The sensitivity of the total ozone distribution in our two-dimensional photochemical model to dynamical inputs has been explored. The residual circulation has been computed from three sets of heating rates but using the same temperature distributions. These heating rates results in advection fields that are appropriately linked with the magnitude of their vertical winds in the lower stratosphere and are called: (1) the strong circulation, (2) the weak circulation, and (3) the combined circulation. Three different formulations of horizontal eddy diffusion have also been investigated, including (1) a small constant diffusion, (2) a computed diffusion from potential vorticity using measured temperatures, and (3) a self-consistent diffusion determined from the meridional velocity. Finally, we studied two different formulations of vertical eddy diffusion: (1) a small constant diffusion and (2) an altitude-dependent larger diffusion. Our modeled global average total zone varies by up to 10%, depending on the various dynamical inputs, and the seasonal and latitudinal variabilities are even more substantial.

The calculations indicate that total ozone in the middle to high latitudes is very sensitive to the advection field below 100 mbar, the region where the heating rates (which are used in computing the advection field) are most uncertain. Modeled total ozone shows better agreement with solar backscattered ultraviolet (SBUV) ozone climatology, when computed horizontal eddy diffusion is used. Our ''best'' modeled total ozone distribution is a result of using the heating rates from Rosenfield et al. (1987) for pressures less than 100 mbar, along with heating rates from Dopplick (1974, 1979) for pressures greater than 100 mbar. Our model results, especially when comparing model and SBUV ozone between 10 and 1 mbar, along with heating rates from Dopplick (1974, 1979) for pressures greater than 100 mbar, indicate that there is an inconsistency between our ''best'' horizontal eddy diffusion and the ''best'' mean residual circulation. Our studies indicate a relative similarity of a typical ozone perturbation scenario among the different dynamical inputs. All of the perturbation studies predict the smallest ozone changes in the tropical latitudes, with larger ozone changes in the middle to high latitudes. The major changes in the various perturbation results are in the middle to high latitudes: (1) the peak ozone change is off of the pole in the southern hemisphere (near 60¿S latitude) using the strong circulation but centered on the pole for the various other dynamical studies, and (2) the maximum ozone change is centered at times varying from early spring to early summer, depending on which dynamical input is used. ¿ American Geophysical Union 1989