Improvement of the Ocean Mixed Layer Model via Large-Eddy Simulation and Inverse Estimation

Abstract The ocean mixed layer model (OMLM) is improved using the large-eddy simulation (LES) and inverse estimation method. A comparison of OMLM (Noh model) LES results reveals that underestimation turbulent kinetic energy (TKE) flux in causes a negative bias depth (MLD) during convection, when wind stress weak or latitude high. It further found entrainment thickness underestimated. effects alternative approaches parameterizations OMLM, such as nonlocal mixing, length scales, Prandtl number, TKE flux, are examined with an aim to reduce bias. Simultaneous optimizations empirical constants various versions Noh different parameterization options then carried out via iterative Green’s function approach data constraining data. An obtained, which reflects new features, including enhanced improve performance all cases, namely, wind-mixing, surface heating, cooling cases. effect grid resolution on optimal also investigated. Significance Statement This work illustrates novel vertical mixing upper ocean, plays important role climate models. utilizes from realistic turbulence simulation, called proxy observation for analyze parameterization, statistical method, estimation, obtain optimized used parameterization. same can be applied other