A time-dependent, three-dimensional, multi-ion numerical model of the global ionosphere was used to study the asymmetry in large-scale ionospheric features between the northern and southern hemispheres. The comparisons were done for June and December solstice conditions at solar maximum for quiet geomagnetic activity. Simple conditions and diurnally reproducible ionospheric features were established in order to elucidate the intrinsic hemispherical differences that are associated with the different displacements between the geomagnetic and geographic poles and the different atmospheric conditions. In comparing the ionospheric densities in the northern and southern hemispheres for a given season, we found the following: (1) The winter hemispheres display the most marked universal time (UT) variations due to the displacement between the geomagnetic and geographic poles, (2) The summer high-latitude and equatorial densities in both hemispheres are morphologically similar.

However, there is a factor of 2 difference in Appleton anomaly densities in certain places between the June and December solstices due to the seasonal dependence of the atmospheric densities, (3) In the winter hemispheres, the corresponding electron densities are again morphologically similar but can be very different quantitatively due to the different dipole offsets, (4) For a given season, the density difference between the northern and southern hemispheres displays a marked UT dependence. For similar winter conditions, the densities in the northern and southern hemispheres can differ by an order of magnitude in places at certain UTs, (5) The winter mid-latitude trough is the feature that exhibits the largest northern-southern hemisphere difference due to the different dipole tilts, and (6) The ''winter anomaly'' is present in the northern hemisphere at almost all UTs, while it is essentially absent in the southern hemisphere. In the northern hemisphere, the anomaly maximizes between 1000 and 1600 UT, with the winter/summer peak electron density (NmF2) ratio reaching 1.6. In our model, the anomaly results primarily from the adopted atmospheric densities. ¿ American Geophysical Union 1989