Modelling the geoid and sea-surface topography in coastal areas

Abstract

This paper looks at the relation between the time-averaged level of the sea surface and a gravimertic geoid, as determined in coastal areas. Measurements in local regions can now be accurate enough to demonstrate that the geoid and mean sea level are not even parallel to each other, let alone identical. The accuracy and pattern structure of surface gravity data in some shelf seas is comparable with those on land, so that a marine geoid can be derived from surface data without using satellite altimetry. The geodetic objective is then to combine the two to determine sea surface topography. In principle, gravimetric studies provide the absolute datum so that local oceanographic models on the shelf can be combined with sea surface topography models related to the global ocean circulation. In contrast, sea surface topography information near deep ocean coasts must come from external sources and satellite altimetry used to give the gravity data needed to offset the less good coverage by ship-borne gravimetry.Marine Bouguer anomalies enable two specific problems of gravity anomaly patterns near the continent ocean transition to be overcome. The necessary extension of Stokes' condensation reduction is developed and illustrated along a north-south profile from the Mediterranean across the Cote d'Azur. The effect on gravity of deep ocean water introduces a geoid correction in the form of a dipolar ridge whose amplitude at the shore is about 11 cm. In addition to geostrophic currents, a semi-quantitative model for the thermohaline effects on sea surface topography is discussed in relation to sea level differences between the Atlantic and Mediterranean.In considering appropriate algorithms for local geoid computation, Kirby's Iterative Fourier Combination routine for combining altimetry and surface gravity is extended to account for global sea surface topography. The impact of very fast spherical harmonic analysis algorithms is discussed and a simple physical model is given which explains the short coherence lengths found for the global gravity field. This necessary assumption for any local geoid computation was hitherto purely empirical.Finally, the use of land data such as tide gauges, ellipsoidal heights from GPS, and orthometric heights from first order levelling are reviewed as ways of corroborating geodetic estimates of sea surface topography and its relation to levelling datums. Successful examples are given from southern England.