Abstract: According to the data of annual maximum and minimum tidal levels as well as annual mean tidal levels of Yangtze River recorded in Zhenjiang and Jiangyin Stations, the trends of the three tidal level time series from 1950 to 2009 were detected by using two non-parametric statistical methods (Mann-Kendall test and Spearman rank correlation test), and the jumps of the above three tidal level time series were examined with three test methods namely moving t-test, moving rank sum test and Pettitt test. All the results were of high coincidence, showing that there were significant rising trends in both annual maximum and minimum tidal levels, while such rising trend in annual mean tidal levels was insignificant at Zhenjiang Station. For Jiangyin Station, only the annual maximum tidal level had obvious rising trend, while the trends of the annual mean and minimum tidal level were not significant. Results of the three tests for jumps showed that the most significant abrupt changes in annual maximum and annual mean tidal level time series occurred mostly around 1979 and an abrupt change in annual minimum tidal level took place in 1987 at Zhenjiang Station. A significant abrupt change in annual maximum tidal level at Jiangyin Station also occurred in around 1979, but there were no significant variation detected in its annual mean and minimum tidal level. With annual maximum tidal level as an example, the causes of tidal level rise were analyzed. This nonstationarity of tidal level might generate negative impacts on hydrologic design, including the increase in difficulty of hydrologic frequency analysis and the decrease in flood control standards of hydrologic design values.

Key words: annual maximum tidal level, annual minimum tidal level, annual mean tidal level, nonstationarity, hydrologic design