Turbulence and the Detailed Structure of a Subtropical Jet Stream

Abstract

Jet stream wind data obtained during Project Topcat in August 1963 at Woomera, when rid of fortuitous detail by averaging serial soundings (about 12 at hourly intervals), show detail in the cross-aids flow which can be ascribed with fair certainity to sub-synoptic male eddy viscosity. This fact is employed to estimate stress, eddy viscosity and viscous dissipation over one mean profile taken in the jet aids. These are found to have values compatible with existing knowledge, and, with eddy stress in the strong shears beneath a jet stream of order 1 dyne cm?2, one can explain at least a high proportion of the upper branch of Krishnamurti's mean crow-axis flow. The variation of vertical velocity with height is qualitatively that inferred from Angell's and Krishnamurti's observations. An empirical linear relation found in mean data referring to strong shear layers, between static stability and the square of the vertical shear, when taken in conjunction with the equation expressing the budget of turbulent kinetic energy, suggests that significant turbulence is generated only when the shear exceeds a critical value (about 10?2 sec?1) and that the generation rate is proportional to the amount by which the square of the shear exceeds the square of this critical value. The ratio KM/KH is estimated to have a value of 1½?2. A thorough check of these indications is desirable. Aircraft observations of turbulence well downstream from Woomera showed a shallow layer of moderate clear air turbulence at a level higher by about 25 mb than that expected on the basis of the Woomera computations, but they do not conflict with the notion that the turbulence was created in the strong shear layer (Ri < 1) which is subject to mesoscale disturbances affecting shear and stability. A possible role for Ace waves, in the production of CAT as suggested by Scorer, is underlined by the profile of his l2 parameter, which shows a maximum at the level where CAT was found. Viscous dissipation computed from aircraft observations of energy spectra in the inertial subrange is compatible with the production rate of turbulent energy estimated here.