A real-time coastal ocean prediction experiment for MREA04

To provide a short-term ocean forecast for sea level variation, current, temperature, and salinity, an ocean nowcast/forecast system has been developed. The system is an integration of a data-assimilating, dynamical ocean model, a statistical data-analysis model, and various data streams for ocean bathymetry, climatological data, surface forcing, open boundary forcing, and observations. The system assimilates satellite data and in-situ measurements to produce an estimation of the current ocean state or nowcast and is forced with a meteorological forecast to produce an ocean forecast. During the MREA04 sea trial, the system was implemented for a region off the Portuguese coast with two-way nested grids and produced real-time ocean forecasts for the period of the experiment. The high density of real-time, in-situ observations during MREA04 provided a unique opportunity for the system to assimilate the in-situ observations in addition to satellite data and to perform a statistically meaningful evaluation of the system's forecast capability. The evaluation shows that the nowcast/forecast system has good skill in predicting the tide and fair skill in predicting the ocean temperature and salinity with overall rms errors of 0.5 °C and 0.15 psu for temperature and salinity, respectively. Assimilating in-situ CTD data produced a better nowcast/forecast than assimilating only satellite data. The forecast error increases as the forecast time increases, but the forecast error does not increase significantly over the nowcast error, which indicates that the error in the nowcast is the major source of the forecast error.