Applications - Surface Based Remote Sensing of the Atmospheric Boundary Layer - page 94

and windprofilers perform these three-dimensional wind soundings by emitting
upward three to five beams inclined to each other by about 15 ◦ to 20 ◦ (Doppler-
beam-swinging method). RADARs and LIDARs usually make conical scans around
the zenith with comparable aperture angles. This scanning mode is also called VAD
(velocity-azimuth-display) mode (Browning and Wexler 1968 ; Banta et al. 2002 ).
For calculation of a mean wind, it has to be assumed that the wind field is homo-
geneous over the horizontal distance enclosed by the beams or the conical scan.
Thus, remote-sensing methods due to their design yield values that are representa-
tive of larger air volumes. This fact hampers the direct intercomparison with cup
anemometer data in locations where the wind flow is non-uniform over the region
of space encompassed by the mast and LIDAR scan; but, on the other hand, remote-
sensing methods are very likely more suited for wind energy applications because
the air volume scanned by the measurement device and the air volume which passes
through the rotor plane of a wind turbine are of comparable size.
4.3.1.1 Acoustic Techniques
The operation of Doppler-SODAR devices in the Doppler-beam-swinging (DBS)
mode was the first remote-sensing technique to continuously observe vertical pro-
files of wind speed and direction in the ABL. With these instruments, the increase
of wind speed with height and the turning of the wind direction with height over dif-
ferent types of land use (e.g. over large cities) can be investigated. Likewise, wind
maxima at inversions and fronts can be observed. Also the modification of the wind
profile by the presence of hills, escarpments, and land use changes can be studied.
Diurnal wind phenomena related to topographic features (sea shores, valleys,
mountains) and nocturnal phenomena such as low-level jets are dealt with further
below in Section 4.5 .
Wind and Turbulence Profiles over Flat Terrain
Over flat terrain the vertical wind profile most likely follows the known theoretical
wind profile laws given in Chapter 2 . The left frame of Fig. 4.11 shows an exam-
ple, where monthly mean vertical profiles of the scale parameter of the Weibull
distribution (the two-parameter Weibull distribution is frequently used to describe
the frequency distribution of wind speeds, its scale parameter is proportional to the
mean wind speed, while its shape parameter defines the shape of the distribution)
over flat terrain are compared to analytical profiles for the Prandtl layer (see (2.5a))
and the wind resource model WAsP (Troen and Petersen 1989 ). In the example dis-
played in the left frame of Fig. 4.11 , the measured wind profiles follow the analytical
profiles quite well for heights up to 60-80 m. This can be taken as an indication
that the average depth of the Prandtl layer was of this magnitude. Modifications
of such profiles mainly occur by land-use changes and by the passage of synoptic
disturbances (fronts).
It was already noticed by the very first researchers who looked into wind pro-
file data (see Section1.1) in 1915 and 1921 that the diurnal variation of wind speed