Black Sea

After over 5 years of daily forecasts, the system was discontinued on 01/March/2017; results up to that date remain online on our ftp site. Work is ongoing on another demonstration project that will be put online here in the near future.

GHER model sea surface elevation [meter]

GHER model sea surface temperature [°C]

Every day, one week of forecasts for the Black Sea surface elevation and temperature ensemble forecasts are shown here. They are the mean of a data-assimilating ensemble of models. The complete results are put online here.

The numerical ocean model is the GHER model. Initialization is obtained from the HYCOM+NCODA global 1/12° analysis from the Naval Research Laboratory (USA). Atmospheric data is obtained from the GFS model from the National Centers for Environmental Prediction (USA). Satellite and in situ observations are obtained through the MyOcean and Coriolis portals. River discharge and Bosphorus flow are obtained from climatology.

An ensemble of models is used to evaluate expected model uncertainty and as a base for data assimilation. Realistic perturbations are applied at the same time to the initial conditions (using WCE 1.0), to the atmospheric fields (using EOF decompositions and recompositions, our perturbations being of the order of the difference between NCEP and ECMWF forecasts), to the river inflow, and to some internal model parameters. The ensemble has no fixed size; members are created and run as long as there is time… Every day, ARGO profiles (if available) and sea surface temperature are assimilated into the ensemble, using the Ocean Assimilation Kit software. The forecast and associated uncertainty estimates are verified a posteriori with subsequent observations.
0) { $now_running=”ensemble member $ca”; }
else { $now_running=”unknown state !”; }
echo “The server is currently running: $now_running”;
?>

You can download the raw data as netcdf files from our ftp server

Disclaimer: this is a demonstration product provided free of charge. seamod.ro does not warrant or suggest that this data is fit for a particular application. If you use it, please cite sources.