The Mercator 1992-2002 PSY1v2 ocean reanalysis for tropical and North Atlantic

Eric Greiner⁽¹⁾

⁽¹⁾ MERCATOR, 8-10 Rue Herme, Parc Technologique, 31526 Ramonville St Agne, France

Abstract

-eddy permitting for mesoscale studies and embedding of area limited models -close enough to data to give a good context for specific cruise campaigns -homogeneous enough to allow the analysis of climate time series -continuous enough to use model outputs as bio-geochemical dynamic inputs -using ECMWF ERA-40 reanalysis

Outlines of the reanalysis will be exposed. The following performances will be detailled :

- SLA error < 5cm (analysis) - 8cm (forecast) - SST error < 0.7°C - Temperature error < 0.7°C (a) - 1.5°C (f) - Salinity error < 0.3psu

- good balance analysis/forecast - no shocks but some local transitories - good circulation, no bad region

Except some local biases up to 10cm/0.7°C, the coherence between the altimetry+MSSH, the in situ data and the model forced by ERA-40 is really good after assimilation. In particular, the interannual signal is reproduced in a fully coherent way.

The 1992-2002 decade was strongly impacted by the NAO change. In 1994, a warm anomaly accumulated in the South-East of the Gulf Stream entered in this current, and reached New-Foundland in 1995. This strong SST anomalies may have switched the NAO index from the very positive values of the late 1980s and early 1990s, to negative values. This shift onto a negative NAO regime enabled this warm anomaly to continue along the Subpolar Gyre. Meanwhile, thanks to winter deepening of the mixed layer, and lateral eddy diffusion, the anomaly was spread. Yet, the damped warm anomaly reached the Labrador Sea in 1998, and was probably an important factor in the extremely high SSTs of this year. Connexions with the subtropical gyre, the weakening of the Cape Verde upwelling, and the North Sea temperature are also evidenced. Overall, the strong North Atlantic rising of 3mm/y during Topex-Poseidon first decade was mostly an extreme NAO event, favored by a perfect timing of all patterns.

Workshop presentation

Full paper