Reducing ocean model imbalances in the equatorial region caused by data assimilation

• Published in: Quarterly journal of the Royal Meteorological Society Vol. 143; no. 702; pp. 195 - 208
• Main Authors: Waters, J., Bell, M. J., Martin, M. J., Lea, D. J.
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01.01.2017; Wiley Subscription Services, Inc
• Subjects: Data assimilation; equatorial region; initialization shock; model imbalances; ocean; Oceanic analysis; Velocity
• ISSN: 0035-9009; 1477-870X
• DOI: 10.1002/qj.2912

The equatorial region is a particularly challenging area for ocean data assimilation. In this region the dominant balance near the surface is between the ocean pressure gradients and the applied wind stress. When increments are applied to an ocean model near the Equator, the pressure gradients are modified and this can cause an imbalance with the unchanged wind stress. This can lead to an initialization shock where spurious equatorial waves and vertical velocities are triggered. The equatorial waves can degrade the performance of the ocean model while the increased vertical velocities have a significant negative impact on biogeochemical models forced by these physical ocean fields. Previous studies have proposed a scheme to reduce the ocean biases associated with slowly varying errors in the applied wind stress by applying a pressure correction calculated from accumulated temperature and salinity increments. This scheme has previously been shown to reduce the mean vertical velocities and improve modelled circulation. However, when applying this scheme, we still see equatorial imbalances from data assimilation on shorter time‐scales. We consider a new method which is an extension of the pressure correction and aims to act as a balance to the equatorial increments and reduce intialization shock. We refer to this as an incremental pressure correction. We provide a description of this method and an analysis of its energetics. In initial experiments with a global ocean data assimilation system, this new method is shown to suppress the generation of equatorial waves in a single observation experiment and to significantly reduce the standard deviation of vertical velocities in a re‐analysis experiment.