You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »


image © Enjoylife2/iStock/Thinkstock 

Dynamic Ocean Model - NEMO

Purpose

The dynamic ocean model used for medium-range and seasonal forecasts of ocean structure is the Nucleus for European Modelling of the Ocean (NEMO).  It is coupled with all IFS forecast models (ensemble control forecast, medium range ensemble, sub-seasonal range ensemble, seasonal ensemble).  

The medium range ensemble uses the atmosphere-wave-ocean coupling framework from the start of the forecast.  This is because it is important to capture two-way feedback between the atmosphere and the sea-surface temperatures, sea-ice extent and ocean waves (e.g. a slow-moving tropical cyclone can cool the sea surface).  

The atmospheric model covers the whole globe.  However, the ocean model has the additional problem of lateral boundaries along coasts causing effects such as boundary currents (e.g. the gulf stream).  Also near the continents the sea depth decreases abruptly at the continental shelf.  A new bathymetry (introduced Cy50, May 2026) is used to better represent the sea bed around coasts.

For both atmospheric and ocean models it is important to assimilate sea surface temperatures and also the salinity and temperature structure within the ocean itself.

Throughout the forecast period, NEMO provides the oceanic temperature structure near and at the surface.   ECWAM provides wave data, and therefore an indication of surface roughness.   Full coupling, incorporating sea roughness effects, governs the fluxes of heat, moisture and momentum into the lowest layers of the atmosphere.  The formation, evolution and decay of ice over open waters is controlled by SI3 (part of NEMO).  In effect, NEMO and SI3 together move ice around (according to ocean drift etc.) and melt or form ice (according to sea-surface temperatures etc.). 


The sea surface temperature has a strong effect upon the development, evolution and persistence of tropical storms.   Equally it's important to assimilate the temperature and salinity structure of the ocean.  Surface wind stress associated with tropical storms, and even that associated with mid-latitude depressions, can induce upwelling that brings cooler water up to the sea surface.  This can affect the subsequent evolution of weather systems.  A turbulence scheme is included to better model vertical mixing processes in the ocean.

Some improvements in forecast results:

  • Modifications to the analysis and forecast scheme reduces the known sea surface temperature warm bias in the Southern Ocean and in the Gulf Stream off Newfoundland.
  • Hourly surface forcing from ERA5 atmospheric reanalysis brings much more accurate representation of short-term variability (e.g. diurnal sea surface temperature changes). 
  • NEMO uses SI3 for more comprehensive handling of sea-ice and albedo.   NEMO used LIM2 and climatological albedo values in Cycles in Cy49r1 and earlier. 


NEMO Structure

NEMO is a three-dimensional general circulation ocean model.  It can reproduce the general features of the circulation and the thermal structure of the ocean with seasonal and inter-annual variations.  It is a primitive equation model adapted to simulate regional and global ocean circulation. Improved handling of sea-ice (Cy50, autumn 2026 and later) allows the model grid to extend closer to Antarctic than cycle 49r1 and earlier.

The observed and prognostic variables are:

  • Sea-surface temperature.

  • Sea-ice information on location and extent.

  • Sea level anomaly information (wave heights).

  • Ocean bathymetric temperature and salinity.

  • Three dimension velocity field.

The ocean model has different resolutions.  For use with:

  • the medium range and sub-seasonal models resolutions,
    • horizontal resolution: ~0.25 degree (~28km) for medium range and sub-seasonal atmospheric models.  
    • vertical resolution: 75 ocean levels and the vertical resolution varies with depth.   Levels are closest together near the sea surface (18 levels in the first 50m, the top level is <1m depth) to capture the temperature and salinity structure of the uppermost layers and the thermocline.
  • the seasonal model resolutions:
    • horizontal resolution: ~1 degree (~112km).
    • vertical resolution: 42 ocean levels and the vertical resolution varies with depth.  Levels are closest together near the sea surface to capture the temperature and salinity structure of the uppermost layers. 

Sea-ice model (SI3)

SI3 ("sea-ice cubed") is a prognostic sea-ice model that deals with the dynamic and thermodynamic evolution of the sea surface.  This enables sea-ice cover to evolve dynamically.  It has an improved representation of the seasonal evolution of the ice.  It has:

  • multi-category ice representation with prognostic salinity and melt pond dynamics.
  • spatially and temporally varying sea-ice salinity with turbulence improving vertical mixing processes.
  • development of melt ponds.
  • new prognostic variables to describe the thermodynamic sea-ice properties.

