Analysis And Prediction Of High Tide Level In The Pearl River Estuary

The Pearl River Delta is one of the most densely populated and economically developed regions in China.The fluctuation of high tides has a profound influence on the safety of shipping,embankment and water supply.Hence,understanding and forecasting the high tide level precisely of the Pearl River Estuary are of significance to the production and life of the region.The mean monthly high tide level series of Hengmen and Modaomen among the eight major outlets of the Pearl River from 2001 to 2015 were studied.The research focused on the fluctuation rules and identification of main components of the mean monthly high tide level.Meanwhile,the influences of various hydrometeorological factors on the mean monthly high tide level in the Pearl River Estuary,such as air pressure,temperature,wind speed,runoff and West Pacific water level,were analyzed.Predictions of the mean monthly high tide level were made in the end.Analyses revealed that the mean monthly high tide level series of Hengmen and Modaomen from 2001 to 2015 showed significant upward trends.Fourier transform,continuous wavelet transform and maximum overlap discrete wavelet transform were used to analyze the periodic fluctuations of mean monthly high tide level series in Hengmen and Modaomen.It turned out that the mean monthly high tide level series mainly consist of Sa(solar annual)tide,Ssa(solar semiannual)tide,and 1/3 annual tide.Sa tide is stationary,while the short period components including 1/3 annual tide display non-stationary,reflecting seasonal fluctuations of the high tides.Further,wavelet coherence was applied to analyze the correlation between the mean monthly high tide level and the corresponding five related factors of air pressure,temperature,wind speed,runoff,and West Pacific water level.It was revealed that the seasonal variation of mean monthly high tide level is coherent with runoff,while the annual fluctuation is coherent with the air pressure,temperature,runoff and West Pacific water level.Moreover,ARIMA and NAR neural network were adopted to forecast the mean monthly high tide level of Hengmen and Modaomen respectively,and the NAR provided better forecast performance.