Using satellite radiance data to validate resolved and parameterized orographic gravity waves in global models

Orographic gravity wave (mountain wave) drag is an important component of the global middle atmospheric circulation. As these dynamics are unresolved in global climate models, their effects on the resolved flow must be parameterized. High-resolution global numerical weather prediction models are now explicitly resolving the long-wavelength “outer scales” of the mountain wave spectrum, but must still parameterize the larger unresolved component. Similarly, new satellite remote sensing instruments can now resolve long-wavelength mountain waves, while remaining insensitive to shorter wavelength motions that are thought to carry much of the momentum flux. Here we bring all these tools together to perform detailed case studies of individual long-wavelength stratospheric mountain wave events. We focus on the large-amplitude stratospheric mountain wave that occurred over southern Scandinavia on 14 January 2003, which produced polar stratospheric ice clouds that were profiled with aerosol lidars from the NASA DC-8. Swath radiance imagery from the Advanced Microwave Sounding Unit-A (AMSU-A) and Advanced Infrared Sounder (AIRS) both show this wave propagating to the stratopause to dominate the Arctic stratospheric mesoscale variability on this day. The wave shows an interesting rotation in horizontal phase structure with height, which we model and explain using idealized Fourier-ray experiments. High-resolution analysis and forecast/hindcast fields from the T_L511L60 European Centre for Medium Range Weather Forecasts (ECMWF) Integrated Forecast System (IFS) and the T479L60 Navy Operational Global Atmospheric Prediction System-Advanced Level Physics High Altitude (NOGAPS-ALPHA) yield a similar-looking wave disturbance that propagates through the full depth of the stratosphere. Forward modeling of these simulated fields yields simulated AMSU-A and AIRS perturbation radiances that compare well with the observations. NOGAPS-ALPHA runs at different horizontal and vertical resolutions demonstrate the sensitivity of the explicitly resolved wave to spatial resolution and horizontal (spectral) hyperdiffusion. We then compare the observed and simulated wave's properties with the diagnostic predictions of a variety of orographic gravity wave drag parameterizations, operating on profiles over southern Scandinavia from NOGAPS-ALPHA runs at various spatial resolutions. The preliminary conclusions and potential implication for parameterizations and global models are briefly discussed based on results for this wave and another mountain wave event over Greenland.