The extent of sea-ice, and the variation of the ice shelf, have important effects upon the energy and moisture balance at the atmosphere/surface boundary.  This is incorporated into the dynamic ocean model.  The ice extent will change through the forecast period in response to sea temperatures and air temperatures, ocean currents and wind.  The varying areas of sea-ice diagnosed by sea-ice model SI3 are used to define albedo over sea-ice for each grid square.  Any snow on the ice surface is modelled as a single layer (unlike the four snow layers modelled over land).  Introduction of the sea-ice model SI3 reduces the warm bias seen in winter over the ice surface, especially in cloud free situations.

SI3 replaces LIM2 ice model (Louvain-la-Neuve sea ice model Version 2) used by Cy49 and earlier.


 

Fig2A.4-1: Schematic representation of how SI3 handles sea-ice.   In each grid square, the fractional cover of up to five categories of sea-ice are identified (ie several tiles).  Categories include shallow or thick ice, any of which may have snow cover depending on snowfall produced by the atmospheric model.  Each category has four evenly spaced ice layers and thermodynamic processes within SI3 act on these and govern melting or formation of the ice in each category.  Any snow on the ice surface is modelled as a single layer (unlike the four snow layers modelled over land).  As ice forms or melts, there is a redistribution in the fraction of each thickness category.  Ridging and rafting of ice under pressure is included which alters the roughness, and snow can be blown from the ice surface according the forecast winds.  The resulting sea-ice fraction and surface characteristics define the albedo of the grid square.  The dynamics of NEMO is then solved for the whole grid box not the categories.    


Data Assimilation - NEMOVAR

NEMOVAR is a three-dimensional variational assimilation (3D-Var) system adapted to the NEMO model.  Ocean temperature and salinity vary only slowly and a 3D-Var system is used because there is little need to fit observations to a precise time.  NEMOVAR output provides initial conditions for the coupled model and also provides a first guess for the next NEMOVAR assimilation cycle.   

The ocean analysis system consists of a real-time stream (ORTS6) and a reanalysis stream (ORAS6).   OSTIA serves as a crucial, high-resolution (0.05°) boundary condition for sea surface temperature  and sea-ice concentration though it can be several days delayed.

Coupled ocean-atmosphere assimilation of microwave imagers and geostationary infrared data gives increments to ocean and sea-ice analyses. as well as upper air.

ORTS6 - Real time assimilation. 

  • Sea surface temperature

    • Information on sea surface temperature (including large lakes e.g. Great Lakes, Caspian Sea, Sea of Azov) are assimilated from:
      • latest Operational Sea Surface Temperature and Sea-Ice Analysis (Met Office OSTIA, re-gridded to IFS resolution). Note OSTIA data can be up to 2 days old.
      • skin sea-surface temperature estimates from satellite microwave imagers (AMSR2 and GMI). These are used in near surface air temps in coupled assimilation.
      • other sea-surface temperatures (ships, buoys etc) are not assimilated directly but are relaxed towards the microwave sea surface temperature data. 

    • Sea-surface temperatures are adjusted to be consistent with the sea-ice concentration in a grid square.  If the sea-ice concentration is:

      • higher than a given threshold (currently 55% cover) then the model sea surface temperature is set to 0°C.
      • lower than the given threshold (including if no ice) but the model sea surface temperature is <0°C then the observed sea surface temperature is also taken more into account.
    • Statistics on availability and assimilation of sea surface temperature information are available.

  • Sea-ice information

    • Information on sea-ice location and extent (including large lakes e.g. Great Lakes, Caspian Sea, Sea of Azov) are available from:
      • Operational Sea Surface Temperature and Sea-Ice Analysis (Met Office OSTIA, re-gridded to IFS resolution).
    • Statistics on availability and assimilation of sea-ice concentration are available.

  • Sea level anomalies 

    • Information on sea level anomaly and significant wave heights are available from:
      • satellite altimeter data.  Sea level anomaly data improves
        • the representation of the seasonal and inter-annual variability of sea level changes.  This is important for sub-seasonal and seasonal range weather forecasts.
        • the representation of ocean currents.
        • long term check on ocean heights as a consequence of global warming.
    • Statistics on availability and assimilation of sea level anomaly and wave height data are available.

