Key points are not available for this paper at this time.
An unusually long data set was acquired at the sodium lidar facility at Colorado State University (41N, 105W), between Sep 18 and Oct 01, 2003, including a 9‐day continuous observation. This time is long enough to average out the perturbations of gravity waves and short‐period planetary waves. As such, it can be used to define tidal‐period perturbations in temperature and horizontal wind. Assuming the sodium mixing ratio is a constant of motion, the observed tidal‐period oscillation in sodium density follows that of vertical wind. Thus, the data set defines tidal‐period perturbations of temperature and wind vector. The observed amplitudes and phases were compared to Global Scale Wave Model predictions (both GSWM00 and GSWM02). We found excellent agreement in diurnal phases and reasonable agreement in semidiurnal phases. However, GSWM02 overestimates diurnal amplitudes and both model versions underestimate observed semidiurnal amplitudes. Since the data period is long enough for the study of planetary waves and of tidal variability, we perform spectral analysis of the data, revealing a strong quasi 3‐day wave in meridional wind, a 14 hour perturbation in zonal wind, and both 14‐hour and 10‐hour periods in meridional wind, likely the result of nonlinear interactions. The observed semidiurnal amplitudes are much larger than the corresponding diurnal amplitudes above 85 km, and over a few days the diurnal and semidiurnal amplitudes vary by factors of 2–3. Causes for the observed tidal variability in terms of planetary wave modulation and tide‐gravity wave interaction are explored qualitatively.
She et al. (2004) studied this question.