The seasonal evolution of submesoscale turbulence statistics from an O(1 km)-resolving mooring array

Abstract:

The objective of the Ocean Surface Mixing, Ocean Submesoscale Interaction Study (OSMOSIS) is to develop parameterizations of processes that deepen and shoal the ocean surface boundary layer (OSBL). To meet this objective, a novel observational program was established to measure the evolution of the OSBL over an annual cycle (September 2012-September 2013). Observations include mooring-based measurements of temperature, salinity and velocity in the upper ocean (depths between 50 and 500 m), as well as continuous hydrographic measurements down to 1000-m depths from two ocean gliders within the mooring domain. The moorings were deployed in the North Atlantic over the Porcupine Abyssal Plain (48.69°N, 16.19°W), a region characterized by moderate mesoscale variability. Here, we present observations from five moorings spaced approximately 1.5 km apart.

Lateral gradients in velocity and buoyancy, coupled with seasonally varying stratification, depict a richness of structure characteristic of the submesoscale. Relative vorticity displays positive skewness during winter months (D-M) and zero skewness during summer months (J-N), with magnitudes reaching close to f. The former results indicate a preference for cyclonic motion during winter. Divergence and strain rate show less variation with season but divergence displays a slight negative bias, indicating a tendency for downward motion near the surface ocean. Lateral buoyancy gradients depict strong seasonal cycles, as well, with magnitudes during summer months exceeding those in winter months by a factor of 2 to 5. In all cases, standard deviations of the aforementioned quantities decrease with increasing depth. Observed statistics are supported by similar calculations from a seasonally varying, high-resolution (dx = dy = 500 m) numerical model. We close with estimates of potential vorticity, q, and discuss processes (e.g., frontogenesis, wind and buoyancy forcing) that may modify q and hence these turbulence statistics.