Estimating the interannual variability of vertical velocity within the global ocean thermocline from observation-based geostrophic meridional velocities.

Vertical velocities are several orders of magnitude smaller than the horizontal ones when looking at patterns larger than the sub-mesoscales in the open ocean. Hence, direct measurement attempts of open ocean w are scarce. Methods for estimating w in the real ocean combine theory and observation-based fields. In the present work, climatological circulation patterns in Linear Vorticity Balance (LVB: βv=f∂w/∂z) are first identified in an eddy-permitting OGCM. Then, for such regime of circulation, we show that it is possible to reconstruct a robust w field for the climatological mean.

In the first part, we present a thorough baroclinic analysis of the climatological LVB. Below the Mixed Layer, the LVB holds to first order in the tropical and subtropical gyres interior and part of subpolar and austral circulation throughout the water column. Within western boundary currents, the equatorial band, areas of the subpolar gyres and the Circumpolar Circulation, significant departures occur due to the dominance of other terms in the vorticity budget, such as nonlinearities or friction. Although the ocean transport adjustment occurs on time scales constrained by basin-crossing times of Rossby waves, we show that the LVB often holds at much shorter time scales of a few years. When the climatology is reduced, the LVB's strength to describe the ocean circulation is relatively maintained. However, the time-dependent of the vorticity balance becomes significant and impacts the vorticity balance in western boundary currents and western tropical regions.

These results allow us to reconstruct the interannual variability of w for flows in LVB using geostrophic meridional velocities and satellite wind fields within large fractions of the global ocean. In the last part, we explore the differences at regional scale between our observation-based reconstruction and two other available estimates of w: one produced by an ocean reanalysis and the other reconstructed with observations and the Omega equation theory.