Abstract

Understanding how much mixing occurs in a stratified turbulent mixing event is a key question for understanding a variety of oceanic processes. However, directly measuring turbulent fluxes is difficult. Instead, parameterizations may be employed to estimate details of the mixing in terms of measurable quantities. One approach is to use characteristic lengthscales, such as the Thorpe and Ozmidov scales, to describe the state of the flow; relationships between such lengthscales can then in theory be used to estimate turbulent quantities of interest. While this approach has seen some success when applied to numerical simulations, there remain several challenges when it comes to using this approach in real observational datasets – for example, the underlying mechanism driving the mixing may not be known, or the measurements may only come from isolated profiles in the larger flow. In this work, we use a suite of direct numerical simulations of stratified shear instabilities as examples of both scouring and overturning mixing events. We subsample the data to mimic oceanographic measurements and compute lengthscales from the individual profiles. In doing so, we explore how well these profile-based estimates are able to characterize the state of the turbulent flow, particularly the mixing efficiency, and test how well fluxes can be computed from this limited data.