Stratospheric Turbulence And Vertical Effective Diffusion Coefficients

The determination of winds and temperature by rawinsonde allows us to calculate the Richardson number, a measure of atmospheric instability. From this, we can then determine the turbulent atmospheric parameters, diffusivity (K), rate of heat deposition (epsilon), and the vertical turbulent intensity (w2). Comparison of these data, for the spring periods, with contours of residual stratospheric radioactivity demonstrate that, at the tropopause, the isocountours of turbulent diffusivity accurately follow the isodensity contours of the residual radioactivity. We also demonstrate that the rapid cessation of the turbulent diffusivity accurately predicts the level of the tropopause. (Author).

Vertical transport in the stratosphere occurs by many processes and the question of the relative importance of stratified turbulence is considered in the context of a 'vertical stack model' described in a previous report. The 'vertical stack model' was designed to estimate the effect of intermittent turbulent layers (separated by essentially laminar flow) and it is here examined in some detail. In particular, a number of less obvious implications are discussed and a list of experimental tests of the model is suggested. It is concluded that the rate of turbulence for vertical transport in the stratosphere remains an open question.

This paper reviews the estimates to date of the vertical 'effective diffusion coefficient' for stratospheric small scale turbulence transport, KB. These estimates range (in order of magnitude) from 1.0 sq. meter/s to 0.01 sq. meter/s, that is to say from a value which would make turbulence a dominant factor in stratospheric transport to a value which would make it totally insignificant. Such a large range implies much ignorance in this subject. The various techniques are closely examined and the unanswered experimental questions are exhibited. The conclusion is reached that more experimental work needs to be done before one has a reliable estimate for KB.

In order to estimate the rate of vertical transport due to turbulence in the stratosphere, it is necessary to know the degree of mixing that takes place in 'Kelvin-Helmholtz billow events.' This is estimated by means of a discrete model and the results are compared with published experimental observations. From these considerations it is concluded that a very large degree of mixing probably takes place in such events, and that, therefore, one must use the relation between the bulk vertical eddy diffusivity and layer diffusivity originally proposed by Rosenberg and Dewan instead of the simple one usually employed.