以水动力学原理为基础,运用累积距平法、有序聚类分析法、滑动F识别与检验法和M—K趋势与突变分析法4种方法,对珠江三角洲1959--2005年47a的7月和1月水位时间序列进行突变点分析。结果表明:7月平均高水位的变异点集中在1974—1982年和1989—1995年两个时段,1月平均低水位的变异点集中在1970-1980年和1984—1993年两个时段,且大部分潮位站存在一个变异点,少数存在两个。人海口门站1月平均低水位和7月平均高水位序列变异点前,后的均值出现增加的特征,1月平均低水位巴值基本保持不变,7月平均高水位D值呈现增加的特征;在网河区1月平均低水位序列和7月平均高水位序列变异点前/后的均值和G值变化虽均较为显著,但趋势不一致。
With the data of long-term series of mean low water level in January and mean high water level in July from 1956 to 2005 at eight tidal stations in the Pearl River Delta, cumulative departure analysis method, sequential clustering method, scanning F test and Mann-Kendall method were used to explore statistical features of the water level series. Each of those statistical methods has its own weak and strong points, however, comparison of the results from different statistical methods can well satisfy the requirements of reliability of research results. Application of more than one statistical technique in the study can well eliminate the uncertainty of the results by single statistical analysis method. The foregoing analysis and discussions produced the following important and interesting results and conclusions: generally, change points divided mean high water levels in July into two periods, from 1974 to 1982 and from 1989 to 1995; The change-points of mean low water level in January appeared from 1970 to 1980 and from 1984 to 1993. There was only one change-point at most of the stations, but there were two at some stations. This study detected change-points of the water level series and analyzed statistical properties of the water level sub-series posterior and anterior to the change points. The findings of this paper would be helpful for the management of the PRD and human mitigation to natural hazards under the changing environment.