The Danish Meteorological Institute (DMI) operates an ocean forecasting system for the Baltic Sea and North Sea systems. The DMI model system is based on the Federal German Maritime and Hydrographic Agency (BSH) ocean circulation model (BSHcmod; Dick, 2001). BSHcmod was implemented at DMI in 2001 and has been running operationally there since then. At DMI the model code is subject to a continues development and optimization and is named DMI-BSHcmod. The model comprises a coarse resolution model region that covers the Baltic Sea and the North Sea, and a two-way nested region of the Kattegat, Danish Straits and western Baltic Sea (see Figure).

BSHcmod is a 3D, baroclinic circulation model that is using the hydrostatic and Boussinesq assumptions. The primitive equations are discretized on an Arakawa C-grid. Tracers are advected using a flux corrected transport scheme to avoid negative concentrations and overshooting. The coefficients for horizontal eddy viscosity and diffusion are parameterized according to Smagorinsky (1963), whereby they are dependent upon the horizontal grid spacing and deformation field. A mixing scheme has been implemented in DMI-BSHcmod, based on the k-ω turbulence model extended for buoyancy affected flows (Umlauf et al., 2003). It has parameterization of both breaking surface and internal waves, three different algebraic stability functions are applied for the vertical diffusivities of momentum, heat and salt (Canuto et al., 2002). The model domain has two open boundaries: One in the northern North Sea between Schotland and Norway and the other in the English Channel (see Figure). On the open boundaries the model is forced by surges, tides and lateral temperature and salinity fields. Surges are calculated from a barotropic model of the North East Atlantic Ocean. The tides are determined from tidal constituents at the boundaries. On outflow through the open boundaries newly updated salinity and temperature data are stored in buffer zone so that these values are applied during subsequent inflow. The model includes river run-off from 79 rivers, either in real time or calculated from a hydrological model or from climatology. The model is coupled to sea ice model which includes equations for both the dynamical (Hibler, 1979) and thermodynamical (Parkinson and Washington, 1979) development. On the surface the model is forced by hourly meteorological fields (winds at 10 m, air temperature at 2 m, mean sea level pressure, relative humidity and cloud cover) from DMI's operational numerical weather prediction model DMI-HIRLAM (High Resolution Limited Area Model) which provides 60 hours forecasts four times a day. The surface heat flux is parameterized using a bulk formulation. Data-assimilation of observed sea surface temperature (SST) is included in the model. The SST assimilation procedure uses an optimal interpolation scheme in combination with a simplified Kalman filter. The method is further described by Høyer and She (2007) and Larsen et al. (2007).

The operational North Sea-Baltic Sea model is set up on a 1/6° × 1/10° grid (~6 nautical miles). In this domain there are 50 levels with a top layer of 8 m, 2 m grid spacing down to a depth of 80 m, and increasing grid spacing down to the deeper parts. The two-way nested region covering Kattegat, the Danish Straits and the western Baltic Sea has a grid spacing of 1/60° × 1/36° (~1 nautical mile).

Canuto, V. M, A. Howard, Y. Cheng and M. S. Dubovikov, Ocean turbulence. PartII: Vertical diffusivities of momentum, heat, mass, and passive scalars, Journal of Physical Oceanography, 32, 240-264, 2002.

Dick, S., E. Kleine, S. H. Müller-Navarra, H. Klein and H. Komo, The Operational Circulation Model of BSH (BSHcmod)-Model description and validation, Berichte des Bundesamtes für Seeschifffahrt und Hydrographie, 29, 49 pp., 2001.

Hibler, W. D., A dynamic thermodynamic sea ice model, Journal of Physical Oceanography, 9, 815-846, 1979.

Høyer, J. L. and J. She, Optimal interpolation of sea surface temperature for the North Sea and Baltic Sea, Journal of Marine Systems, 65, 176-189, 2007.

Larsen, J., J. L. Høyer and J. She, Validation of a hybrid optimal interpolation and Kalman filter scheme for sea surface temperature assimilation, Journal of Marine Systems, 65, 122-133, 2007.

Parkinson, C. L. and W. M. Washington, A large-scale numerical model of sea ice, Journal of Geophysical Research, 84, 311-337, 1979.

Umlauf, L, H. Burchard and K. Hutter, Extending the k-ω turbulence model towards oceanic applications, Ocean Modelling, 5, 195-218, 2003.

Smagorinsky, J., General circulation experiments with the primitive equations, I. The basic experiments, Monthly Weather Review, 91, 99-164, 1963.

The DMI ocean model forecast system currently makes forecasts twice a day with analyses at 00:00 UTC and 12:00 UTC. The hindcast length (prior to the start of the forecast) is 12 hours.

Saved hindcast variabels are: Sea level, zonal and meridional currents, temperature, salinty and ice cover.

Hindcast data are saved in netCDF.

Hindcast data are available via DMI's OPeNDAP server. General information on OPeNDAP can be obtained from http://opendap.org/

Free access. DMI ocean model hindcast data are available from http://mersea.dmi.dk

Model output of coastal sea level is validated against tide gauges on a regular basis. Occasionally, salinity and temperature are validated against available time series.

None.

The DMI model system currently make forecasts twice a day with analyses at 00:00 UTC and 12:00 UTC. The forecast length is 60 hours.

Saved forecast variabels are: Sea level, zonal and meridional currents, temperature, salinty and ice cover.

Forecast data are saved in netCDF. Whenever a new forecast is available the previous forecast is removed from the server.

Forecast data are availabe via DMI's OPeNDAP server. General information on OPeNDAP can be obtained from http://opendap.org/

Free access. DMI ocean model forecast data are available from http://mersea.dmi.dk

Model output of coastal sea level is validated against tide gauges on a regular basis. Occasionally, salinity and temperature is validated against available time series.

Color maps and animations of the forecasted variables are uploaded twice daily to http://ocean.dmi.dk/anim/index.php.