T/P maps (more than 4 years) have been analysed in order to infer the variability in space and time of the sea level in the CANIGO area.

(35 Kb)

The map of the SLA standard deviation (cm) shows the spatial distribution of this variability. We can observe a background variability of 2-3 cm, which is an upper limit of the altimetric error budget. Then three areas of strong variability are well identified. The north-western part of the domain, over the Mid-Atlantic Ridge presents the strongest values (15-20 cm). Along 34°N, a variability between 7-11 cm due to the eddy field associated with the Azores Current and its meanders, which spreads east to 12°W. Then a smaller area south of the Canary archipelago, with values up to 6-8 cm, due to the eddy activity of the Canary Current perturbated on its path through the islands.

The sea level variations can differ from one year to another (i.e., the interannual signal). We have calculated the mean sea level map and the standard deviations separetly for each year:

(18 Kb)		The mean sea level maps (cm) are calculated by averaging all the SLA maps of each year. Then each map is spatially smoothed to filter the small features, in order to look at the "large scale" variations. This signal can vary form -3 to 4 cm. The highest sea level is observed in 1995, which was the warmest year of the period. That is, this sea level elevation may be associated with the dilation of the water layers above the thermocline (i.e., the steric effect). Note also that there is a phase reversal between the years 1994 and 1996.
(58 Kb)		The standard deviation maps (cm) are calculated with the same SLA maps than those used for the mean map calculation above. The general variability distribution pattern is close to the variability over 4 years. But one can see that the Azores Current eddy field is weaker in 1993, and stronger in 1994 and 1995, laying farther east in 1994. The variability is also stronger near the Portuguese continental shelf in 1996. The Canary Current perturbation, south of the archipelago is stronger in 1993 and 1994. Such an interannual variability is not surprising since it was already mentioned with Geosat data (Le Traon and De Mey, 1994) and also in currentmeter long time series (Müller and Siedler, 1992).