South foehn in the Wipp Valley – Innsbruck region: Numerical simulations of the 24 October 1999 case (MAP-IOP 10)
- First Online:
- Cite this article as:
- Zängl, G., Gohm, A. & Geier, G. Meteorol Atmos Phys (2004) 86: 213. doi:10.1007/s00703-003-0029-8
Numerical simulations of the 24 October 1999 south foehn (MAP-IOP 10) are performed with the Penn State/NCAR mesoscale model MM5 for the Wipp Valley and the adjacent parts of the Inn Valley. The model is run in a multiple-nest configuration, the area of interest being resolved at a mesh size of 800 m in most experiments. The study serves to complement an earlier work in which typical flow features of the foehn in the Wipp Valley region were investigated by means of idealized simulations, assessing whether it is possible to reproduce the temporal evolution and the spatial structure of a particular foehn case. A further objective of the paper is to examine the dependence of the model performance on the horizontal resolution, giving some information which resolution will probably be needed for future high-resolution forecasts.
An encouragingly large part of the observed flow features could be well reproduced in the simulations. Except for a small region to the east of Innsbruck, the foehn breakthrough is predicted correctly to within an hour. The spatial structure of the so-called pre-foehn, an enhanced westerly wind occurring at Innsbruck prior to the breakthrough of the foehn, also agrees very well with the observations. Moreover, the maximum extent of the foehn in the Inn Valley, the structure of the gravity wave field above the Wipp Valley and the upvalley progression of a shallow cold front in the evening are consistent with the observations. Except for a few places where the airmass boundary between the warm foehn air and the adjacent colder air is not captured correctly throughout the time, the simulated surface temperatures range within 2 K of the observed values. Discrepancies between the model results and the observations are found in the vicinity of Innsbruck where a flow-splitting phenomenon induces a very complex flow pattern at low levels. Another source of problems is the surface potential temperature along the Wipp Valley. The observed potential-temperature increase between the Brenner Pass and Innsbruck, which appears to be related to turbulent vertical mixing of stably stratified air, is underestimated by the model. Reducing the horizontal resolution from 800 m to 1.4 km deteriorates the model performance in marginally resolved side valleys, but the results obtained for the Wipp Valley and the Inn Valley are still of high quality.