Hunt, B. G., and S. Manabe, 1968: Experiments with a stratospheric general circulation model: II. Large-scale diffusion of tracers in the stratosphere. Monthly Weather Review, 96 (8), 503-539.

Abstract: The 18-level primitive equation, general circulation model described in Part I was usted to study the diffusion of two idealized tracers in the stratosphere. One tracer was designed to simulate broadly the behavior of the radioactive tungsten which escaped into the stratosphere following nulcear tests in the Tropics, the other was taken as a photochemical ozone distribution. Both the meridional circulation and the large-scale eddies were of primary importance for the polewards transport of the tracers in middle and high latitudes, but the supply of tracer for these eddies was principally maintained from the higher levels by the downward branches of the meridional circulation. Two meridional cells were found to occur in the stratosphere, a tropical direct cell and a higher latitude indirect cell, and these provided a natural explanation for many of the observed features of the tracer distributions in the actual atmosphere. The only major tropospheric-stratospheric exchange took place in the subtropics through the tropopause gap, the vertical eddies and the meridional circulation being of comparable magnitude for this exchange.
The synoptic situation in the atmosphere was found to be of fundamental importance for the large-scale diffusion of the tracers in middle latitudes, and the downgradient transport of tracers in the lower stratosphere was primarily accomplished by the upper level troughs of the planetary scale wave system.
Although the model used in this investigation was based on radiative conditions corresponding to annual mean insolation it appeared to be representative of winter conditions, and was in agreement with many observational features.
Schematic diagrams illustrating the principal features of the large-scale diffusion of the two tracers are given in figures 12 and 24.