Gravity-wave refraction by three-dimensionally varying winds and the global transport of angular momentum

Current gravity wave parameterization schemes are based on a model of gravity wave propagation that assumes a horizontally homogeneous mean state. The paradigm of wave-mean interaction that emerges from such models indicates that persistent forcing of the mean flow occurs only where waves dissipate. Recently, idealized studies have indicated that non-trivial mean flow forcing can emerge without dissipation when waves are refracted by horizontally inhomogeneous winds. This talk will discuss the relative importance of these non-dissipative wave-mean interactions in the real atmosphere, as well as clarifying conceptual issues of localized wave-mean interactions in three dimensional winds.

Conceptually, the generalization from theories of wave-mean interaction in horizontally homogeneous flows to the case of localized wavepackets propagating through three-dimensionally varying fields presents difficulties that are not widely appreciated. Standard pseudomomentum arguments are highly dependent on simplified geometry, and become more complicated in three dimensions. We propose a generalization of the Eliassen-Palm flux that respects global conservation of angular momentum, and that can be used to compute the mean forcing associated with non-dissipative refraction of a ray tube.

To assess the impact of these new forces, a three-dimensional global ray tracing code in spherical coordinates has been developed. Using the WKB ray tracing formalism, it computes the propagation and refraction of ray tubes through steady, three dimensionally varying atmospheric winds, either analytically specified or given on a grid. Unlike pre-existing ray tracing models, it correctly models the refraction of waves by the curvature of the earth's atmosphere, allowing accurate computation of wave paths on global scales.

This model has been used to compute the monthly mean vertical profile of vertical angular momentum flux due to topographic gravity waves in the CCCma Third Generation Atmospheric GCM, both with and without non-dissipative wave--mean interaction effects. This experiment allows an estimate of the impact of non-dissipative wave-mean interactions on the global angular momentum budget.