Quantifying Lidar Turbulence Error Sources with an Uncertainty Framework

Doppler wind lidars (DLs) are quickly becoming a practical alternative to meteorological towers in the wind energy industry. DLs have been used for wake and wind resource assessment studies for several years and are currently gaining acceptance for use in power performance testing. While DLs generally measure the same average wind speeds as cup anemometers on a tower, several factors cause DLs to measure different values of turbulence. These factors have different magnitudes depending on atmospheric stability and measurement height and cannot be easily separated in real-world measurements. As turbulence is an extremely important parameter for both turbine loads and power production, it is essential to characterize and quantify sources of lidar turbulence error in order for lidars to gain wider acceptance in the wind industry.

In this work, an uncertainty framework for ground-based DLs is presented. The framework relates sources of lidar turbulence error, such as instrument noise and variance contamination, to physical characteristics of the atmosphere, such as aerosol concentration and spatial heterogeneity. Once sources of error are identified, models can be developed to remove these errors from the lidar-estimated turbulence. The key characteristic of these error models is that they are derived from atmospheric parameters that can be measured by a DL. Thus, rather than using a climatologically averaged model to reduce lidar turbulence error, the models adapt with current flow conditions and can be utilized under different site conditions.

In order to quantify the various error sources at different measurement heights and under different stability conditions, a set of large-eddy simulations (LES) with a virtual lidar tool is utilized. Results from the LES give insight to the most significant error sources under different atmospheric conditions. In addition, the LES is used to develop and test models for uncertainty reduction, leading to further refinement of the lidar uncertainty framework.