Heat and salt budgets over the Gulf Stream North Wall during LatMix survey in winter 2012.

Alejandra Sanchez-Rios, Oregon State University, Corvallis, OR, United States, Robert Kipp Shearman, Oregon State University, College of Earth, Ocean, and Atmospheric Sciences, Corvallis, OR, United States, Eric A D'Asaro, Applied Physics Laboratory University of Washington, Seattle, WA, United States, Craig Lee, Organization Not Listed, Washington, DC, United States, Jonathan Gula, University of Brest, Laboratoire d’Océanographie Physique et Spatiale, Brest, France and Jody M Klymak, University of Victoria, Victoria, BC, Canada
Abstract:
As part of the ONR-sponsored LatMix Experiment, ship-based and glider-based observations following a Lagrangian float are used to examine the evolution of temperature, salinity and density along the Gulf Stream north wall in wintertime. Satellite observations during the survey and the in-situ measurements showed the presence of submesoscale (<10 km) features along the front. Models have successfully reproduced similar features, but observations are lacking, particularly at the small scales needed to understand their role in the transport of heat and salt across the front and out of the mixed layer. Calculating the trend in time at each depth and cross-front location we found an increase of heat and salinity in regions where the strongest cross-front gradients of velocity were observed at the mixed layer and around 150m depth, these changes are density compensated and suggest isopycnal mixing and a connection between the mixed layer and subsurface layers. The large Rossby number (Ro>1) calculated for this regions corroborates the possibility of submesoscale dynamics. Using a heat and salinity budget, we show that surface forcing, entrainment from below and advection by the mean flow velocities are not sufficient to explain the observed rate of change of heat and salinity in the mixed layer. Although confidence estimates prevent an accurate flux divergence calculation, Reynold flux estimates are consistent with a cross-frontal exchange that can reproduce the observed temporal trends.