A Naval Research Laboratory coastal ocean prediction system has been setup and used to perform a real-time ocean forecast during a rapid environmental assessment 2004 (MREA04) experiment that took place off the Portuguese coast. The system is driven by atmospheric forecasts from the Navy's Coupled Ocean and Atmospheric Prediction System (COAMPS) to produce a nowcast and, up to 72-hour forecast every 24-hr of sea level variation, 3D ocean current, temperature and salinity fields during the experiment.

The NRL coastal ocean prediction system consists a nested (4-km/1-km), 41-level sigma-z data-assimilating ocean model based on Navy Coastal Ocean Model(NCOM). The open boundary conditions including sea surface elevation, transport, temperature, salinity and currents are provided by the NRL 1/8 degree Global NCOM prediction system. A one way coupling scheme is used to ingest those boundary conditions into the coastal model. Hourly surface heat fluxes, as well as solar radiation, wind stresses and sea level air pressure from COAMPS are applied for surface forcing.

The system was spin up for a month, prior to the field exercise, using the COAMPS forcing. During the experiment, the model is restarted from the previous day's nowcast. Once model is restarted it continuously assimilates the temperature/salinity analyses generated from satellite altimeter(GFO, Jason-1, ERS-2) sea surface height anomalies, AVHRR MCSST data, and from in-situ CTD temperature/salinity profiles. A three-dimensional, temperature/salinity estimation was produced from satellite data first using a statistical model called the Modular Ocean Data Assimilation System (MODAS). The CTD temperature/salinity profiles were than combined, using optimum interpolation, with the satellite estimations to produce a new analysis that is assimilated into the model. Forecasts are than produced with available COAMPS forecasts. Once the daily nowcast/forecasts are produced they are distributed through the Internet for evaluation.