A two-dimensional time-dependent tracer model was constructed to study the double-peak structure (equatorial minimum and subtropical maxima) observed in the distribution of nitrous oxide during the spring equinoctial condition. The meridional transport circulation was determined from computed diabatic heating rates, and the horizontal eddy diffusion coefficients were computed as residuals by potential vorticity conservation, using observed mean winds to compute the budget of zonal mean potential vorticity. This model was used to predict the 30-day evolution of the observed nitrous oxide field. The integrated results were generally in good agreement with observed tracer fields up to the midstratosphere. However, the agreement between the model calculations and the observed mixing ratios was not as good in the equatorial upper stratosphere at the equinoxes. Stronger downward motion compared to motion derived from diabatic heating rates based on climatological temperature and ozone fields is required in the equatorial region in order to simulate the observed double-peak structure. To intensify the equinoctial subsidence, the semiannual oscillation in the zonal mean wind and temperature fields were enhanced to double the climatological values in the equatorial region. The diabatic vertical velocity fields produced in this way significantly changed the resulting tracer patterns.

The double-peak structure produced by the model with the modified meridional circulation was much closer to the observed double peak than the one produced by the model based on the original diabatic circulation. A possible change of the ozone concentration associated with the specified temperature change was estimated. The effect of this ozone change on transport was to enhance the strength of the semiannual oscillation, although the magnitude was very small. This work indicates that standard climatologies may underestimate the strength of the equatorial semiannual oscillation. The diabatic circulation must be stronger than implied by these climatologies. Otherwise, it does not seem possible to account for the observed evolution of the nitrous oxide mixing ratio distribution. ¿ American Geophysical Union 1991