A multi-parameter parameterization of ocean diapycnal mixing efficiency: global estimates inferred from Argo-float based profiles

We propose a dynamic parameterization for irreversible mixing efficiency (E) based on an extensive Direct Numerical Simulation (DNS) database of inhomogeneously stratified shear-induced turbulence. The Osborn formula for estimating diapycnal diffusivity K_ρ with its canonical flux coefficient of 0.2 is replaced by a generalized-Osborn formula with a flux coefficient that varies non-monotonically with both the buoyancy Reynolds number, Re_b =ε/(νN²) (where ν is the kinematic viscosity and ε and N² denote turbulent dissipation and density stratification respectively) and the gradient Richardson number, Ri = N²/S² where S denotes the vertical shear. Using estimates of Re_b inferred from the global array of Argo floats, we analyze the global maps of Γ = E /(1−E ) and discuss their important spatial variability and Ri-dependence. Moreover we elaborate on the depth dependent global averages of K_ρ .