The Partition of Finescale Energy into Internal Waves and Subinertial Motions

Abstract

Finescale vertical wavenumber strain spectra (strain: normalized buoyancy frequency variability or vertical derivative of isopycnal displacement) are consistently less steep than shear spectra (shear: vertical derivative of horizontal velocity) for vertical wavelengths smaller than several tens of meters. Interpreting the diminished ratio of shear to strain (shear/strain ≡ horizontal kinetic/available potential energy) at higher vertical wavenumber as due to a greater contribution from high-frequency internal waves is not consistent with extant internal wave?wave interaction theories. A contribution from low-aspect-ratio subinertial density fine structure (flattened structures referred to herein as vortical modes) is therefore hypothesized. Vertical wavenumber spectra for vortical mode shear and strain are inferred from the observed spectra. Observed correlations between shear squared and buoyancy frequency squared exist that cannot be explained by either linear internal waves or geostrophic vortical modes. A model of internal wave?vortical mode interactions is used to interpret the observed correlations and partition the finescale spectra into internal wave and vortical mode components. A simple Doppler shift model is used with current meter data to refine the partitioning. The inferred vortical modes have an aspect ratio of approximately f/N (f: Coriolis frequency, N: buoyancy frequency), an rms velocity of 0.7 cm s?1, and bandwidth-limited gradient spectra. At vertical wavelengths larger than 30 m the vortical modes are inferred to be quasigeostrophic and in thermal wind balance. The data are interpreted as exhibiting an approximate equipartition between waves and vortical modes at vertical wavelengths smaller than 10 m.