OFFSHORE RESOURCE ASSESSMENT - Wind Resource Assessment: A Practical Guide to Developing a Wind Project - page 213

as surface-based measurements, another satellite-based sensor, or a numerical weather
model).
Like other satellite-based microwave sensors, SAR systems provide a snapshot (or
image) of the near-surface winds at a particular time. Since the image retrievals must
be requested from the multipurpose sensor, the number of archived SAR images varies
from one region to another. Furthermore, because they cover a smaller area (a few
hundred kilometers wide at most), SAR satellite passes over the same location are
less frequent than those of satellite-based SCATs and radiometers. One consequence
is that SAR systems, unlike SCATs and radiometers, do not provide a gridded data
set, and some expertise in radar image processing is needed to derive the wind speeds
from the raw images.
14.4.2 Wind Resource Assessment Using Satellite-Based
Microwave Sensors
Wind resource maps from radiometers and SCATs can provide a useful initial indi-
cation of the wind resource in a particular region. Among their advantages are that
they are available for free and they offer global coverage with daily or more frequent
satellite passes. An extensive archived data set from the SSM/I radiometer is available
since 1987. However, the wind speeds derived from SCATs such as QuikSCAT are
considered to be more accurate than those derived from radiometers. While the period
of record of the QuikSCAT data set, 1999-2009, 4 is not as long as that of SSM/I, it
is sufficient to obtain fairly reliable estimates of long-term mean wind speeds. Several
studies have shown that the instantaneous wind vector accuracy (root-mean-square
error) for QuikSCAT data is around 1.0-2.0 m/s, which may be somewhat better than
that of numerical weather prediction models but is worse than that of anemometers,
lidar, or sodar. Figure 14-12 shows two examples of the kinds of global ocean surface
wind resource maps that can be obtained from the QuikSCAT data.
Satellite-based SAR imagery, on the other hand, offers considerable promise for
project-scale resource assessment, as it can achieve a much better resolution than
radiometers and SCATs and can also measure winds near coastlines (as demonstrated
in the SAR image in Figure 14-2a, which shows complex near-shore wind patterns
in the St. Lawrence River). Although the spatial resolution of SAR is in principle
very fine indeed, in practice, problems such as speckle limit the grid spacing for most
applications to a few hundred meters. This is nonetheless comparable to the spacing
between wind turbines in a wind project. Thus, SAR can provide information about
the wind resource distribution of practical use for designing wind projects.
Several studies have shown that wind speeds derived from SAR are about as
accurate as SCAT-derived speeds when compared to buoy measurements. Other than
the inherent limitations of the method, the utility of SAR-derived wind speed data
depends mostly on the number of SAR images available for a particular site of
4 NASA PO.DAAC web site: http://podaac.jpl.nasa.gov/. Ifremer web site: http://www.ifremer.fr/cersat/
en/data/overview/gridded/mwfqscat.htm. QuikSCAT stopped operating in November 2009.