Home

Registration

Preparation of papers

Presenters

Programme

Contact

Clothilde Langlais^1,2
B. Barnier ², P. Fraunie ¹, D. Jacob ³ and K. Beranger ⁴

(1)Lseet, Toulon France
(2)LEGI, Grenoble France
(3)Max Planck Institute for Meteorology, Hamburg, Germany
(4)LOCEAN, Paris France

Main authors appear in bold

Driving coastal circulation models requires an accurate representation of the fluxes at the air-sea interface. This places a rigorous demand on atmospheric forcing data sets: accuracy, consistency over long periods, and coverage of large areas at sufficiently high resolution. In a shelf area like the Gulf of Lion dominated by highly variable coastal processes, the resolution in space and time of the forcing data sets is an important requirement to models. In this gulf, the complex shelf circulation is mainly influenced by the wind stress curl: wind forcing is shown to drive the high frequency variability, acting at time scales of few days on the upwelling/downwelling system, when low frequency seasonal variability dominates the coastal hydrology. In the case of regional and coastal modeling, the use of atmospheric forcing data obtained by a downscaling of global reanalysis appears as a possible solution to obtained a forcing consistent over long period. However, the relevance of such downscaling meteorological forcing is still an open question, and validation studies are required.
This paper compares two sets of atmospheric forcing in the area of the Gulf of Lions during a ten year period (1990 - 2000): one is provided every 6 hours by the most recent ECMWF reanalysis on a 1.125° resolution grid (ERA 40) and the other is provided every hour by the REMO dynamical downscaling of the former ECMWF reanalysis ERA15 at a 18km resolution. We compare the low frequency signal of the two data sets and investigate the spatio-temporal variability gained with the REMO resolution. Finally the paper analyses the impact of the high resolution REMO forcing, using a ten years long run of a regional Gulf of Lion implementation of the NEMO model at 1/64° (1.25km) resolution and 130 vertical Z-levels.

Poster presentation

Back