Applications - Surface Based Remote Sensing of the Atmospheric Boundary Layer - page 102

data show that the value of the exponent is often much larger than its assumed value
of 0.2, and that it varies with height.
Over the U.S. Great Plains, Banta et al. ( 2002 , 2006 ) found that the directional
shear with height was generally small when LLJs of about 12 m s − 1 or more were
present. Under these conditions, along-wind vertical slice scanning can be used to
produce wind-speed profiles at intervals of 30 s or less. When significant directional
shear is present, the wind profiles would need to be monitored using conical scans
at one or more elevation angles, which may take 2-5 min, depending on how many
elevations are used for the sampling.
Peña et al. ( 2009 ) compare wind profiles well beyond 80 m, obtained with a small
wind LIDAR, with their new extended laws (see eqs. (2.21) and (2.33) ) for wind
profiles in the ABL. In addition to the usefulness of mean-wind data for long-term
wind energy site assessment, LIDAR profile data could also be useful for real-time
wind turbine operations. Spatially averaged, statistically significant profile data are
available at time intervals of 2 min or less, which could be used to adjust turbine
operational settings to accommodate changes in wind speed and direction. Although
the range resolution of HRDL is 30 m, the desired vertical resolution of ~10 m
or less can be achieved by scanning (or pointing) at lower elevation angles, or by
performing elevation (vertical-slice) scans.
An overview of optical techniques for detecting wind profiles for wind energy
applications is given in Emeis et al. ( 2007c ). A special application result from a
wind LIDAR specially developed for wind energy tasks is presented in Kindler et al.
( 2007 ). Here, a conically scanning LIDAR which uses beam focusing for range
detection is used and tested.
Conical scanning with LIDARs for the determination of wind profiles can lead
to problems in complex terrain, because this technique assumes horizontal homo-
geneity within the volume covered by the conical scan. However, in complex and
mountainous terrain this assumption is often not valid. A first attempt for an esti-
mation of the occurring error in such situations has been made from flow model
simulations by Bingöl et al. ( 2009 ). They found errors in the order of 10%, but their
model was the linearized model WAsP. So, this issue has to be investigated further
with more sophisticated flow models.
Doppler LIDAR data at high resolution are also useful for characterizing tur-
bulence properties of the atmosphere, either as mean statistics or as revealing
turbulence events. Banta et al. ( 2006 ) defined vertical averaging bins for sequen-
tial vertical-slice scans and calculated mean and variance profiles of the streamwise
wind component. These profile statistics were composited to relate mean and tur-
bulence profile properties to those of the LLJ. Techniques for calculating other
turbulence properties have been demonstrated, including second, third, and fourth
moments of the vertical velocity (Lenschow et al. 2000 ) and velocity structure
functions and coherence (Lothon et al. 2006 ).
Several methods have also been proposed for estimating the eddy dissipation rate
from Doppler LIDAR data (e.g. Banakh and Smalikho 1997 ; Frehlich and Cornman
2002 ; Smalikho et al. 2005 ), which provide information on the fine-scale turbulence.
Such fine-scale information is valuable for indicating overall turbulence levels or the