High‐resolution ensemble data assimilation and forecasting of mesoscale and submesoscale circulation in the Arabian Gulf

Abstract High‐resolution ocean models are increasingly capable of resolving energetic mesoscale and submesoscale dynamics, but integrating data assimilation (DA) into such systems remains challenging. Static background error covariances can limit the effectiveness of assimilating observations, particularly in shallow and semi‐enclosed basins, such as the Arabian Gulf, characterized by rapidly varying circulation. To address these challenges, we develop a high‐resolution (1 km) regional ensemble DA system, the first of its kind for the Gulf, based on the Massachusetts Institute of Technology General Circulation Model and the Data Assimilation Research Testbed. The system incorporates both standard and hybrid ensemble adjustment Kalman filter formulations to assimilate satellite‐derived sea‐surface temperature and sea‐surface height (SSH), as well as in‐situ temperature and salinity profiles. Assimilation improves surface and subsurface fields, with the hybrid ensemble adjustment Kalman filter yielding more spatially coherent analyses, reduced sea‐surface temperature and SSH forecast misfits relative to observations, and improved overturning circulation relative to standalone simulations and global reanalyses. Independent validation using SSH from the Surface Water and Ocean Topography mission during its 2023 fast‐sampling phase shows improved spectral characteristics and enhanced high‐frequency mesoscale and submesoscale variability, underscoring the benefit of ensemble‐based DA with high‐resolution modelling in small, shallow basins such as the Arabian Gulf.