Anemometers: Cups, Vanes, Sonics and Lidar

Cup anemometers measure wind speed via rotation rate of hemispherical cups around a vertical axis. The rotational speed is proportional to wind speed after calibration. These instruments exhibit mechanical inertia, meaning they under-respond to rapid gusts. The anemometer’s distance constant (d₆₃) is the distance a wind must travel to elicit 63% of the final response; for a typical cup anemometer, d₆₃ ≈ 2.5 metres. At 10 m/s, the response time is approximately 0.25 seconds. Vanes, mounted alongside, measure direction using tail-fin alignment into the wind, damped to prevent oscillation. Directional accuracy is typically ±3° under steady conditions. However, turbulence or rapid shifts degrade performance. In Ireland, many Met Éireann synoptic stations use Thies or Orban cup/vane units. These are sited at 10 m AGL over level terrain, per WMO guidelines. The 10-minute average reported in public observations smooths out transient errors, but peak gusts are sampled at 3-second intervals, introducing a known low bias of 5–10% compared to true peak due to inertia. For example, a gust rising from 10 m/s to 25 m/s in 1 second may be recorded as 22–23 m/s. This underestimation is consistent across exposed coastal sites such as Malin Head, where squalls from Atlantic depressions frequently exceed 20 m/s. The Wind Agent’s Agreement Spine compares such observations with model ensembles to assess data reliability. Cup anemometers require regular maintenance: bearing wear increases starting threshold and reduces calibration accuracy. Annual recalibration is recommended, traceable to national standards.

Ultrasonic anemometers measure wind by transmitting high-frequency sound pulses between paired transducers. The time-of-flight difference between opposing paths is proportional to the wind component along that axis. A three-axis setup resolves u, v, w vectors, enabling 3D wind measurement at high frequency (10–64 Hz). This allows direct computation of turbulence statistics, such as gust factor or friction velocity. Unlike cup anemometers, sonics have no moving parts and respond instantaneously, with a starting threshold near 0 m/s. However, they are sensitive to precipitation, insect buildup, and thermal gradients. In heavy rain, water droplets scatter sound waves, causing spurious readings. For example, a 20 m/s horizontal wind measured with a Gill R3 sonic may show 2–3 m/s vertical contamination during drizzle. Temperature compensation is critical: uncorrected, a 10°C gradient across the path introduces ~0.5 m/s error. In Ireland, sonic anemometers are used at research sites like Valentia Observatory for flux studies and at wind farms for shear profiling. Their output feeds into The Wind Agent’s Agreement Spine, where high-frequency agreement across sensors increases confidence in exceedance forecasts. At Mace Head, sonic data are used to validate boundary layer models. The 3D resolution allows detection of low-level jets and shear layers not visible in cup data. Data gaps occur during prolonged wetting, requiring quality control flags. Sonic anemometers must be levelled within 1° to avoid vector projection errors.

Pitot tube anemometers measure dynamic pressure via a differential pressure sensor between a forward-facing port and static ports. The wind speed is derived from Bernoulli’s equation: $ v = \sqrt{2 \Delta P / \rho} $, where $ \Delta P $ is dynamic pressure and $ \rho $ is air density. At 15°C and sea level, $ \rho ≈ 1.225 \, \text{kg/m}^3 $. A $ \Delta P $ of 200 Pa corresponds to $ v = \sqrt{2 \times 200 / 1.225} ≈ 18.0 \, \text{m/s} $. These are common in aviation and racing but rare in meteorology due to sensitivity to angle of attack and icing. Propeller anemometers rotate a helical blade into the wind, with rotation rate proportional to speed. They are often paired with a vane for directional alignment. Their starting threshold is higher than sonic types—typically 0.8–1.2 m/s—due to bearing friction. A common model, the R.M. Young 05103, has a rotor time constant of 1.5 seconds, meaning it takes 1.5 seconds to reach 63% of final speed after a step change. This lag is significant in gusty maritime conditions, such as those off the west coast during Atlantic fronts. Propellers also suffer from overspeeding in turbulent flow, where eddies induce artificial rotation, leading to overestimation by 5–15% in complex terrain. In Ireland, propeller types are occasionally used on offshore buoys but are not standard at Met Éireann land stations. Their use is limited by exposure-related drift and maintenance demands.

Lidar (Light Detection and Ranging) measures wind by emitting pulsed laser beams and detecting Doppler shifts in backscattered light from aerosols. Vertical profiling lidars, such as the Leosphere WLS70, scan in multiple directions to reconstruct 3D wind vectors at heights from 40 m to 200 m. A typical scan cycle (e.g., 30° elevation, 4 cardinal directions) yields 10-minute average profiles at 10–50 m resolution. The horizontal wind speed is computed from line-of-sight velocities using velocity-azimuth display (VAD) processing. Lidar has no moving parts and avoids tower interference, but requires sufficient aerosol loading; performance degrades in very clean air. In Ireland, lidar units are deployed temporarily at wind farm sites for resource assessment. Sodar (Sonic Detection and Ranging) uses sound waves instead of light, emitting acoustic pulses and measuring return from temperature fluctuations in the air. Range is shorter—typically 200 m—and sensitive to ambient noise. Both systems provide height-matched data critical for The Wind Agent’s Shear Glass, which displays wind at 10, 80, 120, and 180 m. For example, during a nocturnal low-level jet event, lidar may detect 18 m/s at 100 m while surface cups show 8 m/s—information vital for crane or drone operations. Data from these remote sensors are integrated into the exceedance fan to assess probability of threshold breach at specific elevations.

