THE CLIMATE ADJUSTMENT PROCESS - Wind Resource Assessment: A Practical Guide to Developing a Wind Project - page 166

data set. With so few observations, a correlation based on daily mean speeds may
yield a poor result, and therefore, weekly or monthly means may be called for. This
substantially reduces the amount of independent information available for establishing
the target - reference relationship.
Despite these disadvantages, it is often a good idea, especially for projects in
complex terrain, to obtain rawinsonde data from nearby stations either for direct use
in MCP or for verifying the homogeneity through time of available surface weather
data.
12.3.4 Modeled Data Sets
In recent years, the use of reference data sets created by atmospheric models has
become more common, although it is not yet the industry norm. They are sometimes
called virtual meteorological masts ,or VMMs . The most well-known type of modeled
data is called reanalysis data . It comes in several varieties and is produced by a num-
ber of national weather agencies, including the National Centers for Environmental
Prediction (NCEP)/National Center for Atmospheric Research (NCAR) and the Euro-
pean Center for Medium-Range Weather Forecasts (ECMWF). The NCEP/NCAR data
are available for free and therefore tend to be the most widely used.
All reanalysis data sets are created by using historical weather observations (gen-
erally from surface, rawinsonde, satellite, and aircraft-borne instruments) to drive a
global or regional NWP model. From these model runs, weather parameters (including
temperature, pressure, wind, precipitation) are extracted for every grid point and every
level in the model. Reanalysis data sets were created to support climate studies. Unlike
real-time weather forecasting models, which are frequently modified, the reanalysis
models are fixed for the entire historical simulation.
Reanalysis data have a number of positive attributes, including convenience, multi-
ple levels and types of weather parameters, and a long data record (more than 60 years
for some data sets). Because the gridded data are available everywhere covered by the
model, there is no difficulty finding suitable grid points. This eliminates much work
searching for surface weather stations and data sets, and it provides a common data
source for all MCP studies. In parts of the world where surface weather observations
are unreliable, reanalysis data (and other modeled data sets) may be the only feasible
source of reference data for MCP.
However, reanalysis data also have significant disadvantages and must be used
with caution. First, the correlation of the reanalysis winds with tower observations
depends on the complexity of the terrain and the resolution of the reanalysis model.
The NCEP/NCAR global reanalysis data set, in particular, is relatively coarse, with a
resolution of about 2 ◦ in latitude and longitude (a little over 200 km) and thus may
give poor results in mountainous terrain, at coastal boundaries, and in other places
where there is a sharp wind gradient.
More importantly, the homogeneity of reanalysis data is limited by that of the
observational system used to drive the model, which has changed dramatically over
the decades. The bulk of the weather observations in the 1950s and 1960s came from
weather balloons supplemented by land and ship-based surface observations. Weather