FLOW CHARACTERISTICS ABOVE A VERY LOW AND GENTLY SLOPING HILL ANN-SOFI SMEDMAN and HANS BERGSTROM Department of Meteorology, Box 516, S-751 20 Uppsala, Sweden (Received in final form 30 January, 1984) Abstract. Tower measurements of wind and turbulence in near neutral conditions at the top of a very low and gently sloping hill (height -20 m, with a length scale - 1000 m) are analysed in terms of current flow-over-hill theory. Measurements of wind maximum height and the change of the variances of the three wind components from the inner to the outer region are found to be in agreement with predictions from the theory. Spectra of the longitudinal and vertical wind components in the inner region, scaled according to Panofsky et al. (19X2),come close to the corresponding Kansas curves in the high frequency range. They have higher energy in the low frequency region, probably a spectral lag effect caused by rougher upwind terrain. In the outer region, the spectra coincide with the corresponding Kansas curves if normalized by their respective variances and plotted against flf,. 1. Introduction In late 1978 the Swedish National Board of Energy Source Development decided to construct two 3-MW wind energy conversion systems, one of which was located at Maglarp, close to the southern coast of Sweden. The site is situated about 2 km from the seaat a low hill (20 m above sealevel) in a slightly rolling farmland area, as shown in Figure 1. To provide proper design criteria for the installation, profile measurements of temperature, wind speedand wind direction were started about a year later on a 120m high tower, and in 1981 two field studies including turbulence measurements were performed. As the area around the site is rather flat, the first step in the analysis was to interpret the data in terms of Monin-Obukhov similarity. However, the result was not very encouraging and the question was raised whether the low hill could affect the flow substantially. For instance Jensen and Petersen (1978) pointed out that a surface elevation change of 2 m over 50 m caused accelerated upper winds and typical shearing stress profiles. In this case, however, the elevation change is much smaller, 20 m over 700-2000 m. Nevertheless, an attempt is made to interpret the data in terms of flow- over-hill theory. 2. Wind Flow over Hills The presence of a hill produces a number of modifications on the air flow in the atmospheric boundary layer. Both the wind speedand the turbulence structure change as the hill is traversed. There are now many studies using two-dimensional models as Boundary-Layer Meteorology 29 (1984) 21-37. OOOS-8314/84/0291-0021$02.55. 0 1984 by D. Reidel Publishing Cornpan>,.