Wavelet analysis of turbulence in cirrus clouds

Two ̄ights of the UKMeteorological Oce's Hercules aircraft through daytime frontal cirrus around Scotland have been analysed using wavelet analysis on the vertical velocity time-series from the horizontal runs. It is shown that wavelet analysis is a useful tool for analysing the turbulence data in cirrus clouds. It ®nds the largest scales involved in producing turbulence, as does Fourier analysis, such as the 2-km spectral peaks corresponding to convective activity during ̄ight A283. Wavelet spectra have the added advantage that the position is shown, and so they identify smaller-scale, highly localised processes such as the production of turbulent kinetic energy by the breaking of KelvinHelmholtz waves due to the vertical shear in the horizontal wind. These may be lost in Fourier spectra obtained for long time-series, though they contribute something to the average spectral density at the appropriate scale. The main disadvantage of this technique is that only octave frequency bands are resolved.