  • Ocean bathymetric temperature and salinity

    • Bathymetric information is available from:
      • ARGO floats (profiling floats):  These take measurements as they sink to about 2000m then return to the surface to download the information via satellite.  The cycle is then repeated. 
      • AXBTs (Airborne eXpendable Bathy Thermographs): These floating buoys are  airdropped and then deploy long wires with temperature sensors. These sample and transmit ocean temperatures down to about 500m.  AXBT data is frequently used to increase knowledge of the ocean structure along the forecast track of a tropical storm and is an important factor in determining the intensity a hurricane may reach.
      • instrument-carrying marine mammals:  These carry sensors that download measured information via satellite when the animal returns to the surface.  Depths reached can be as much as 1500m below sea level.
      • moored buoys:  Some moored buoys have sensors suspended at several levels below the ocean surface.
      • CTDs (Conductivity(salinity), temperature, depth):  These take measurements as they are lowered 
    • Bathymetric observations are not used in regions where the model sea depth is <500m.  This avoids assimilating data on the continental shelves where the model has poor representativeness.
    • Statistics on availability and assimilation of bathymetric data are available.


ORAS6 - Ensemble of Data Assimilations: 

It is important to represent uncertainty in the ocean initial conditions and in model structure.  ORAS6 (an oceanic EDA system) produces 11 perturbed analyses (temperature/salinity profiles, sea level anomalies or waves, and sea-ice concentration).  These contribute, through ocean-atmosphere coupling, to the ensemble of forecasts used for probabilistic predictions at medium range, sub-seasonal range, and seasonal range.

ORAS6 provides ocean and sea-ice initial conditions for all ECMWF coupled forecasting systems.  It is the ocean and sea-ice ensemble reanalysis, and provides surface boundary conditions fo use with atmospheric reanalysis ERA6.  The area covered extends further towards Antarctica than cycles Cy49 and earlier.

The perturbation of the the sea-surface forcing are based on uncertainties derived from model climatological data derived over an extended period (1950 to date).  These include:

  • perturbations of sea surface temperature, sea-ice concentration, wind stress, and fluxes of heat and freshwater.  
  • a flow-dependent background error covariance using ensemble of data assimilations.
  • a variational assimilation of sea surface temperature (SST) with flow-dependent errors.

Sea-ice thickness properties in ORAS6 have changed substantially from ORAS5 due to differences in forcing and the sea-ice model.  In some regions of the Arctic, the sea-ice in ORAS6 can be biased too thin.  There is a more accurate representation of short-term variability (e.g. diurnal cycle os surface sea temperatures).

Information on ensemble reanalysis system for ocean and sea-ice: ORAS6 is available.

Example sea-surface temperature (SST) and ice concentration

Sea-surface temperature are initialised using:

  • analyses received daily from the Met Office (OSTIA, 5 km resolution).
  • NEMO and SI3.  

NEMO forecasts changes in the sea-surface temperature (SST) and SI3 forecasts sea-ice evolution.  These are used interactively by all IFS atmospheric models.  Medium range ENS and sub-seasonal range ENS use the same initial ice extent.  See also remarks on water surface temperature and sea-ice. 


Fig2A.4-2: Sequence of sea-ice and sea-surface temperatures from the medium range ensemble control run data time 00UTC 27 April 2017.  T+0hr (00UTC 27 April 17), T+120hr (00UTC 02 May 17),  T+240hr (00UTC 07 May 17), and  T+360hr (00UTC 12 May 17).  On such plots the climatological average sea-ice cover is shown in pink (contour and stippling, for >50%), just discernible in the northern Gulf of Bothnia and in the White Sea.   Dark purple areas (SST between 0C and -2C) are prone to ice formation if not already in existence.   Areas of sea-ice are shown as turquoise. 

Note: 

  • Movement of ice (turquoise) in the northern Gulf of Bothnia  due to the winds.
  • Steady rise of sea-surface temperatures in the Black Sea, and especially in the shallow waters of both the Sea of Azov and the northern Caspian Sea.  In the White Sea (east of Finland, top of plot) sea-ice cover is less than the climatological average for this time of year.  Using these plots, the user can assess where sea-ice cover is above/below average.

Considerations

The impacts of differently-evolving distributions of sea surface temperature and ice cover should be considered when comparing different forecasts, even when they are from the same data time. 

Additional Sources of Information

(Note: In older material there may be references to issues that have subsequently been addressed)


See also Section on Atmosphere/Ocean coupling.


(FUG Associated with Cy50r1)


  • No labels