Lidar observations of the middle atmospheric thermal tides and comparison with the High Resolution Doppler Imager and Global-Scale Wave Model 1. Methodology and winter observations at Table Mountain (34.4 degrees N)

The tidal signature in the middle atmospheric thermal structure was investi gated using more than 140 hours of nighttime lidar measurements at Table Mo untain (34.4 degrees N) during January 1997 and February 1998. The lidar pr ofiles (30-85 km) revealed the presence of persistent mesospheric temperatu re inversions around 65- to 70-km altitude with a clear local solar time (L ST) dependence. Daytime temperature profiles (65-105 km) obtained by the Hi gh Resolution Doppler Imager (HRDI) on board the Upper Atmosphere Research Satellite (UARS) in January and February from 1994 to 1997 and zonally aver aged at the latitude of the Table Mountain Facility (TMF) were considered t ogether with the lidar results. The daytime HRDI and nighttime lidar temper ature differences from their respective daytime and nighttime averages were compared to the equivalent differences predicted by the Global Scale Wave Model (GSWM). A remarkable consistency was observed between the lidar and t he HRDI upper mesospheric thermal structure, with a continuous downward pro pagation of warm temperatures from 100 km at 1000 LST to 75 km at 2000 LST and 65-70 km at 0300-0500 LST, surrounded above and below by colder tempera tures. This structure was predicted by GSWM but with a 2- to 4-hour delay a nd a weaker amplitude. On the lower side of this structure (i.e., 65-70 km) a thin layer, characterized by early night cold temperatures and late nigh t warm temperatures, was identified as the result of the downward propagati on of the temperature inversions. Using a new analysis technique, which we have named "constrained wave adjustment" for convenience in future referenc e, and assuming that the observed temperature variability was entirely driv en by tides, some estimations of the diurnal and semidiurnal phases and amp litudes were made from the lidar measurements between 40- and 85-km altitud e. Although it does not allow a complete and accurate extraction of the tid al components, this new method appeared to work well fur the present TMF st udy. The estimated diurnal amplitude exhibited a minimum at 63 km with a fa st phase transition, characteristic of the transition between the upper str atospheric trapped modes (phase at 1800 LST) and the upward propagating mod es. This transition layer was predicted by GSWM to be at 5-km lower altitud e. This altitude shift was present throughout the middle mesosphere. Immedi ately above the transition layer, the very fast growing diurnal amplitude b etween 65 and 72 km was followed by a substantial decrease and by the emerg ence of the semidiurnal component, resulting in the formation of the mesosp heric temperature inversion layers. However, the amplitude of the inversion s remained large compared to the theoretical tidal predictions and a differ ent formation mechanism should possibly be considered. Recent modeling stud ies have shown that gravity wave breaking can be significantly affected by the tidal background winds and some preferential wave breaking times could emerge that are dependent on the phase of the diurnal tide and the characte ristics of the dissipating waves. This "LST filtering" could result in LST- dependent temperature inversion layers similar to those observed by lidar. The new analysis technique presented in this paper was also applied to stud y the temperature tidal oscillations at Mauna Loa Observatory (19.5 degrees N). The results are presented in a companion paper [Leblanc et al., this i ssue].