With ionospheric weather defined as the hour-to-hour, day-to-day and week-to-week variability, a complete understanding requires an organization of results within a framework of seasonal and solar cycle climatologies. There is also the need to organize studies of variability with reference to a generalized set of prevailing climatological conditions which might be identified as ``quiet,'' ``moderately active,'' and ``highly disturbed.'' In general, however, approaches to studies of variability have tended to focus on transitions from ``non-disturbed'' periods to ``highly disturbed'' periods as manifested during geomagnetic storms. This represents the preponderance of efforts involving ionospheric weather, and it is the case reported here.
During and following a geomagnetic storm, ionospheric responses can appear chaotic because of the coupled intensity and commencement-time effects of the disturbance. To understand the cause-effect relationships Fuller-Rowell et al. [1994] performed four storm simulations using the UCL (University College of London) coupled ionospheric-thermospheric model under equinox conditions, with each storm 12 hours long and commencing separately at 0600, 1200, 1800, or 2400 UT. The results were used to extract the basic physical processes in a simple descriptive way. Features included a Joule-heat-driven wind surge that penetrated the opposite hemisphere resulting in poleward winds. Results also showed that equatorward winds in sunlight produce positive ionospheric changes during the main phase and that increases in molecular nitrogen in regions of sunlight caused negative ionospheric changes. The strength of the increases were found to depend on the longitude and the local time of the sector during the storm onset.
In another study of ionospheric storms, Codrescu et al. [1992]
compared time-dependent simulations of the NCAR/TIGCM with a distribution
of eight northern latitude ionosonde stations at East Asian
longitudes during the period of March 19-28, 1979. The measured
parameters included the critical frequency f
F
and the height
of the 
-region along with meridional winds deduced
from an analysis of 
. The agreement between the data and
the model specification of 
was very good poleward of
45
magnetic latitude, but decreased equatorward. The agreement
for 
however was good throughout all comparisons. The
model seemed to ``overreact'' in calculating the amplitude and
latitudinal penetration of the perturbation, suggesting that
the parameterized auroral forcings were too large. Additional
analysis showed that the model-measurement comparisons were in
good agreement for negative storm effects above 40
where meridional
winds and composition were believed to be realistically modeled.
Agreement decreased with decreasing latitude, presumably because of
the increased influence of dynamo electric fields and interhemispheric
flux exchanges which are not self-consistently modeled in the TIGCM.
Yeh et al. [1994] provided more extensive measurements of storm-time
effects in a study of the global behavior of ionospheric responses to
the great magnetic storm of October 1989. They found that
ionospheric responses at the various observing stations differed
appreciably with a dependence on the local time of storm commencement.
They also found long lasting and large-scale changes manifested in
severe depressions of 
at higher latitudes, in the
suppression of the equatorial anomaly, and in the development of
large horizontal gradients at certain latitudes.
In more regional studies, Buonsanto et al. [1992a] developed a
Millstone Hill radar dataset detailing storm-induced ionospheric
effects over a 35
latitude span during two geomagnetic
disturbances in March and April of 1990. They observed deep
nighttime density troughs above and equatorward of Millstone Hill, as
well as extremely low daytime densities with h
in the
molecular-ion dominated region below 200 km. In a related work,
Buonsanto and Foster [1993] used Millstone Hill and Arecibo radar data
to illustrate the importance of both magnetospheric electric
field penetration and disturbance neutral winds on the
low-to-middle latitude ionosphere during the magnetic storm of March
20-21, 1990.