engleski

Tidal response of the Adriatic Sea inferred from altimetry hydrodynamic modeling and gauge data

Satellite altimetry has reached such a level of precision that it allows direct evaluation of marine tides, raising the question of comparison with other methods. Although the Mediterranean Sea is not known for pronounced tides, in some of its parts, like the northern Adriatic, shallow topography contributes to harmonic amplitudes of appreciable magnitude. Here we gauge the success of 4 altimetric and hydrodynamic models in relation to the established image of the Adriatic tides. To that end, the harmonic constants for 33 Adriatic coastal and island stations have been compiled, and a cubic fit procedure used to extend that information along Adriatic Topex/Poseidon (T/P) track segments. From the first 165 T/P cycles on these tracks we have also derived the constants for the diurnal (K1, O1) and semi- diurnal (M2, S2) harmonics by direct harmonic analysis of the altimetric residual sea surface heights. The gauge and altimetry data have then been compared to the same output from 4 recent tidal models (POL - Tsimplis at al., 1995; CAV - Canceill at al., 1995; CSR3.0 - Eanes, and Bettadpur, 1995; and FES95.2 - LeProvost at al., 1996), the latter two incorporating T/P altimetry, the CAV model using Adriatic gauge data (Ancona), and the POL model using neither gauge nor altimeter data. Summarizing, less difference has been observed among the models, than between the model and empirically derived solutions. Comparison of descending track solutions in terms of amplitude and phases reveals more model-empirical solution variability at the track end with less pronounced tidal signal (lower end for K1, upper for M2 - near amphidromy). Apart from POL K1 phase anomaly (departure from other solutions by some 10 degrees throughout the track segment), all 4 models predict the M2 (K1) amplitude within about 1 (0.5) cm mutual/relative difference, maintaining the phase discrepancies within about 5 degrees. The FES M2 and K1 solutions follow the CAV so closely, in both amplitude and phase, that hydrodynamically they clearly present the same solution. For both M2 and K1 constituents the CSR and FES predictions exhibit similar behavior, but for M2 the FES predicts ~0.6 cm larger amplitude than CSR , whereas for K1 it is the other way around. Their relative phase difference remains within about 2 degrees or less, everywhere except near the amphidromic point. The gauge-fitted alongtrack solution exhibits the smallest amplitude for both M2 and K1, while phase behavior varies depending on the harmonic and part of the track. Both the gauge and their cubic-fit extended track point values are typically at variance with model solutions, in Greenwich phases for example by 20-50 %. The in- and out-of-phase components of M2 and K1 derived from T/P altimetry along Adriatic track segments typically agree rather well in sign and value with coastal and island stations.

satellite altimetry; tidal modeling; Adriatic Sea

nije evidentirano