Wind turbines harvest energy from the wind in the atmospheric boundary layer and reach heights of up to 200 m above the ground. Winds in this region of the boundary layer are driven by a complex interaction of synoptic forcing with local and upwind conditions. The National Renewable Energy Laboratory operates the National Wind Technology Center (NWTC) at one such site near Boulder, Colorado; winds at the NWTC are known to follow a seasonal pattern, changing from higher speed westerly flow in winter to weaker summer winds from the North and South.

In this presentation we present a method to automatically and objectively classify wind speed and direction at the NWTC using k-means clustering. This method uses 14 years of 10-minute average wind speeds and directions measured at different heights on an 80-m meteorological tower to reveal the four dominant wind phenomena at this tower. The observed wind phenomena all exhibit different seasonal patterns and are seen at all heights on the tower. A comparison of frequency data for each of the phenomena with several climate indices including the Arctic Oscillation and Multivariate ENSO Index shows different degrees of correlation between the frequency and direction of winds at this location with regional climate variability. The results for this site are consistent with other studies of the impact of climate variability in the Rocky Mountains, and also agree qualitatively with studies of wind characteristics at other sites in the North American prairies.

Our presentation will illustrate the clustering method and the relationships between the winds at this site and several climate indices. We will also suggest how these relationships can provide a foundation for quantifying future wind variability at this site and others.