Fossil turbulence and internal waves

Publisher Website
Google Scholar
Abstract
Measurements of small scale temperature and velocity fluctuations in the ocean show that in most layers the turbulence occurs in isolated patches occupying a small fraction of the fluid volume. Most of the temperature microstructure patches observed are partially mixed remnants of previous overturning turbulence events after the turbulence has been damped out. The fluctuations are often partially restratified at the larger scales of the original turbulence. The smaller scales are more isotropic. An equilibrium microscale is preserved by the laminar straining motions of either the ambient internal waves, remnant internal waves produced when the turbulence was damped, or by laminar restratification. Such remnant waves and remnant temperature fluctuations are referred to as fossil turbulence. Fossil turbulence can be distinguished from active turbulence because active turbulence can only exist when inertial forces of the eddy motions are larger than either buoyancy or viscous forces: this requires 1.2 LR?λ?15 LK, where λ is the wavelength of possible turbulent motions. In the ocean the buoyancy scale 1.2 LR is often found to be less than the viscous scale 15 LK: this implies no active turbulence at any scale in the microstructure. Kinetic energy of the turbulence which produced the microstructure is converted to internal wave energy. A spectral description of the process is presented, and the role of turbulence as a source of internal waves is discussed. For available oceanic data, the buoyancy scale 1.2 LR is always found to be less than the original overturning scale 1.2 LR0 = 3.5 (D C0/N)1/2, which implies that the microstructure of active turbulence unaffected by buoyancy (o‐subscripts) has not been observed. A turbulence activity parameter AT is used to classify the available microstructure observations according to hydrodynamic state: all the data indicates various degrees of fossilization, with restratification anisotropy associated with the most advanced stages. Estimates of space‐time average velocity and temperature dissipation rates in the ocean from microstructure measurement are complicated by the extreme patchiness in space and intermittency in time of the dissipation events. Based on recent evidence of the degree of horizontal patchiness, underestimates by several orders of magnitude are probable if short data records are used. Such undersampling errors may account for the large discrepancy which exists between towed body and dropsonde estimates of mean dissipation rates in the core layer of the equatorial undercurrents and a similar discrepancy between dropsonde estimates of vertical diffusivity compared to canonical values inferred from mean properties in the main thermocline. Some corrections for time intermittency may be possible using information of previous dissipation activity preserved by the fossil turbulence.

This publication has 0 references indexed in Scilit: