1Colorado Research Associates Division, NorthWest Research Associates, Inc., Boulder, CO 80301, USA
2NASA Ames Research Center, Moffett, CA 94035, USA
3Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA
Abstract. ER-2 MMS and MTP wind and temperature measurements during the CRYSTAL-FACE campaign in July 2002 were analyzed to retrieve information on small scale gravity waves (GWs) at aircraft's flight level. For a given flight segment, the S-transform was used to search for and identify small horizontal scale GW events, and to estimate the apparent horizontal wavelengths of the events. The horizontal propagation directions of the events were determined using the Stokes parameters method combined with the cross S-transform analysis. The MTP temperature gradient method was used to determine the vertical wavelengths of the events. GW momentum fluxes were calculated from the cross S-transform. Other wave parameters such as intrinsic frequencies were calculated using the GW dispersion relation. More than 100 GW events were identified. They were generally short horizontal scale and high frequency waves with λz of ~5 km and λh generally shorter than 20 km. Their intrinsic propagation directions were predominantly toward the east, whereas their ground-based propagation directions were primarily toward the west. Among the events, ~20% of them had very short horizontal wavelength (<10 km), very high intrinsic frequency (ω/N≥0.8), and relatively small momentum fluxes, and thus they were likely trapped in the lower stratosphere. The averaged magnitude of vertical flux of horizontal momentum was ~0.026 kg m−1 s−2, and the maximum magnitude was ~0.13 kg m−1 s−2.
Using the estimated GW parameters and the background winds and stabilities from the NCAR/NCEP reanalysis data, we were able to trace the sources of the events using a simple reverse ray-tracing. More than 70% of the events were traced back to convective sources in the troposphere, and the sources were generally located upstream to the events. Finally, a probability density function of GW cooling rates was obtained in this study, which may be used in cirrus cloud models.