Boundary mixing along the Northern Deep Water Gulf of Mexico

An anthropogenic tracer reported in Ledwell et al. (in preparation ) was released along an isopycnal at approximately 1100 water depth, 150 m above the local bottom on the continental slope of the Northern Deepwater Gulf. Sampling after 4 months returns an interior diapycnal diffusivity estimate of $1.3 \times10^{-4}$ m$^2$ s$^{-1}$. Sampling at one year returns a background interior mixing rate an order of magnitude smaller, $1.5 \pm 0.5 \times10^{-5}$ m$^2$ s$^{-1}$, and an inference that the 4 month estimate is dominated by a boundary process.

In this work we examine ancillary evidence collected during the year long field program that might explain the observed diapycnal dispersion. We find only background $(1 \times 10^{-5}$ m$^2$ s$^{-1})$ mixing levels in the interior Gulf of Mexico, diagnosed from the application of finescale parameterizations to LADCP/CTD data. If these interior estimates are representative of climatological mixing rates, it also implies the 4 month tracer dispersion estimate is dominated by a boundary mixing process. However, estimates of boundary mixing from the LADCP/CTD data at the time of the 12 month survey are not sufficient to explain the 4 month tracer dispersion.

Plausibility that the inferred boundary mixing could result from non-propagating form drag is explored with a 2-dimensional time dependent advection diffusion model. The model is used with the current meter data to establish the expectation that flow over complex topography (e.g. salt domes) situated on the continental slope provides a turbulent energy source sufficient to explain the tracer dispersion. Implied is an $O(1)$ effective drag coefficient.