Measurements of the Spatial Variability of Mean Wind Profiles Using Multiple Doppler Lidars over Distances less than 1 Km

Abstract:

Small differences in wind speed can translate to large differences in wind energy (WE) revenues, so WE decision making requires accurate measurements of wind profiles through the turbine rotor layer of the lower atmosphere. Advances in understanding and modeling of boundary-layer processes, also needed by WE, requires such measurements through an even deeper layer—at least the lowest few hundreds of meters. An important use for such accurate measured wind-profile data is in the initiation and verification of NWP models. This prospect raises several fundamental questions, such as, what does the modeled profile represent, how was the measured profile determined, and what if the profile had been measured from a different site within the grid cell? To address these questions, two experiments were conducted at the Boulder Atmospheric Observatory (BAO) in modestly complex terrain downwind of the mountains. The Lidar Uncertainty Measurement Experiment (LUMEX) in June-July 2014 featured 5 Doppler lidars (2 scanning), and XPIA in April-May 2015, 11 Doppler lidars, including 5 scanning systems. Two broad goals of these projects were to assess differences in scanning and other data acquisition procedures on the measurements, addressed in (Pichugina et al.) at this conference, and to evaluate the effects of varying spatial separations on differences in the measured winds, addressed in the present paper. Sonic anemometers every 50 m on the 300-m BAO tower were used as a reference for the wind calculations, as well as another profile location. Lidar scan data indicated terrain-related regions of stronger flow within the scan volume of more than 1 m/s that were at least semi-recurrent. This variability produced significant differences in mean rotor-level winds by 2 identical profiling lidars separated by 500 m. During XPIA, four of the scanning Doppler lidars performed intersecting elevation scans (vertical-slice or “RHI”) to create ‘virtual towers’ at various separation distances from less than 100 m to more than 500 m. These datasets show increasing RMS differences in winds as the profile separation distances increase.