Fabry-Perot interferometric (FPI) measurements of thermospheric zonal neutr
al winds at Arequipa, Peru (16.7 degreesS, 71.5 degreesW, -2.7 degrees dip)
, and Carmen Alto, Chile (23.1 degreesS, 69.4 degreesW, -10.2 degrees dip),
were collected during the solar minimum periods of September-October 1996
and 1997. The data set included 39 nights from Arequipa and 14 nights for C
armen Alto, with 8 nights of simultaneous observations. Analysis of average
d results found the peak evening zonal neutral wind speed of similar to 127
+/- 15 m/s eastward for the Arequipa observatory, which is located near th
e magnetic equator, to occur between 21:30 and 22:30 LT. In contrast, the p
eak evening zonal winds of similar to 100 +/- 10 m/s eastward observed from
Carmen Alto, which is located near the crest of the equatorial ionization
anomaly (EIA), occurred similar to0.5-1 h later. These measurements represe
nt the first case of groundbased FPI observations of the so called equatori
al temperature and wind anomaly (ETWA) over such a small latitude range in
the same longitude sector. This reduction in speed of similar to 20-25% at
Carmen Alto relative to Arequipa is attributed to increased ion drag at Car
men Alto caused by the higher electron density within the EIA region at alt
itudes of 220-300 km. Model studies were conducted using electron density a
nd neutral atmosphere parameters form the parameterized ionospheric model (
PIM) and the mass spectrometer incoherent scatter (MSIS) models, respective
ly, to calculate the ratio of ion-neutral collision frequencies at the two
sites. We found that the increase in electron density within the EIA was su
fficient to account for the observed reduction in the zonal wind. Thus, thi
s analysis confirms the dominant role of ion drag in modulating thermospher
ic dynamics at equatorial latitudes. A comparison of the FPI results with t
he predictions by two current neutral wind models, the Horizontal Wind Mode
l-90 and the NCAR Thermospheric Ionosphere Electrodynamics General Circulat
ion Model (TIEGCM), reveals that neither is able to reproduce accurately th
e latitude dependence reported here. Model refinements for electrodynamics
and improved resolution are suggested. (C) 2001 Elsevier Science Ltd. All r
ights reserved.