Using vertical phase differences to better resolve 3D gravity wave structure

Atmospheric gravity waves (GWs) are a critically important dynamical mechanism in the terrestrial atmosphere, with significant effects on weather and climate. They are geographically ubiquitous in the middle and upper atmosphere, and thus, satellite observations are key to characterising their properties and spatial distribution. Nadir-viewing satellite instruments characterise the short horizontal wavelength portion of the GW spectrum, which is important for momentum transport; however, these nadir-sensing instruments have coarse vertical resolutions. This restricts our ability to characterise the 3D structure of these waves accurately, with important implications for our quantitative understanding of how these waves travel and how they drive the atmospheric circulation when they break. Here, we describe, implement and test a new spectral analysis method to address this problem. This method is optimised for the characterisation of waves in any three-dimensional data set where one dimension is of coarse resolution relative to variations in the wave field, a description which applies to GW-sensing nadir-sounding satellite instruments but which is also applicable in other areas of science. We show that our new "2D+1 ST"method provides significant benefits relative to existing spectrally isotropic methods for characterising such waves. In particular, it is much more able to detect regional and height variations in observed vertical wavelength and able to properly characterise extremely vertically long waves that extend beyond the data volume.