The wave with a 1-day period was interpreted as a diurnal tide. A comparison with the migrating tide in the global scale wave model showed that the observational results had larger amplitude and shorter vertical wavelength than the model. The difference between the observation and the model may be caused by the superposition of the non-migrating tide.
The wave with the near-inertial period was interpreted as an internal inertial gravity wave. A hodograph analysis was performed in order to investigate the wave properties. It was found that the wave which appeared at a height around 19 km (just above the tropopause level) propagated southwestward with a ground-based group velocity of about 1.4 m/s.
The longer period disturbances which appeared at heights of 12-17 km had layered structures with the vertical scales of 2-4 km. They were considered to be due to inertial instability, based on the facts that the potential vorticity of the background atmosphere was nearly zero and that their phase structures were consistent with theory. It was shown by a backward trajectory analysis that the air parcel with negative potential vorticity had its origin in equatorial Indonesia. It was also shown by a forward trajectory analysis that the air parcel was transported to the Pacific to the south of Japan. This is consistent with the existence of similar layered disturbances that are shown using rawinsonde data at a station there.