Japanese-US team zoomed into the highly exposed Greater Tokyo area and exhaustively analysed the unmatched wealth of scientific and historical data available for this region.

The Swiss Re-sponsored study estimates that the overall likelihood for intensity-6 (JMA intensity scale) or greater shaking in Tokyo, Kawasaki, and Yokohama for the next 30 years is in the preliminary range of 35%–40%. The research has also found that while a recurrence of the great 1923 Kanto earthquake was less likely than previously thought, earthquakes similar to the 1855 Ansei-Edo event are larger and more common than previously assumed. Although both events cause JMA intensity-6 (severe) shaking in Tokyo and Yokohama, the Ansei-Edo type event drives the combined probabilities, thus becoming the dominant factor in Tokyo’s hazard landscape in the next 30 years. Even though these results are based on different methods and data evaluations than the governmental study used, they still concur with it well in both qualitative and quantitative terms.

Based on the wealth of new insights, Swiss Re will review – and possibly update – its own earthquake risk assessment model, so that latest scientific progress is properly reflected. This review will be based on published results from both the government- and the Swiss Re-sponsored studies. We encourage the other insurance industry players to analyse these research results and weigh the potential impact and implications they might have.

participants
Designed and maintained by Volkan Sevilgen, USGS Disclaimer