TWO KINDS OF STATISTICAL INTERPRETATIONS FOR DESIGN SEA LEVEL IN THE VIEWPOINT OF RETURN LEVEL ESTIMATION AND PREDICTION OF THE FUTURE EXTREME

Abstract

 Two-level design for the protection of coastal zone is a well-worn but worthwhile approach in civil engineering, and it is usually characterized as high/low occurrences of those natural hazards, though it does not convince us of the real scales. Once in a thousnad years against the L2 hazards will be beyond our individual understanding in comparison with the frequent occurrence of the L1 hazards. We propose here a new interpretation of those levels: L2 as the prediction of the coming maximum in the future; L1 as the estimation of the return level, whose impotant key is to employ a single common period for both levels, with an example demonstarated.