MAGNETIC ACTIVITY DEPENDENCE OF HIGH-LATITUDE THERMOSPHERIC WINDS ANDDENSITIES BELOW 200-KM

Satellite-based measurements are utilized to elucidate the latitudinal , local time, and magnetic activity dependence of winds and densities in the scantily observed atmospheric region between 170 and 220 km abo ve 45-degrees magnetic latitude. One data set consists of atmospheric densities from high-accuracy (time resolution almost-equal-to 5-6 hour s) orbital analyses of three Doppler Beacon satellites in orbit during 1973. The perigees of these satellites are generally restricted to 16 0-180 km, 1200-1400 LST, and geographic latitudes greater than about 4 5-degrees. Statistical relationships are derived between the density c hanges and the planetary magnetic index, K(p), and the 5-hour mean of the auroral electrojet index, AEBAR. The former relationships are comp ared with those derived from the MSISE90 empirical model (Hedin, 1991) , which is found to overestimate the rate of increase of density with respect to K(p). Another data set consists of densities and cross-trac k winds from the Satellite Electrostatic Triaxial Accelerometer (SETA) experiment for the March 21 to April 9, 1979, period, which includes several intervals of elevated magnetic activity. Besides comparing var ious time series, the data are also binned according to 10-degrees lat itude increments and unit increments of K(p) to derive trends. Some ty pical results include the following, corresponding to average changes in the 45-degrees to 65-degrees magnetic latitude band as K(p) is incr eased from 1 to 6: (1) for the nightside (almost-equal-to 2230 LT), a change in cross-track (nearly zonal) wind from 25 +/- 25 m s-1 (eastwa rd) to -125 +/- 25 m s-1 (westward), and an increase of about 20% in d ensity; and (2) for dayside (almost-equal-to 1030 LT), a change in cro ss-track wind from 25 +/- 25 m s-1 (eastward) to 125 +/- 25 m s-1 (eas tward), and a density increase of 25%. For some individual sudden enha ncements in magnetic activity, changes in winds and densities can be m ore than double the above average values. Comparisons are also made wi th the NCAR TIGCM (National Center for Atmospheric Research thermosphe re-ionosphere general circulation model) simulation for the complete 2 0-day interval, and with recent empirical models of densities (Hedin, 1991) and winds (Hedin et al., 1991), with data points in each case de rived for the satellite paths, instrument orientations, and sampling r ates identical to that of the SETA experiment.