Development of a Season/Flow Dependent Gravity Wave Drag Parameterization for the NOAA FIM Global Atmospheric Model

Orographic gravity wave drag (GWD) has been shown to significantly affect the strength of the upper tropospheric jet stream. Exclusion of GWD in numerical weather prediction (NWP) models tends to produce stronger westerlies than observed, particularly in the Northern Hemisphere. Overly intense westerlies imply a stronger equator-pole temperature gradient, which in turn overly deepens mid-latitude cyclones and causes these storms to move faster than analyzed. GWD can reduce this bias, but it must be parameterized as the gravity waves that slow the polar jet occur on smaller space and time scales than are explicitly resolved by even high-resolution global NWP models. The new NOAA ESRL Flow-following, Finite-volume Icosahedral Model (FIM) utilizes the NOAA NCEP Global Forecast System (GFS) physics suite, including an important update in 2012 to the orographic GWD parameterization. However, fundamental differences in the FIM and GFS dynamical cores require an adaption or ‘tuning' of the GWD parameters. In addition, the input GFS topographic grids that drive the GWD were optimized for the GFS and thus were adjusted to the FIM grid. Preliminary experiments demonstrated a large impact of both the GWD parameters and the parameterization's technical implementation on NWP skill scores. Two three-month retrospective FIM runs for both the 2013 boreal summer and the 2014 boreal winter revealed large seasonal differences in GWD impact on standard NWP metrics, in particular the 5-day 500 mb, Northern Hemisphere geopotential height anomaly correlation. Investigation of the most effective values for the GWD parameters as a function of model spatial resolution and the implementation of a seasonal and flow dependent GWD parameterization will be discussed.