The Weimer and IZMEM statistical convection models are driven with a time s
eries of interplanetary magnetic field (IMF) measurements made onboard the
Wind spacecraft. The model outputs are used to infer the ionospheric convec
tion velocity at Casey, Antarctica (80.8 degrees S geomagnetic latitude), a
nd then compared with measurements of Doppler velocity made using a Digison
de, and measurements of F-region convection implied by a collocated magneto
meter. During a single, representative campaign interval, 13-17 February 19
96, the Weimer model explained 19% (42%) of the variation in Doppler speed
(direction) observed by the Digisonde, and 21% (14%) of the equivalent conv
ection components observed by the magnetometer,This compares with IZMEM whi
ch explained 16% (46%) of the variation in Doppler speed (direction) observ
ed by the Digisonde, and 34% (32%) of the equivalent convection components
observed by the magnetometer. In general, there was better agreement betwee
n convection direction than convection speed. Some of the disagreement was
probably due to differences between the IMF measured by Wind located simila
r to 170 R-E upstream in the solar wind and the IMF actually arriving at th
e magnetopause. However, the results of this study do show that measurement
s of ionospheric velocity using different experimental techniques need heav
y averaging to identify a common component of velocity controlled by the IM
F vector. The present time series approach was also used to estimate 16 +/-
5 min as the time required for the ionospheric convection to reconfigure i
n response to IMF changes occurring at the magnetopause. (C) 1999 Elsevier
Science Ltd. All rights reserved.