Search

The CEDAR storm study is a cooperative effort to study specific storm intervals with particularly good data coverage. The aim is to involve interested scientists with F-region/thermosphere data or modelling capabilities in coordinated efforts to understand the coupled upper atmosphere regions at different heights and on a global scale. The major storms studied to date occurred in March 1990, June 1991, November 1993, and May 1995. In this paper the objectives and organization of the CEDAR Storm Study are described, and results from several recent projects are reported. These include attempts at a global description of high latitude ion convection related to magnetosphere changes, Joule and particle heating, neutral winds, temperatures, and composition; plasmaspheric irregularity imaging; ionospheric model validation; the relationship between SAR arcs, subauroral ion drift (SAID) events, and the ionospheric trough; high latitude patches and blobs; ionospheric storm effects observed by radio and optical tomographic techniques; upper thermosphere neutral winds at high and middle latitudes; and effects of energy transfer into the lower thermosphere and mesosphere on neutral winds.

Three types of radars have demonstrated, during the 1980's, the capibility of measuring winds with useful resolution (2–3 km in the vertical, sampling rates, 1–12 hr−1): medium frequency (MF) using the spaced antenna method, LF systems using similar techniques, and meteor radars, some associated with S.T. radars. The techniques used, assumptions and examples of winds data are shown for each major type of system. The MLT network is now active in the WITS/ CEDAR and STEP (1990–1995) programs, for the latter the project is called MLTCS ( M esosphere, L ower T hermosphere C oupling S tudy). It involves climatologies of mean winds, tides and planetary waves, interactions between these waves, and solar, magnetospheric influences. Some examples of these data are shown.

During the last decade a large number of radars (∼12) have been developed, which have produced substantial quantities of tidally-corrected mean winds data. The distribution of the radars is not global, but many areas are well covered: the Americas with Poker Flat (65°N), Saskatoon (52°N), Durham (43°N), Atlanta (34°N), Puerto Rico (18°N); Europe with Kiruna (68°), Garchy (47°N) and Monpazier (44°N); and Oceania with Christchurch (44°S), Adelaide (35°S), Townsville (20°S), and Kyoto (35°N). Zonal and meridional wind height-time cross-sections from 60 80 km (MF/Meteor Radar) to ∼110 km have been prepared for the last 5–6 years. They are compared with cross-sections from CIRA-72 for zonal winds, and Groves (1969) for meridional winds. It is shown that while CIRA-72 is still a useful model for many purposes, significant differences exist between it and the new radar data. The latter demonstrate important seasonal, latitudinal, longitudinal and hemispheric variations. The new meridional cross-sections are of great value. The common features with Groves (1969) are the equatorward cells in summer near 85 km; however their strength (∼10 ms −1 ) and size are less. Systematic and somewhat different variations emerge at higher (⩾52°N) and middle (35–44°) latitudes.

A substantial quantity of wind data have been assembled from radar systems since CIRA-72 was formed: most of these radars include height ranging, and operate on a regular and even continuous basis. Systems include meteor and MF (medium frequency) Radars: an MST (mesosphere-stratosphere-troposphere) Radar (meteor mode); and an LF (low frequency) drift system. Latitudes represented are near 20° N/S, 35° N/S, 45° N/S, 50°N, 65° N/S. In all cases tidal oscillations were calculated so that corrected mean winds (zonal, meridional) are available - the meridional was not included in CIRA-72. Means for groups of years near 1980 are available, as well as individual recent years (1983, 1984) to allow assessment of secular trends: revised and improved analysis has been completed for several stations. Height-time cross-sections have been formed for each observatory: heights are typically ∼75–110 km, with time resolution of 7–30 days. Such detailed cross-sections were almost unknown before 1972. Comparisons with CIRA-72 are shown, and these emphasize the differences between hemispheres (NH, SH) in the radar winds. Other new winds from rockets and satellite radiances are contrasted with the radar set. There are important differences with the satellite-derived geostrophic winds (1973–78): possible explanations involve secular trends, longitudinal variations, and ageostrophy.