Data obtained by Sondre Stromfjord incoherent scatter radar during rec
ent Lower Thermospheric Coupling Study (LTCS) campaigns have been anal
yzed to examine seasonal changes in high-latitude lower thermosphere n
eutral winds. Data included in this study were obtained with a 2x160 m
u s multipulse as well as with higher-resolution multipulse and altern
ating code techniques and were analyzed with different auto correlatio
n function (ACF) fitting routines. We report neutral wind results obta
ined during LTCS experiments with the new acquisition and analysis sys
tem and compare these results with those acquired using earlier versio
ns of the software. We investigate the effects of different assumption
s in the ACF fitting program on the data return and quality. Wind resu
lts are intercompared to examine average seasonal structure and variab
ility. Strong semidiurnal oscillations characterize high-latitude lowe
r thermospheric winds at 105 km irrespective of season; however, summe
r amplitudes are nearly a factor of 2 greater than those observed duri
ng winter. Average winter winds appear to have more variability than t
hose deduced for summer. Equinox winds at fall and spring show similar
oscillations but are nearly reversed in phase. At 115 km, diurnal osc
illations grow and their amplitudes often exceeds typical semidiurnal
amplitudes. Tidal amplitudes observed in radar data are compared with
the predictions of several models, including National Center for Atmos
pheric Research thermosphere-ionosphere-general-circulation model (TIG
CM) [Roble et al., 1988], global scale wave model (GSWM) [Hagan et al.
, 1994], and Forbes-Vial model (FV91) [Forbes and Vial, 1991]. These c
omparisons suggest that theoretical predictions significantly underest
imate tidal amplitudes calculated from Sondrestrom incoherent scatter
radar measurements.