New simulations suggest that the magnitude of tsunamis predicted to hit Japan have previously been underestimated.

New simulations suggest that the magnitude of tsunamis predicted to hit Japan have been significantly underestimated.

The difficulty here is that these predicted giant tsunamis are still expected to be very rare events. It is thus unclear what is more practical, to build things at great cost so that they can survive these rare events, or to live with the risk and rebuild each time after disaster strikes.