Uncertainty quantification and optimization of an eddy diffusivity, mass flux model for turbulent vertical mixing in the ocean surface boundary layer

Gregory LeClaire Wagner1, Andre Souza2 and Raffaele M Ferrari1, (1)MIT, Cambridge, United States, (2)Massachusetts Institute of Technology, Cambridge, United States
Abstract:
We report on the development of a new parameterization of turbulent vertical mixing in the ocean surface boundary layer for the "Climate Machine" (CliMa), a new climate model being developed by the Climate Modeling Alliance. Our 1.5-order, eddy-diffusivity, mass-flux scheme combines a prognostic turbulent kinetic energy equation and and mixing-length-based eddy diffusivity closure with a model for tracer transport by non-local convective plumes. We use a suite of large eddy simulations that spans a range of realistic physical scenarios to optimize and estimate the uncertainty of the closure’s free parameters, and compare parameter uncertainty and model error to the K-Profile-Parameterization optimized against the same numerical results.