Impact of Surface Forcing on the Forecast Skill of NCOM-4DVAR

The quality of the surface fluxes that are used to force an ocean prediction system can have a significant impact on the system’s overall prediction skill, especially in coastal regions where the flow can have the tendency to be more ageostrophic. In this study, four one-year-long experiments (May 2013 – April 2014) were performed with the Navy Coastal Ocean Model 4-Dimensional Variational Assimilation (NCOM-4DVAR) system along the coast of southern California. NCOM-4DVAR is an analysis and forecasting tool that employs the representer method to assimilate temperature, salinity and sea surface height observations from satellites, buoys, gliders, and drifters in all 4 dimensions; NCOM is then used to propagate a forecast from the analysis. The four experiments differ only with the surface forcing; each experiment is forced with surface fluxes from one of the following different sources: 1) Coupled Ocean Atmosphere Mesoscale Prediction System (COAMPS), 2) Navy Global Environmental Model (NAVGEM), 3) COAMPS with NFLUX corrections, and 4) NAVGEM with NFLUX corrections. NFLUX (NRL Ocean Surface Flux System) is a tool that uses 2DVAR to blend a range of satellite observations of surface air temperature, moisture, wind speed, solar and longwave radiation with model forecasts to provide corrections to the model’s estimation of the surface heat fluxes. The forecast skill will be computed for these four experiments using independent data, and the model variability resulting from the different surface forcing will be analyzed.