Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 74
Edited by: B.H.V. Topping and B. Kumar
Paper 24

Application of Artificial Neural Networks to Tidal Forecasting

D.S. Jeng+, T.L. Lee+ and K.J. Hsu*

+School of Engineering, Griffith University Gold Coast Campus, Queensland, Australia
*Department of Construction and Planning, Leader University, Tainan, Taiwan, R.O.C.

Full Bibliographic Reference for this paper
D.S. Jeng, T.L. Lee, K.J. Hsu, "Application of Artificial Neural Networks to Tidal Forecasting", in B.H.V. Topping, B. Kumar, (Editors), "Proceedings of the Sixth International Conference on the Application of Artificial Intelligence to Civil and Structural Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 24, 2001. doi:10.4203/ccp.74.24
Keywords: neural network, tidal level forecasting.

Information of tide level at a site is required for engineering design as well as operation-related activities in the ocean environments. The tidal level does not only affect marine structure design, but also the groundwater in a coastal aquifer. The beach water table fluctuation due to spring-neap tides is one of example [1]. Thus, knowledge of local tidal-level variation has a direct implication to engineering practice.

Numerous models for tidal forecasting have been carried out in the past. To describe the characteristics of the tide-level variations in an open sea, Darwin [2] proposed a classic equilibrium tidal theory. However, Darwin's model [2] was incapable to accurately estimate the tidal level for the complicated bottom topography, especially in near shore areas. Later, Doodson [3] employed the least-squares method to determine of harmonic constants. This harmonic analysis has been widely for tidal forecasting in the past. The accuracy of harmonic models entirely depends on accurate observed data over a long-term tidal record, which is used to determine the parameters of the tidal constituents. This is the major shortcoming of the harmonic models. Recently, Yen et al. [4] applied the Kalman filtering method to determine harmonic parameters with a limited amount of tidal measured data. However, the model is only applicable for short-term prediction, rather than long-term predication.

In this paper, we establish an artificial neural network model to predicate tidal level. The strength of using ANN model for tide forecasting is the accurate predication with short-term observation data, which has been the shortcoming of harmonic analysis. The observation data in Tanshui Harbour (located in Northern Taiwan) are used a case study here. The numerical results demonstrate the ANN model performs well in long-term tide forecasting (one year) with a short-term observation data (15-day).

L. Li, D. A. Barry, F. Stagnitti, J. Y. Parlange and D. S. Jeng, "Beach water table fluctuations due to spring-neap tides: The moving boundary effects", Advances in Water Resources, 23, 817-824, 2000. doi:10.1016/S0309-1708(00)00017-8
G. H. Darwin, "On an apparatus for facilitating the reduction of tidal observations", Proc. Roy. Soc., Ser. A, 52, 345-376, 1892. doi:10.1098/rspl.1892.0082
A. T. Doodson, "The analysis and predictions of tides in shallow water", Int. Hydrogr. Rev., Monaco, 33, 85-126, 1958.
P. H. Yen, C. D. Jan, Y. P. Lee and H. F. Lee, "Application of Kalman filter to short-term tide level prediction", Journal of Waterway, Port, Coastal and Ocean Engineering, ASCE, 122(5), 226-231, 1996. doi:10.1061/(ASCE)0733-950X(1996)122:5(226)

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description