Modifications to the K-Profile parameterization with nondiffusive fluxes for Langmuir turbulence

View More View Less
  • 1 University of California, Los Angeles
© Get Permissions
Restricted access

Abstract

The K-profile parameterization (KPP) is a common method to model turbulent fluxes in regional and global oceanic models. Many versions of KPP exist in the oceanic sciences community and one of their main differences is how they take the effects of nonbreaking waves into account. Although there is qualitative consensus that nonbreaking waves enhance vertical mixing due to the ensuing Langmuir circulations, there is no consensus on the quantitative aspects and modeling approach. In this paper we use a recently-developed method to estimate both components of KPP (the diffusive term, usually called local, and the nondiffusive component, usually called nonlocal) based on numerically-simulated turbulent fluxes without any a priori assumptions about their scaling or their shape. Through this method we show that the cubic shape usually used in KPP is not optimal for wavy situation and propose new ones. Furthermore we show that the formulation for the nondiffusive fluxes, which currently only depend on the presence of surface buoyancy fluxes, should also take wave effects into account. Finally, we investigate how the application of these changes to KPP improves the representation of turbulent fluxes in a diagnostic approach when compared to previous models.

Corresponding author: Marcelo Chamecki: chamecki@ucla.edu

Abstract

The K-profile parameterization (KPP) is a common method to model turbulent fluxes in regional and global oceanic models. Many versions of KPP exist in the oceanic sciences community and one of their main differences is how they take the effects of nonbreaking waves into account. Although there is qualitative consensus that nonbreaking waves enhance vertical mixing due to the ensuing Langmuir circulations, there is no consensus on the quantitative aspects and modeling approach. In this paper we use a recently-developed method to estimate both components of KPP (the diffusive term, usually called local, and the nondiffusive component, usually called nonlocal) based on numerically-simulated turbulent fluxes without any a priori assumptions about their scaling or their shape. Through this method we show that the cubic shape usually used in KPP is not optimal for wavy situation and propose new ones. Furthermore we show that the formulation for the nondiffusive fluxes, which currently only depend on the presence of surface buoyancy fluxes, should also take wave effects into account. Finally, we investigate how the application of these changes to KPP improves the representation of turbulent fluxes in a diagnostic approach when compared to previous models.

Corresponding author: Marcelo Chamecki: chamecki@ucla.edu
Save