AWS Truepower, LLC released a statement today calling into question media reports of declining wind speeds across the Northern Hemisphere. Such a decline, if true, could have serious consequences for the development of wind energy projects, which depend on a stable wind climate. Based on its own review and analysis of available data and studies, however, AWS Truepower believes that no significant decline has been definitively observed.
“As far as we can determine,” says Dr. Michael Brower, a physicist and Chief Technical Officer of AWS Truepower, “average wind speeds in locations and at heights of interest for wind energy have been neither decreasing nor increasing in a statistically significant way in most parts of the world.”
An article published on October 17 in Nature Geoscience by Robert Vautard and colleagues from two European climate research institutes has been cited as showing that surface wind speeds have decreased during the past few decades across the Northern Hemisphere. This is the latest in a series of media reports publicizing such a trend.
“The key problem with these reports,” says Dr. Jeff Freedman, a Senior Research Scientist with AWS Truepower, “is that they focus mainly on findings based on surface wind speed measurements taken at airports and the like. Such measurements are notoriously unreliable for long-term wind studies because of changes over the years in measurement equipment, instrument heights and locations, and surroundings.”
The Nature Geoscience study reviewed data from 5412 surface stations around the globe, but in an indication of how difficult it is to find reliable long-term surface wind data sets, a quality-control procedure employed by the researchers resulted in 85% of these stations being excluded from the analysis. From the remaining 822 stations, the authors found that observed wind speeds have decreased over the past 30 years at average rates ranging from 1.8% to 5.9% per decade over most land surfaces of the Northern Hemisphere.
However, two other data sources evaluated in the same article – the NCEP/NCAR global reanalysis data set and the ECMWF ERA-interim reanalysis data set, both the output of weather models employing data from a wide variety of sources – do not support this finding. On the contrary, the NCEP/NCAR global reanalysis data indicate no significant trend in wind speeds in the past 30 years, whereas the ECMWF data set shows rates of decrease that are 50% to 90% smaller than those seen in the surface observations over the past 20 years, according to the article. The ECMWF trends are concentrated in northern Europe and parts of central Asia.
The most likely explanations for the discrepancies between the various data sources involve the surface wind observations. The authors note that an increase in surface roughness caused principally by tree growth around the stations could explain between 25% and 60% of the observed decrease.
Yet Dr. Freedman observes that wind energy projects are likely to experience much smaller changes in wind speed due to this phenomenon. “Wind speeds at the height of modern large turbines [typically 70 to 90 m, or 230 to 330 ft, above ground] are much less influenced by surface conditions than they are at the 10 m (33 ft) height of most of these measurements.”
AWS Truepower observes further that changes in measurement practices in many parts of the world may have created false trends in the data. In the United States, for example, nearly all airport weather stations were moved in the past 15 years from 6.1 m (20 ft) masts to 10 m (33 ft) ground masts, the current international standard, and automated data recording was introduced.
“It is well known in the US wind industry,” according to Dr. Brower, “that these changes have resulted in marked discontinuities and inconsistencies in surface wind measurements – and most often a drop in the reported average wind speeds.”
“In much of the rest of the world,” adds Dr. Brower , “inconsistent maintenance practices and encroaching urbanization have caused substantial declines in observed wind speeds in many locations. Because they are so pervasive, such trends are often difficult to detect and remove even with the most stringent quality-control routines.”
The fact that other data such as weather balloon data and reanalysis data do not support declines in observed surface wind speeds has received relatively little attention in the press. Such data sets tend to be more reliable indicators of long-term wind speed trends because they are not sensitive to local surroundings and because they reflect the larger-scale temperature and pressure gradients that ultimately drive the wind.
AWS Truepower’s research has covered several published and unpublished studies on long-term wind speed trends on both regional and global scales. The company’s scientists find the results of these studies inconclusive for a number of reasons, including 1) the limited number of surface stations with long-term, continuous, homogenous wind speed records, 2) inconsistencies between regions and within similar periods of the observed wind speed trends, 3) discrepancies with reanalysis data and weather balloon observations, and 4) the fact that the heights and locations of most surface weather stations are not representative of wind energy projects.
Climatological wind studies have often been hampered by the inconsistent and unknown quality of wind data sets, unexplained discontinuities in time series, and false trends caused by land cover changes and building construction around the observation sites. Moreover, although the vast majority of long-term climate stations currently report wind speeds at 10 m height, the heights have varied over the years, from 3 m to 20 m, as have the locations. Instrumentation type (e.g. cup, vane, or sonic anemometry), reporting intervals (early reports for many stations were every 3 hours; most stations now report every hour), and recording methods (from manually read dials and strip charts to automated data loggers) have also changed.
In summary, AWS Truepower concludes that the available data do not support the finding that there is a significant global or hemispheric trend in wind speeds that may be of concern for the wind energy industry. AWS Truepower continues to monitor and analyze data and published research. In addition, the company publishes quarterly bulletins on wind conditions in North America and Europe (available at http://www.awstruepower.com/knowledge-center).
Freedman, J.M., 2008: Trends, data, and climate change. AWEA Wind Resource and
Project Energy Assessment Workshop. Portland, OR.
Klink, K. 2007: Atmospheric circulation effects on wind speed variability at turbine height. J.
Appl. Meteorol. and Clim., 46, 445-456.
Pryor, S. C., R. J. Barthelmie, D. T. Young, E. S. Takle, R. W. Arritt, D. Flory, W. J. Gutowski Jr., A. Nunes, and J. Roads 2009: Wind speed trends over the contiguous United States, J. Geophys. Res., 114, D14105, doi:10.1029/2008JD011416.
Vautard, R., Cattiaux, J., Yiou, P., Thépaut, J.-N. & Ciais, P., 2010: Nature Geoscience, 3, 756 – 761. advance online publication doi:10.1038/NGEO979 (2010)