A two-dimensional tidal model for the Mediterranean Sea (original) (raw)

1995, Journal of Geophysical Research

The tidal propagation in the Mediterranean Sea is described through a highresolution, two-dimensional hydrodynamic model forced by the equilibrium tide and the incoming tide at the Strait of Gibraltar. The four most significant constituents, M2, S2, K1, and O1, are included in the model. Good agreement with a set of 63 coastal gauges is achieved. The significance of the equilibrium tide and the forcing at the open boundary is investigated. The incoming wave from the Strait of Gibraltar is important in tuning the tides in the whole of the Mediterranean. For the north Aegean Sea the solution without the forcing at Gibraltar results in a doubling of the amplitudes of the semidiurnal tides. western basin is connected to the Atlantic Ocean through the Strait of Gibraltar. Both straits are narrow enough to constrain atmospheric forced waves with periods up to a few days [Garrett and Majaess, 1984]. The Adriatic and Aegean Seas are connected to the eastern basin through the Straits of Otranto and Crete, respectively. The bathymetry of the Mediterranean Sea is quite complex with both the east and west basins being more than 3 km deep in places. The Adriatic Sea extends as a canal with NW-SE axis and can be divided into two basins with different bathymetric characteristics. The north basin slopes down regularly toward the southeast with depths, in general, less than 300 m. The southern basin is much deeper with maximum depths over 1200 m. The configuration of the Aegean Sea is very complex. Its communication with the eastern Mediterranean is through the straits east (Crete-Karpathos, Karpathos-Rhodes) and west (Kithira-Crete) of Crete. Defant [1961] divided the basin into two regions, one that starts from the eastern strait (Crete-Karpathos-Rhodes straits) and extends in a NW-SE axis up to the north coasts of Greece and another which starts at the western strait (Kithira-Crete strait) and is effectively isolated (as far as tidal regimes are concerned) from the northern Aegean by the Cyclades island complex. Apart from the strait of Gibraltar, two smaller openings of the Mediterranean to the Bosporous (NE Aegean) and the Suez (SE Mediterranean) channels are negligible for tidal propagation studies. Paper number 95JC01671. 0148-0227/95/95 J C-01671 $05.00 The tides of the Mediterranean are produced by the addition of the direct action of the equilibrium tide together with the part of the Atlantic tidal wave that can penetrate through the succession of straits [Pugh, 1987]. The interaction of the resulting wave with the bathymetry produces tidal components which exceed values of 10 cm only in certain areas, namely, the north Adriatic, the Gulf of Gabes, and the north Aegean. Enhanced tidal currents known since antiquity are produced in Messina Strait, between the Italian peninsula and Sicily, and in Euripus Strait, between Evoia Island and mainland Greece. It is also worth noting that the second barotropic natural mode of oscillation of the Mediterranean basin is close to the frequency of the semidiurnal tides, thus complicating the propagation of these tidal waves [Schwab and Rao, 1983]. The propagation of the tide through the Strait of Gibraltar has long been a point of disagreement. Several authors [Harris, 1897; Stemeck, 1916; Villain, 1949; Lafuente et al., 1990] state that the incoming Atlantic tide sets up some sort of standing wave in the western basin. On the other hand, Maloney and Bums [1958, p. 2] consider the tide in the Mediterranean as a direct response to forcing by the tide-generating potential and state that "... the erection of a barrier across the Strait of Gibraltar would have little or no effect on the overall tide of the Mediterranean Sea." Defant [1961], in his extensive review of the subject, stated that a great amount of tidal energy penetrates through the Strait of Gibraltar. The employment of simple calculations [Candela et al., 1990] revealed that about 94% of the incident Atlantic wave is reflected at the entrance of the Strait of Gibraltar and that the incoming energy could account for a uniform amplification of 1.4 cm of the equilibrium signal (about 10%). Candela et al. [1990], from analysis of tidal currents in the Strait of Gibraltar, found a net input of energy for the M 2 component of 8 x 108 W but, in view of the associated standard deviation of 10 9 W, concluded that this was consistent with zero net flux. Existing hydrodynamic numerical models do not succeed in describing the tidal propagation in all areas, due to lack of tide gauge data, coarse spatial resolution, unreliable bathymetry, or a combination of these factors. Indeed, only four numerical models covering the whole of the Mediterranean are known to the authors. The first [Dressier, 1980] (hereinafter referred to as DR80), with a resolution of 1/3 ø, involves only the M 2 component. The second model [Vincent and Canceill, 1993; 16,223 16,224 TSIMPLIS ET AL.: TIDAL MODEL FOR THE MEDITERRANEAN SEA 45 40 35 30 I I I lOOO t' 3000 I lO 20 Figure la. Bathymetric chart of the Mediterranean. P. Canceill et al., Barotropic tides in the Mediterranean Sea from a finite element numerical model, submitted to Journal of Geophysical Research, 1993] (hereinafter these are referred to as CA93), with a spatial resolution of 10 to 20 km, could not reproduce the characteristics of the tide by direct forcing alone and therefore had to constrain its solutions at several points to succeed in describing the major characteristics of the tides. Although this may be a useful assimilation technique, especially for use in altimetric studies, it limits any conclusions on the physics of propagation of the tides in the Mediterranean and depends heavily on the reliability of the harmonic parameters at the forcing points. Another model using an assimilation technique based on tide gauge and altimetric data, with a resolution of 1/2 ø, excludes the Aegean Sea [Sanchez et al., 1992] (hereinafter referred to as SA92). Recently, another model based on tide gauge and altimetric data assimilation techniques has also become available through Internet (L. H. Kantha, personal communication, 1994), and a model by C. J. Lozano and J. Candela (manuscript in preparation, 1995) for the M 2 tide only, but which deals with the tidal energetics as well as the tidal propagation, will be available shortly (D. E. Cartwright, personal communication, 1995). The models present conflicting conclusions regarding the role of the Strait of Gibraltar and tidal propagation. DR80 found small differences in the east basin when the Gibraltar Strait was closed, thus concluding that the tides there are due 45 40 35 30 71 Lions 10 12 21 29 42 50 5 "• 11 13 26 ß 36 • ,58 9 37 •