Two years of Geosat data were used to systematically investigate the mesoscale variability and spatial scales of the North Atlantic. We first calculated sea surface height variability and eddy kinetic energy. These compare well with the eddy kinetic energy obtained with surface drifters. Then the spatial scales of mesoscale variability were characterized in 10° (latitude)×10° (longitude) areas by their wave number spectra and autocorrelation functions. Comparison with existing in situ data shows good agreement, and it is believed that the total error budget is globally below that of the oceanic signal. The wave number spectra show significant differences as a function of latitude and longitude, and these can be related to different types of forcing (instability of a mean current or fluctuating wind). Scales typically decrease from west to east and south to north. Simple proportionality with respect to the first internal Rossby radii does not apply everywhere.

We study the influence of the wet tropospheric correction on mesoscale dynamic topography as derived from satellite altimetry. For this purpose, we use Geosat altimeter data in the northeast Atlantic, and we process separately the tropospheric correction derived from the PERIDOT model following the technique for analysing altimeter height profiles. We show that the humidity spatial scales are larger than those of the dynamic topography. Consequently, the wet tropospheric effect is considerably reduced when orbit error corrections are applied; after long‐wavelength filtering, the along‐track rms of the correction is 13% of that of the oceanic signal. In addition, contrary to the dynamic topographic signal, the tropospheric signal is temporally decorrelated between two successive adjacent tracks so that the wet tropospheric contribution to the objectively analysed synoptic maps is only 10% of the signal of mesoscale variability.

No abstract

No abstract

No abstract