Spatial distributions of scalar mixing in stratified turbulence

We consider a series of direct numerical simulations of stratified turbulence at fixed buoyancy Reynolds number ($Re_b$), spanning a range of Froude ($Fr$) and Prandtl ($Pr$) numbers. We characterize each dataset by considering the spatial distributions of the dissipation rates of turbulent kinetic energy $\epsilon$ and scalar variance $\chi$, providing independent information about the power available to driving mixing and the rate of resulting scalar diffusion, respectively. Mixing rates are found to be highly variable throughout the domain and unexpected correlations between local properties of the turbulent and scalar fields are highlighted. Implications for the measurement and parameterization of turbulent thermal fluxes within the ocean are discussed.