عنوان مقاله [English]
The present article addresses a new approach form using tidal analysis of long and short observations of tidal heights by tidal tools of Mike 21 software. Predicted or observed tides must be available as open boundaries in a hydrodynamic model to simulate tidal currents. Mike 21 is a powerful software in simulating of marine and oceanic parameters. Hydrodynamic models use tidal data as a main input in the open boundaries of 2D/3D simulation. The accuracy of the simulation results from tides and tidal currents as well as sediment transportation and other related marine parameters, is quite dependent on accurate input data in the open boundaries of the hydrodynamic model.
Large and meso-scale simulations usually use sophisticated global tidal models. These tidal models are prepared by assimilating of satellite altimeter data and tide gage data. The accuracy of such a tidal input seems to be quite sufficient. In shallower part of ocean, like estuaries and coastal areas, these data are not so accurate; therefore direct tide observations are important. On the other hand, long-term tide observations are limited due to technical and economical constraints.
In this study, the appropriate solution to tidal analysis and predictions was introduced in the Mike 21 software to produce acceptable and accurate tidal height data. Mike 21 is provided with two different tidal analysis and prediction modules. The IOS and Admiralty methods of tidal analysis and prediction are known for the corresponding high quality functions. Using the software, the tidal analysis of sample data was discussed in order to produce the accurate tidal data. In this article, hourly tidal observations for a year which belonged to three stations in the Oman Sea and Persian Gulf were analyzed. All tidal analyses were performed for one year, 30, 15 and 10 days periods using Canadian IOS and Admiralty method software. The predicted tidal data were compared with respect to observed data for evaluating of computations. Residuals normally will describe such discrepancies, if there are no abnormal fluctuations of tides. These abnormal inconsistencies might be produced by storm surges and other non-tidal effects. The presented results were statistically analyzed. Two different statistical indices, Root Mean Square Error and Reduction in Variance, were used to evaluate predicted tide data. Consequently, the tool for appropriate analysis and predicting of tidal data of the different periods was introduced. It is shown that for long term tide observations, the IOS method is the proper solution for tidal analysis and prediction, while the Admiralty method is fitted for short term analysis and prediction of tides.