Innsbruck Foehn Studies V: GAP Flow

Unique gap flow measurements were made along the Brenner pass cross section (Italy-Austria) during the field phase of the Mesoscale Alpine Programme (Sept. 7 - Nov. 15, 1999). The goal was to improve the understanding of dynamically driven gap winds through and over a lateral and vertical constriction, which manifest themselves as south foehn on the downstream side. In particular, the relative importance of changes in gap width and terrain height for the flow behavior and the forcing of the flow from aloft were to be examined. Researchers from 12 institutions in North America and Europe amassed a temporally and spatially dense data set from several radiosounding stations, Doppler sodars, a wind profiler, a scanning Doppler lidar, 70 surface stations, and 4 research aircraft. The campaign was coordinated by the applicant. The objectives of this proposal are fourfold: 1) to perform exemplary, detailed and comprehensive case studies of gap flow, and to examine the temporal and spatial variability of gap flow. This will form the basis for the other goals: 2) to study the predictability of the gap flow along the Brenner cross sections and to evaluate the performance of (semi-)operational research numerical weather prediction models, and 3) to compile a long-term climatology of gap flow from meso-gamma scale analyses spanning the whole Alps. The analyses themselves will be performed in a different proposal, which is included in the current proposal "bundle".

Almost a century ago ground-breaking research of the phenomen of strong, gusty and often dry and warm winds at the foot of mountain ranges was performed. Results were reported in the "Innsbruck foehn studies parts 1-4". Recently, a large international measurement campaign, the "Mesoscale Alpine Programme" (MAP), took another look at foehn with a vast array of instruments. These data and additional data from a small-scale measurement campaign of another mountain fall wind along the Adriatic coast, the "bura", were used to find out more how and why air "falls" down and speeds up along the downslopes of a mountain range and how often foehn winds occur. Our work put together an almost complete picture of foehn out of the puzzle pieces that previous generations of foehn researchers had accumulated and that MAP had produced. The most appropriate way to picture foehn is by likening it to a wide river or lake where water flows over and shoots down a weir. A lot of water slowly moves towards the weir, then becomes very shallow (typically less than a meter deep) as it moves very fast down the weir. Air behaves similarly, only that it is typically hundreds of meters deep when it plunges down the lee side of a mountain range. Whereas engineers make the top of a weir nicely horizontal and smooth, mountain ranges are jagged and full of indentations and gaps. Naturally, air will plunge through the gaps before it reaches high enough to spill over the mountain crest. Incidentally we found that the famous foehn locations all over the world are downstream of such gaps. Even for an experienced meteorologist, it is not always easy to tell foehn apart from air that flows down mountain slopes at nighttime, when air cools off and thus becomes heavier. Identifying the presence of foehn had been done subjectively by examining the changes of wind speed and direction, temperature and relative humidity with time by a meteorologist. The drawback was that the classification varied between different meteorologists and that looking at decades worth of data or data from many locations was too time consuming. We developed a computer algorithm that reliably identifies foehn periods. This algorithm was used to compile climatologies of foehn on both the southern and northern side of the central Alpine crest. The windiest locations endure foehn on average 20% of the time in the course of a year! Even the biggest computers today are not powerful enough to forecast weather taking into account all the gaps and valleys of a mountain range. Only the coarse outlines of e.g. the Alps are in the computer model. For example, Innsbruck in the Alps of the computer model Alps lies at 1500 m instead of 600 m! Even though there is e.g. no Wipp Valley foehn in the forecasting model there are some tell-tale signs that we could use to forecast the probability of foehn in the Wipp Valley with excellent skill to 3 days in advance.