AUTHOREA
Log in Sign Up Browse Preprints

loading page

Tracer Versus Observationally-Derived Constraints on Ocean Mixing Parameters in an Adjoint-Based Data Assimilation Framework
  • +6
  • David Trossman,
  • Caitlin Whalen,
  • Thomas Haine,
  • Amy Waterhouse,
  • Arash Bigdeli,
  • An Nguyen,
  • Matthew Mazloff,
  • Patrick Heimbach,
  • Robin Kovach
David Trossman
University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences

Corresponding Author:david.s.trossman@gmail.com

Author Profile
Caitlin Whalen
University of Washington, Applied Physics Laboratory, University of Washington, Applied Physics Laboratory, University of Washington, Applied Physics Laboratory, University of Washington, Applied Physics Laboratory
Author Profile
Thomas Haine
Johns Hopkins University, Department of Earth and Planetary Sciences, Johns Hopkins University, Department of Earth and Planetary Sciences, Johns Hopkins University, Department of Earth and Planetary Sciences, Johns Hopkins University, Department of Earth and Planetary Sciences
Author Profile
Amy Waterhouse
Scripps Institution of Oceanography, Scripps Institution of Oceanography, Scripps Institution of Oceanography, Scripps Institution of Oceanography
Author Profile
Arash Bigdeli
University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences
Author Profile
An Nguyen
University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences
Author Profile
Matthew Mazloff
University of California San Diego Scripps Institution of Oceanography, University of California San Diego Scripps Institution of Oceanography, University of California San Diego Scripps Institution of Oceanography, University of California San Diego Scripps Institution of Oceanography
Author Profile
Patrick Heimbach
University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, Jackson School of Geosciences & Institute for Geophysics, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, Jackson School of Geosciences & Institute for Geophysics, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, Jackson School of Geosciences & Institute for Geophysics, University of Texas-Austin, Oden Institute for Computational Engineering and Sciences, Jackson School of Geosciences & Institute for Geophysics
Author Profile
Robin Kovach
SSAI, SSAI, SSAI
Author Profile

Abstract

This study investigates the possibility of using an ocean parameter and state estimation framework to improve knowledge of mixing parameters in the global ocean. Multiple sources of information about two ocean mixing parameters, the diapycnal diffusivity and the Redi coefficient, are considered. It is first established that diapycnal diffusivities derived from multiple observational data sets with a strain-based parameterization of finescale hydrographic structure can be used to ameliorate model biases in diapycnal diffusivities from the Estimating the Circulation & Climate of the Ocean (ECCO) framework and the GEOS-5 coupled Earth system model. The evidence is as follows. Adjusting ECCO-estimated and GEOS-5-calculated diapycnal diffusivity profiles toward profiles derived from Argo floats using the finescale parameterization improves agreement with independent diapycnal diffusivity profiles inferred from microstructure data. In addition, several aspects of the GEOS-5 model solution, such as mixed layer depths and temperature/salinity/stratification (i.e., hydrographic) fields, improve when the Argo-derived diapycnal diffusivities from the finescale parameterization are used instead of the model’s diapycnal diffusivities. The model’s hydrographic changes, which exceed the spread in hydrographic variables from an ensemble of simulations that use different initial conditions, occur due to the dynamic adjustment that arises when diapycnal diffusivity adjustments are applied. An adjoint sensitivity analysis with the ECCO framework suggests that the assimilation of biogeochemical tracers, such as dissolved oxygen concentrations, in future ECCO re-optimizations would improve estimates of the diapycnal diffusivity field.