In this study, the relation between the interior and the surface dynamics for nonlinear baroclinically unstable flows is examined using the concepts of potential vorticity. First, it is demonstrated that baroclinic unstable flows present the property that the potential vorticity mesoscale and submesoscale anomalies in the ocean interior are strongly correlated to the surface density anomalies. Then, using the invertibility of potential vorticity, the dynamics are decomposed in terms of a solution forced by the three-dimensional (3D) potential vorticity and a solution forced by the surface boundary condition in density. It is found that, in the upper oceanic layers, the balanced flow induced only by potential vorticity is strongly anticorrelated with that induced only by the surface density with a dominance of the latter. The major consequence is that the 3D balanced motions can be determined from only the surface density and the characteristics of the basin-scale stratification by solving an elliptic equation. These properties allow for the possibility to reconstruct the 3D balanced velocity field of the upper layers from just the knowledge of the surface density by using a simpler model, that is, an "effective" surface quasigeostrophic model. All these results are validated through the examination of a primitive equation simulation reproducing the dynamics of the Antarctic Circumpolar Current.

The motivation to study the vertical exchanges of tracers associated with mesoscale eddies is that the mean concentration of most oceanic tracers changes rapidly with depth. Because mesoscale processes may transport these tracers vertically, biogeochemists hypothesized that these vertical exchanges may strongly affect global tracer budgets. This hypothesis has motivated a large number of biogeochemical studies that we review here by focusing on the significant advances that have been achieved and the remaining issues and uncertainties. The main question that emerges concerns the importance of the submesoscales (10 km in the horizontal) in these vertical exchanges. Independently, in the past decade, fluid dynamicists examined the three-dimensional properties of submesoscales generated by a mesoscale (100 km in the horizontal) turbulent eddy field. We review their results and discuss how the vertical exchanges associated with these submesoscales may answer the issues raised by biogeochemical studies and inspire future directions.

Ocean eddies (with a size of 100–300 km), ubiquitous in satellite observations, are known to represent about 80% of the total ocean kinetic energy. Recent studies have pointed out the unexpected role of smaller oceanic structures (with 1–50 km scales) in generating and sustaining these eddies. The interpretation proposed so far invokes the internal instability resulting from the large-scale interaction between upper and interior oceanic layers. Here we show, using a new high-resolution simulation of the realistic North Pacific Ocean, that ocean eddies are instead sustained by a different process that involves small-scale mixed-layer instabilities set up by large-scale atmospheric forcing in winter. This leads to a seasonal evolution of the eddy kinetic energy in a very large part of this ocean, with an amplitude varying by a factor almost equal to 2. Perspectives in terms of the impacts on climate dynamics and future satellite observational systems are briefly discussed.