The marine component of the Earth system has an important influence on the atmosphere on a range of timescales. We are developing a fully coupled model of the marine system including the surface waves, ocean and sea ice.

It has long been known that the waves affect the marine boundary layer of the atmosphere by modifying the surface roughness. All forecast systems at ECMWF are coupled to an ocean model. The ensemble and seasonal forecast systems use a coupled atmosphere-ocean model, which includes a simulation of the general circulation of the ocean and the associated coupled feedback processes that exist.

Ocean wave modelling

ECMWF uses and develops the WAve Model (WAM) model. It is coupled to the atmospheric model or run as a standalone model in the Limited-Area Wave (LAW) configuration.

Since 1998 ECMWF has been running a coupled forecasting system where the atmospheric component of the Integrated Forecasting System (IFS) communicates with the wave model (WAM) through exchange of the Charnock parameter which determines the roughness of the sea surface (Janssen, 2004).

Our wave prediction system is based on a statistical description of oceans waves (i.e. ensemble average of individual waves). The sea state is described by the two-dimensional wave spectrum which gives the distribution of wave variance over different frequencies and propagation directions. Wave energy then follows from the product of water density, acceleration of gravity and wave variance.

The evolution of the wave spectrum follows from the energy balance equation which is explicitly solved by the WAM model. It determines the rate of change of the wave spectrum by adiabatic processes such as advection  and refraction of wave energy, and by physical mechanism such as wind-wave generation, nonlinear four-wave interactions, and dissipation of energy by processes such as wave breaking, micro-scale breaking and bottom dissipation.


Ocean modelling

ECMWF uses the community ocean model NEMO (Nucleus for European Modelling of the Ocean) as part of the IFS. The NEMO model provides the dynamic ocean model used in the ensemble prediction system and the seasonal forecast system (S4). The ensemble prediction system of the medium and monthly range forecasts runs the ocean model at 0.25 degree horizontal resolution with 75 levels in the vertical and is initialised with the NEMOVAR (3D variational assimilation system) OCEAN5.  The seasonal forecast system (S4) uses a 1 degree horizontal resolution with 42 levels in the vertical and is initialised with NEMOVAR OCEAN4.

Since 2013 the ensemble forecasts have coupled the atmosphere-wave-ocean model from the start of the forecast. This is important to allow capturing the two-way feedback between the atmosphere and the sea surface temperatures, for example when a tropical cyclone is slow moving it can cool the sea surface.

Since November 2014 the ensemble forecasts have been run with 0.25 degree horizontal resolution with the sea ice model active.

Sea ice modelling

Sea ice is an important component of the Earth system; it is highly reflective, altering the amount of solar radiation that is absorbed; it changes the salinity of the ocean where it forms and melts, and it acts as a barrier to the exchange of heat and momentum fluxes between the atmosphere and ocean.

The current operational ensemble forecast systems model sea ice dynamically using the LIM2 model within NEMO ocean model to represent the dynamic and thermodynamic evolution of sea ice within the coupled forecast system.