THE CLIMATE ADJUSTMENT PROCESS - Wind Resource Assessment: A Practical Guide to Developing a Wind Project - page 157

• The site and reference station must be in substantially the same wind climate .This
means that variations in wind speed at each location should be well correlated
in time. The correlation can be assessed qualitatively by plotting a time series of
observed wind speeds for both the target and reference stations. A quantitative
measure such as the Pearson correlation coefficient ( r ) can also be used. The
square of the correlation coefficient, r 2 , can be thought of as the fraction of the
variation in the values of one variable that can be explained by a linear equation
with another variable.
• The target and reference stations must have homogeneous wind speed records .A
wind speed record is said to be homogeneous if the measurements have been
taken continuously at the same location and height, with the same or equiva-
lent instrumentation. In the case of the reference station, its record should be
substantially longer than, and overlap with, that of the target site.
• The concurrent target-reference period should capture seasonal variations in
the relationship . In practice, this means at least nine continuous months, and
preferably a year or more.

It will become clear in the following sections that the first two requirements, in
particular, are not always easy to satisfy.

12.2.1 Correlation
When a wind project site is in flat, open terrain, it is often easy to find a weather
station in the vicinity that experiences much the same wind climate. In regions of
greater complexity, however, it is not uncommon for the available reference stations
to be in quite different wind climates. For example, the project site may be on an
exposed ridge or mountain top, while the nearest reference stations are all in sheltered
valleys; or the project site may be near a coastline, while the available reference
stations are well inland. The result can be relatively poor correlations between the
target site and reference station (Fig. 12-2).
The weaker the correlation with the reference station, the larger the uncertainty in
the adjusted long-term wind resource at the target site. Assuming normally distributed
annual wind speed fluctuations and a homogeneous reference station data record, the
following simple equation approximates the overall uncertainty in the long-term mean
wind speed as a function of the correlation coefficient, r 2 :

r 2
N R σ

1 − r 2

N T

2 2

T

σ =

R +

σ (12.1)

Here, σ T are the standard deviations of the annual mean wind speeds of the

reference and target sites, respectively, as a percentage of the mean. Often, the same
standard deviation is assumed for both sites. A typical range based on North American
data is 3 - 6%, though it may be larger or smaller in other regions. N R is the number
of years of reference data, and N T is the number of years of concurrent reference and
target data. Because of seasonal effects, this equation should not be used if N T <

σ R and

1.