URL: http://www.mercator-ocean.fr/html/actualites/news/actu_psy3v2_en.html
|
URL: http://www.mercator-ocean.fr/html/actualites/news/actu_psy3v2_en.html |
Sea ice modelling and in situ data assimilation in Mercator global systemIn collaboration with Marie Drévillon, Mercator Ocean April 2007: Mercator Ocean is on time for the new version of its global ocean forecasting system. This date is marking the beginning of the Mersea European operational oceanography programme second target operational period.
This is done: Mercator is taking a step forward complexity with the new version of the global model with 2 decisive upgrades:
European integration
For the initiatedNEMO ocean and sea ice model The global ocean model is OPA (ORCA025 configuration), 1/4° resolution, that corresponds to an average of 10 Km at these latitudes. This configuration is based on NEMO (Nucleus for European Models of the Ocean) modelling system developped in collaboration with the Drakkar programme. The sea ice is fully comprehensive with the implementation of the Louvain Ice Model, LIM2 version, developped by Georges Lemaître Astronomy Institute, Belgium. With sea ice concentration, sea ice and snow thicknesses, sea ice drift and sea ice thermal content prognosed by this multi-layer model based on the Semtner [1] 3-layers and the Hibler [2] visco-plastic formulations, forecasts will handle most of the processes linked to the sea ice lifecycle. Vertical resolution In order to better resolve the upper layers, the vertical grid, 50 levels, has been refined at the surface ending with a discretization of 1 meter until 20 meters depth and of 500 meters for the bottom layers. With this surface refined-mesh, this new vertical grid has been designed to improve the circulation in the coastal shelves and to represent more adequately the impact of the atmospheric diurnal cycle, which is planned to be explicitly modelled in the near future. Forcing fields with bulk formulas The atmospheric forcing fields which will drive the future system are computed using the empirical bulk parameterization described by Goosse et al. [3]. A systematic bias in the precipitations is removed thanks to GPCP (Global Precipitation Climatology Project) observations when available, and for the recent and real time analyses and forecast, the bias is removed thanks to a predictor computed from these observations. New parameterizations As shown by Barnier [4] the combination of recent implementation of an energy-enstrophy conserving scheme for momentum advection with a partial steps representation of the bottom topography yields significant improvements in the mean circulation and in the representation of western boundary currents such as the Gulf Stream and the downstream flow of the North Atlantic Current. Moreover, the model solution is often comparable to solutions obtained at 1/6°or 1/10° resolution on some aspects concerning mean flow patterns and distribution of eddy kinetic energy. New assimilation sheme The data assimilation technique that is beeing used is a multi-data and multivariate assimilation algorithm consisting of a Singular Extended Evolutive Kalman (SEEK) filter analysis method. The SEEK filter is a reduced-order Kalman filter introduced by Pham [5] in the context of mesoscale ocean models. This method assimilates jointly satellite Sea Level Anomalies (SLA) and Sea Surface Temperature (SST), and in situ profiles of temperature and salinity. The error statistics are represented in a sub-space spanned by a small number of dominant 3D error directions. And then... Next step: global ocean forecasting system at eddy resolving resolution (1/12°). A North Atlantic basin version is in progress and interannual experiments including all the updates listed above have been already performed. Useful links
Bibliography
[1] A J Semtner, 'A model for the thermodynamic growth of sea ice in numerical investigations of climate', J. Phys. Oceanogr., 6, 379-389, (1976).
Copyright © 2001-2007 Mercator Ocean. All rights reserved.
| |||||||||||||||||||
)
)
)
)
)
)
