Sensitivity of Gravity Wave Momentum Flux Estimates on Separation Methods

Gravity waves (GWs) can manifest and exert influence across various scales, challenging their direct separation from large-scale background flows through spectral filtering. Horizontal divergence is commonly used as a proxy for GWs, but the rotational motion of GWs and the divergent motion of Kelvin and mixed Rossby gravity (MRG) waves may introduce uncertainties. To evaluate the reliability of the methods above, this study examines the variation in vertical and lateral momentum fluxes among three GW separation methods: spherical harmonic filtering (SHF), divergent wind from Helmholtz decomposition with SHF (DIVF), and inertia–gravity (IG) mode from normal-mode decomposition with SHF (IGF). The analysis of selected boreal winter and summer cases shows that DIVF and IGF exhibit similar results, whereas direct SHF shows notable discrepancies, particularly in the upper troposphere. This finding suggests that the rotational motion of GWs and divergent motion of Kelvin and MRG waves only minimally affect the estimation of GW momentum flux (GWMF), whereas the rotational motion of non-IG (Rossby, Kelvin, and MRG) modes can introduce significant biases. The differences in the upper troposphere are larger than those in the stratosphere, where divergent motion dominates both the total and the GW circulation fields. In the stratosphere, although the vertical GWMF shows minimal variation between methods, the lateral GWMF obtained by direct SHF exhibits visible biases. Given the similarity between DIVF and IGF, using divergent wind components is recommended for diagnosing GW quantities related to horizontal wind perturbations. This comparison offers valuable insights into accurately extracting GWMF. Significance Statement The purpose of this study is to evaluate the effectiveness of direct spectral truncation and divergent wind proxy for separating gravity waves in the upper troposphere and stratosphere from the perspective of wave momentum flux. The findings suggest that divergent wind proxy provides more reliable estimates of gravity wave momentum flux, especially in the upper troposphere where direct spectral truncation introduces significant biases due to the rotational motion of non-IG (Rossby, Kelvin, and MRG) modes. The effects of rotational motion in gravity waves and the divergent motion of Kelvin and mixed Rossby gravity waves are minimal in momentum flux estimation. This work offers valuable insights for guiding gravity wave separation and is expected to promote future research in gravity wave dynamics.