The Impact of Observed Variations in the Shear-to-Strain Ratio of Internal Waves on Inferred Turbulent Diffusivities

• Published in: Journal of physical oceanography Vol. 46; no. 11; pp. 3299 - 3320
• Main Authors: Chinn, Brian S, Girton, James B, Alford, Matthew H
• Format: Journal Article
• Language: English
• Published: Boston American Meteorological Society 01.11.2016
• Subjects: Categories; Correlation; Diffusion coefficients; Diffusivity; Internal waves; Laboratories; Marine; Mooring; Ocean circulation; Oceanography; Profiles; Shear; Strain; Studies; Temporal variability; Temporal variations; Time series; Topography; Turbulence; Variability; Velocity; Vertical profiles; Wind variability; United States > US; California
• ISSN: 0022-3670; 1520-0485
• DOI: 10.1175/JPO-D-15-0161.1

Abstract The most comprehensive studies of the spatial and temporal scales of diffusivity rely on internal wave parameterizations that require knowledge of finescale shear and strain. Studies lacking either shear or strain measurements have to assume a constant ratio between shear and strain ( R ω ). Data from 14 moorings collected during five field programs are examined to determine the spatial and temporal patterns in R ω and the influence of these patterns on parameterized diffusivity. Time-mean R ω ranges from 1 to 10, with changes of order 10 observed over a broad range of scales. Temporal variability in R ω is observed at daily, weekly, and monthly scales. Observed changes in R ω could produce a 2–3 times change in parameterized diffusivity. Vertical profiles of R ω , E shear , and E strain (shear or strain variance relative to Garret–Munk) reveal that both local topographic properties and wind variability impact the internal wave field. Time series of R ω from each mooring have strong correlations to either shear or strain, often only at a specific range of vertical wavenumbers. Sites fall into two categories, in which R ω variability is dominated by either shear or strain. Linear fits to the dominant property (i.e., shear or strain) can be used to estimate a time series of R ω that has an RMS error that is 30% less than the RMS error from assuming R ω = 3. Shear and strain level vary in concert, as predicted by the Garret–Munk model, at high E shear values. However, at E shear < 5, strain variations are 3 times weaker than shear.