The in situ vertical velocity data from the lowest-level aircraft, the NCAR Electra, indicated the presence of gravity waves at 5500 m that were characterized by relatively short horizontal wavelengths of approximately 8 km with a maximum amplitude of 3.0 m/s and a maximum vertical displacement of 800 m. Backscatter data from the downlooking Scanning Aerosol Backscatter Lidar (SABL) data was enhanced by several layers of low-level clouds and moisture that suggested maximum peak-to-trough wave amplitudes of 1 km and typical horizontal wavelengths of 7-10 km present in the lee of the narrow topographic peaks. Spectral decomposition of the topography along the aircraft transects indicates three dominant horizontal wavelengths of 28, 8, and 2.2 km and is consistent with the short wavelength gravity wave response. Two transects from the Electra indicate a sharp vertical velocity maximum and a large upward parcel displacement suggestive of wave breaking near the 5500 m level, which was located in the strong wind shear layer. The in situ vertical velocity data from the Electra, the Met Office C130 and DLR Falcon transects above 5500 m measured weak gravity waves with vertical velocities less than 1.5 m/s in the upper troposphere and lower stratosphere. Backscatter data from the downlooking DLR DIAL lidar onboard the Falcon also confirm that a rapid decrease in the wave amplitude with height was present in the layer between 5500 and 6500 m. Factors that may contribute to the rapid decay of wave amplitude include: the decrease in the Scorer parameter with height that results in an evanescent wave response and ducting of small horizontal wavelengths, gravity wave breaking in the shear layer, and a directional critical level that will partially absorb wave energy.
Numerical simulation results from COAMPS using 5 nested grid meshes with a minimum horizontal resolution of 556 m and 55 vertical levels are used to diagnose the gravity wave dynamics in this case. Mountain waves with dominant horizontal wavelengths of less than 10 km are simulated, in general agreement with the in situ aircraft and lidar observations, however, some phase errors occur in the positions of the significant wave crests and troughs. The model results suggest that low-level topographic blocking results in a reduction of the effective height and width of the Alpine peaks, which leads to a reduction of the amplitude and dominant horizontal wavelength of the mountain waves. Gravity wave breaking characterized by convective instability and a turbulent kinetic energy maximum is simulated in the lee of the first substantial topographic peak in the 3000-5500 m layer, in close agreement with the in situ Electra and SABL data. The relative roles of the directional critical level and the wave trapping will be further explored using three layer linear analytic solutions with an initial state based on the composite upstream dropsonde data.