Ocean drifter velocity data assimilation Part 2: Forecast validation.

The highlights for the manuscript, "Ocean Drifter Velocity Data Assimilation, Part 2: Forecast Validation", include: • 3DVAR velocity data assimilation within an operational prediction system. • Eularian velocities inferred from surface drifter position observations. • Velocity assimilation reduces temperature and salinity forecast error. • Including velocity observations improves the position of fronts and eddies. A large deployment of drifters conducted during August-December, 2020 in the Gulf of Mexico offers a test bed for a data assimilation system developed specifically to include velocity observations. This updated Navy Coupled Ocean Data Assimilation system employs the three-dimensional variational approach and is described in part one of this two-part paper (Helber et al, 2023). In this paper, we examine the impact of velocity data assimilation on the ensuing forecasts of the ocean state including not only velocity but also temperature and salinity fields below the surface. Two high-resolution (1 km) experiments were performed in the Gulf of Mexico; one with velocity data assimilation and the other without. The resulting 48 h forecasts of temperature, salinity, and velocity are examined and compared relative to the observations being assimilated (including the inferred velocities from the drifters) and unassimilated observations of temperature, salinity, and velocity from two gliders near the drifters. In addition, we assess eddy positioning and Lagrangian trajectory separation. Comparisons of these two experiments, with and without velocity data assimilation, suggest that adding velocity observations to the assimilation increases skill in predicting velocity and the subsurface temperature and salinity. [ABSTRACT FROM AUTHOR]