All anemometers have a minimum wind speed required to initiate movement or detection, known as the starting threshold. For cup anemometers, this is commonly cited by WMO as 0.4 m/s; sonic types detect from 0 m/s. Below threshold, wind is recorded as calm even if air is moving. This affects frequency of calm hour reporting, particularly in sheltered inland areas. Over-speeding occurs when an instrument registers higher than true wind due to turbulence or mechanical resonance. Cup anemometers may over-read by 1–3% in turbulent flow, while propeller types can exceed by up to 15% under gusty conditions. This is distinct from under-response in gusts—both errors exist simultaneously. For instance, during Storm Éowyn (24 Jan 2025), a coastal station with a worn cup anemometer recorded a 3-second gust of 38 m/s, but nearby sonic data suggested a true peak near 41 m/s, with the cup under-responding to the peak but over-reading the mean due to turbulence. The Wind Agent’s evidence records log such discrepancies using the Agreement Spine, which weights sensor types by known error profiles. Thresholds and response characteristics must be considered when setting operational limits—never assume observed values are exact.

Exposure to ice, salt spray, and dust degrades anemometer performance. Ice accumulation on cup arms or sonic transducers adds mass or blocks paths, causing under-reading or complete failure. Heated cup anemometers reduce ice risk but may alter thermal response. At Valentia Observatory, unheated sensors have recorded zero wind during freezing fog when actual speeds exceeded 5 m/s. Salt corrosion affects bearings and electronics, particularly on west-facing coasts. Malin Head reports annual maintenance cycles to replace vane bearings due to salt wear. Insects, especially spiders, build webs between sonic transducers, scattering signals. A web spanning 30% of the path can introduce 1–2 m/s error. Regular cleaning is essential. Misalignment—such as a vane offset by 10°—causes persistent directional bias. A 10° error in 20 m/s wind introduces a 3.5 m/s error in crosswind component, critical for operations with directional thresholds. The Wind Agent’s fleet board flags stations with known icing events or maintenance logs. During Storm Ophelia (16 Oct 2017), several stations reported implausible calm periods due to ice blockage—later confirmed by maintenance records. Always consult station metadata when interpreting observations.

Anemometer calibration ensures output matches true wind speed within known uncertainty. Primary standards use wind tunnels with reference Pitot tubes traceable to national laboratories. The WMO recommends recalibration every two years for operational stations. Field calibration is impractical; units are removed and tested in controlled flow. A typical cup anemometer has a calibration equation: $ V = a \cdot f + b $, where $ f $ is frequency (Hz), and $ a $, $ b $ are coefficients. For a Thies model, $ a ≈ 0.22 \, \text{m/s per Hz} $, $ b ≈ -0.1 \, \text{m/s} $. A rotation rate of 50 Hz corresponds to $ V = 0.22 \times 50 - 0.1 = 10.9 \, \text{m/s} $. Uncertainty is typically ±2% above 5 m/s. Sonic anemometers are factory-calibrated and not user-adjustable. Calibration drift increases with bearing wear—up to 5% over three years in harsh environments. In Ireland, Met Éireann maintains a calibration programme for its network. The Wind Agent uses calibration status in its Agreement Spine to weight sensor reliability. Uncalibrated sensors are flagged in alerts. Traceability to international standards ensures consistency across time and space, essential for climate monitoring using ERA5 reanalysis.

Public wind observations in Ireland derive primarily from Met Éireann’s synoptic network, using cup and vane anemometers at 10 m height. These feed national reports and international exchange via WMO. Research sites like Mace Head and Valentia use sonic anemometers for flux and boundary layer studies. Offshore buoys, such as the M4 buoy west of Clare, use heated sonic or propeller types. Temporary lidar campaigns support wind energy projects. The Wind Agent integrates data from all these sources, distinguishing between measured (observed) and modelled (forecast) values. Observations are used to validate HARMONIE-AROME and ECMWF models. The exceedance fan uses ensemble forecasts, while the Shear Glass blends observed profiles with modelled vertical structure. For example, during Storm Eunice (18 Feb 2022), observed gusts at 39 m/s at Roches Point were consistent across cup, sonic, and model data, increasing confidence in the exceedance alert. The live Ireland map shows station types and data status. Users must refer to their own operational documents for decision thresholds—The Wind Agent provides evidence, not authority.