A three-dimensional mountain wave imaged in satellite radiance throughout the stratosphere: Evidence of the effects of directional wind shear

Swath-scanned thermal radiances from the Advanced Microwave Sounding Units (AMSU-A) on the NOAA and Aqua satellites are used to image the horizontal temperature structure of a long-wavelength mountain wave that formed over southern Scandinavia on 14 January 2003. Data from all six stratospheric channels show this wave propagating through the full depth of the stratosphere. In channels 9-11 (attitudes similar to 20-90 hPa) the imaged wave has a phase structure broadly consistent with a stationary wave radiated by the southeastward flow over and above the quasi-elliptical terrain of southern Norway. Channel 12 radiances at similar to 10 hPa, however, show a remarkable abrupt change in imaged wave structure: the horizontal wavelength contracts, phase lines rotate anticlockwise by 30 degrees-40 degrees, and peak activity migrates to the south to lie over Denmark and northern Germany. Similar structure persists in channels 13 and 14 (at similar to 2-5 hPa) with progessively increasing radiance amplitudes. These features are stable over the similar to 10 hours of AMSU-A measurements from five separate overpasses, and are validated against independent radiances acquired by the Atmospheric Infrared Sounder (AIRS) and retrieved AIRS/AMSU-A temperature profiles from the Aqua overpass. This change at the channel 11/12 interface coincides with an onset of anticlockwise rotation (backing) and intensification of background stratospheric winds with height. Fourier-ray and spatial ray modelling incorporating these directionally sheared winds and simplified orographic forcing reproduce the salient features of the observations, but only after wave-induced temperature perturbations have been converted to channel radiances using a forward model. The differential visibility of various components of this three-dimensional mountain wave to the AMSU-A channel weighting functions has a first-order impact on the observations at heights above 10 hPa. Once that is factored in, the combined observations and modelling provide direct experimental support for the Shutts model's predictions of how backing wind vectors affect the vertical evolution of three-dimensional mountain waves. Implications of these observations for orographic gravity wave drag parametrization are briefly discussed. Copyright (c) 2007 Royal Meteorological Society.