Estimating turbulent lengthscales in stratified mixing events from limited measurements

Stratified turbulent mixing events play an important role in oceanic flows. However, direct measurement of their associated fluxes is difficult, necessitating the use of turbulent mixing parameterizations. One approach in parameterizing mixing has been to calculate various lengthscales characterizing the state of the stratified turbulent flow, and then to use relationships between these scales to infer the relevant fluxes — an approach which has seen some success when applied to numerical simulations and some observational datasets. However, when dealing with observational data, several additional challenges arise, including not knowing the specific mechanism driving the turbulent mixing and having limited data throughout the flow domain (for instance, isolated profiles). In this work, we subsample direct numerical simulations of both scouring and overturning stratified turbulent mixing events to mimic oceanographic measurements and compute lengthscales from the limited data. We explore how well profile-based estimates characterize the state of the flow and quantify the impact of limited data on the computed fluxes.