中文摘要：

利用2种高（低）潮序列统计了高（低）潮累积频率，一是从记录原点起算的高（低）潮序列I，一是从月平均海平面起算的高（低）潮序列II。可以发现：不同年份的高（低）潮序列I的累积频率10％（90％）对应的潮位之间存在明显的差异，这主要是由于不同年份月平均海平面变化之间的差异引起的。同时也发现：不同年份高（低）潮序列Ⅱ的累积频率lO％（90％）对应的潮位彼此比较一致，表明用不同年份的高（低）潮序列Ⅱ获得的结果在统计意义上是稳定的。为了得到安全稳定的设计高（低）水位，引人了月平均海平面特征极大（小）值。本文建议，设计高（低）水位取为月平均海平面特征极大（小）值与高（低）潮序列II累积频率10％（90％）对应的潮位之和。

英文摘要：

Two kinds of high（low） tides series are used to calculate the cumulative frequencies of high （low） tides. The one is the high（low） tides series I reckoned from the origin of record, the other is the high（low） tides series II reckoned from the monthly mean-sea-levels. It is found that there exist considera- ble differences among those water levels corresponding to the cumulative frequency 10M（90~） of the dif- ferent years＇ high（low） tides series I, owing mainly to the differences among the different years＇ monthly mean-sea-levels. It is also found that those water levels corresponding to the cumulative frequency 10% （90%） of the different years~ high（low） tides series II are stable in a statistical sense. The characteristic maximum（minimum） of monthly mean-sea-levels is introduced in order to obtain a safe and stable high （low） design water level. In this paper, it is suggested that the high（low） design water level is taken to be the characteristic maximum（minimum） of monthly mean-sea-levels plus the water level corresponding to the cumulative frequency 10 % （90%） of the high（low） tides series II.