Numerous types of lidars have been employed for atmospheric science field studies; a comprehensive summary of lidar theory, and retrieval methods is given in Weitkamp (2005). Doppler wind lidars are used extensively in wind energy and will be discussed in more detail in the next section. The international wind energy industry has taken steps to develop recommended practices for the use of sodar (and other ground-based remote-sensing devices) for wind energy applications. Recommended practices for resource assessment and a summary of other relevant standards can be found in Clifton et al. (2013a).
Other types of lidars (differential absorption lidar and Raman lidar) may be used to measure the concentration of atmospheric gases, including water vapor and ozone. The differential absorption lidar uses absorption, as evidenced by reduced backscatter from greater distances, to measure these concentration of atmospheric gases (Spuler et al. 2015). A Raman lidar detects atmospheric components (such as water vapor) by measuring the wavelength-shifted return from selected molecules (Strauch et al. 1971). A Raman-shifted eye-safe lidar uses very short pulses of invisible, eye-safe, near-infrared electromagnetic radiation to sense aerosol particles and their variations and movement to quantify atmospheric motions (Mayor and Spuler 2004; Dérian et al. 2015).
A pulsed Doppler lidar transmits short bursts of light (whereas radar transmits short bursts of radio waves) at short wavelengths to use aerosol particles as atmospheric scattering targets. Because aerosols are generally suspended in atmospheric flow, they are excellent tracers of air motions. By observing the Doppler shift of the light backscattered by aerosols, the lidar can remotely measure air velocities. While Doppler lidar has been employed in atmospheric science experiments since the mid-1980s (Post and Cupp 1990; Banta et al. 1992; Grund et al. 2001), the advent of cost-effective and reliable fiber-optic lasers in the mid-2000s ushered in an era of widespread deployment of lidars for wind-profiling applications (Courtney et al. 2008).
Pulsed Doppler lidars measure profiles along the beam of the line-of-sight wind velocity. Therefore, a system pointed vertically will only provide a profile of the vertical wind velocity. To measure the horizontal wind, the lidar’s beam must be tilted slightly out of the vertical direction to allow the horizontal wind to contribute to the line-of-sight velocity. With appropriate scanning schemes, and by assuming horizontal homogeneity across the measurement volume, the three-dimensional wind vector can be inferred. Lidars can have ranges on the order of 200 m (for Windcube v1 and v2 systems) up to 20 km (for the Lockheed Martin WindTracer system), with range gates varying between 10 and 100 m.
Doppler wind lidars can operate in several different modes, as summarized in Banta et al. (2015). Two of these modes (velocity-azimuth display (VAD) and Doppler beam swinging (DBS)) are visualized in Figure 4. Both DBS and VAD scanning consider data from one elevation angle and multiple azimuth angles; DBS considers three or four azimuth angles, whereas VAD typically considers at least 20. “Profiling” lidars (commercial examples include Leosphere Windcube v1/v2) rely on a four-beam (v1) or five-beam (v2) DBS approach (as opposed to the three-beam approach shown in Figure 5). A plan-position-indicator (PPI) scan involves changing azimuth angles while keeping a constant elevation angle. At very low elevation angles and short ranges, the PPI can almost be considered a horizontal scan. A range-height-indicator (RHI) scan holds an azimuth angle constant and changes the elevation angle to provide a vertical cross section of the area of interest (the PPI and RHI terminology is derived from the use of these types of scans by radars.) By combining multiple lidars looking at the same point or points in space, multiple components of the flow can be resolved (Newsom et al. 2008; Hill et al. 2010; Carbajo Fuertes et al. 2014; Berg et al. 2015; Klein et al. 2015; Yazicioglu et al. 2016; Debnath et al. 2017a, b).
In contrast to the Doppler wind lidar method, the continuous wave lidar does not use pulses but rather continuously emits laser light through a constantly rotating prism deflected by some angle (typically 30°) from the vertical direction. The ZephIR lidar makes one complete rotation per second (s), sampling the backscattered light at a frequency of 100 megahertz (MHz). Fifty radial velocities, one for every 360°/50=7.2°, are available from each rotation. Each of the several available heights is scanned for 1 s, and the beam is then refocused to the next height in the sequence and the scanning procedure is repeated. Continuous wave lidars tend to have shorter ranges than pulsed Doppler lidars.
The Technical University of Denmark (DTU) SpinnerLidar is a velocity measurement device that can acquire velocity measurements at a temporal and spatial resolution that is higher than offered by most lidar measurement systems. DTU developed the SpinnerLidar to be a turbine- mounted lidar for rapid scanning of the wind field in a two-dimensional plane (Mikkelsen et al. 2013; Sjöholm et al. 2013b; Angelou and Sjöholm 2015; Machefaux et al. 2016; Churchfield et al. 2016; Herges et al. 2017). An image of the DTU SpinnerLidar installed at the Scaled Wind Farm Technology (SWiFT) facility is shown in Figure 6.
The SpinnerLidar scan head consists of two co-rotating ~15° wedge-shaped prisms integrated on a ZephIR 300 continuous-wave coherent Doppler lidar. The lidar produces laser light at a wavelength of 1,565 nm and was configured to stream-averaged Doppler spectra at a rate of about 500 measurements per second. The prisms have a fixed gear ratio with adjustable motor settings to change the duration and number of measurements per scan (motor speed fixed for each scan) (Sjöholm et al. 2013b; Angelou and Sjöholm 2015). At each focus distance, the SpinnerLidar scans the two-dimensional surface of a sphere with an approximately 30° half angle, capturing the line-of-sight velocity component at each measurement location. The lidar can also cycle through focus distances with a change in focus distance occurring in the same amount of time as a full scan. The position of the measurement relative to the symmetry axis is calculated from the instantaneous position of the two wedge-shaped prisms. Changes in the orientation of the rotation axis are accounted for using an integrated three-axis accelerometer
OpenLidar is an initiative that arose from the International Energy Agency Wind Task 32 and is intended to enable collaboration on lidar design, operation, and data processing across the worldwide wind lidar community. The OpenLidar platform is a combination of a modular lidar system architecture (Figure 8), guidelines for documenting modules and their interfaces, and a user-edited website for documentation and software (www.openlidar.net).
The OpenLidar architecture is similar to existing lidar designs. New modules developed under OpenLidar will have clear power, data, and safety requirements, and interfaces. The same modular architecture concept and documentation can be applied to existing modules or lidars. Therefore, participants in OpenLidar can customize modules or design new modules knowing that they will work with other modules designed for this architecture, or integrate modules to create a lidar for specific use cases. OpenLidar facilitates collaboration and experimentation as well as reduces the learning curve involved with developing new lidar designs.
OpenLidar follows a four-stage plan. An initial, high-level system architecture and collaboration website has been created. In the near future, developers hope that participants will start detailed design work on individual modules. The last two stages include building hardware demonstrations and the mass customization of the modules. The current goal of OpenLidar is to raise awareness of the initiative and gain a critical mass of researchers.
OpenLidar is currently being used at the University of Stuttgart to help integrate a new lidar scanner with a lidar module developed by the University of Oldenburg. This collaboration forms part of the Application Oriented Wind Field Research and Measurements for Wind Turbines (ANWIND) project to develop and deploy a rugged, scanning lidar in the offshore environment.
Lidars are generally considered to have a TRL of 8.