SUMMARYThis study examines the applicability of single-layer hydraulic theory to Alpine föhn winds in the example of the ow in the Brenner Pass region (Austria/Italy). The south föhn is described as the gap ow over the Pass and along the associated Wipp Valley. Numerical shallow-water simulations for a wide range of initial conditions, including shallow-and deep-föhn cases, are discussed and compared with selected measurements collected within the Mesoscale Alpine Programme (MAP). The observational analyses are based on Doppler and aerosol-backscatter lidar data for a speci c föhn case, and on surface observations and radiosoundings for all föhn cases observed within the 70-day MAP Special Observing Period (SOP).Radiosoundings at a location upstream (south) and downstream (north) of the Brenner Pass reveal that the average MAP SOP föhn case had subcritical ow south and nearly critical ow north of the Brenner Pass. The hydraulic model indicates ow transition to a supercritical state near the Pass, a hydraulic jump 2 km further north, mainly subcritical ow in the upper Wipp Valley except near the Brenner and, especially for shallow föhn, a tendency to becoming supercritical in the lower Wipp Valley. The model results suggest that the vertical topographic contraction exerts stronger control for the ow at the Brenner gap than the lateral contraction. In accordance with Doppler lidar measurements, the hydraulic model captures the decrease of the wind speed in the across-valley direction from east to west in the northern part of the valley. This asymmetric ow pattern is a result of the complex valley geometry rather than of the in uence of upper-level synoptic winds, as suggested in a previous study.For some observed cases the strength of the temperature inversion at the top of the föhn layer decreased across the Brenner Pass, presumably due to entrainment processes. Consequently, the reduced-gravity decreased as well. There are two major reasons for the only moderate agreement between the hydraulic parameters retrieved from soundings and those provided by the model: the model assumes a constant value of reduced-gravity, and the calculation of hydraulic parameters for föhn cases without a strong and sharp temperature inversion is ambiguous. Essential ow patterns in the simulated elds, such as the location of the pressure minimum and wind speed maxima, compare well with the surface observations along a valley transect. Indications for a steeply ampli ed, or breaking, gravity wave resembling a hydraulic jump to the lee of the two mountain ridges are found in lidar observations as well as in the simulations.