Motion extraction of atmospheric waves from spectroscopic imaging

We describe a method for extracting atmospheric gravity wave motion from spectroscopic airglow images using two-dimensional spectral analysis. This method detects quasi-monochromatic wave components and estimates wavelengths, wave propagation directions, and phase speeds from the cross periodogram of two consecutive time-differenced images. Monte Carlo simulations have been used to validate the phase-difference technique for motion estimation and to compare its performance with a translational block-motion extraction method, which represents the conventional approach. The new technique is shown to be more precise than its predecessor. We have also applied the two techniques to real airglow images for both statistical and case studies. The variances of the phase speeds generated by both methods are insignificant relative to the geophysical variation of the gravity wave phase speeds. An individual wave study illustrates that the phase-difference method provides more stable estimation of phase speeds than the block-motion method.