Coastal altimetry for the computation of a Mean Dynamic Topography in the Mediterranean sea

Satellite Sea Level Anomaly (SLA) observations are crucial in an operational oceanographic system due to their high coverage on sea surface currents and elevation and their strong constraint on water column integrated steric contributions. The use of Sea Surface Height (SSH) measurements by altimeter satellites in the Mediterranean Forecasting System (MFS) requires an accurate Mean Dynamic Topography (MDT) field with a high horizontal resolution which must be added to SLA observations. Here a new MDT computed through a direct method is proposed to solve the main limitations to the current MDT, evaluated from a model-dependent first guess. The direct method consists in the difference between an altimetric Mean Sea Surface Height (MSSH) and a geoid model. Moreover, a novel altimetric dataset reprocessed near the coast is adopted in order to improve the representation of coastal dynamics. Altimetric data from a single satellite, Jason-2, are used to generate a SSH dataset. This is used along with the EGM2008 geoid model to compute along track MDT observations. Optimal Interpolation algorithms are used to regrid along track MDT on MFS model grid. Derived geostrophic velocities are then computed. The validation of the altimetric dataset against the operational dataset showed improved performances in terms of time series completeness and standard mean error. From the analysis of the MDT and the retrieved geostrophic velocities we can conclude that the direct method allowed us to reconstruct basin scale and large scale MDT features but not meso/small scale and coastal dynamics. Main limitations in our results are due to the low accuracy of geoid model and the Jason-2 tracks spacing.