Correction of motion-induced and intrinsic averaging errors in turbulence measurements by floating Doppler wind lidars

Offshore wind energy has grown rapidly in recent decades due to the stronger and more uniform winds available at sea, but despite significant cost reductions it remains one of the most expensive renewable energy sources, prompting the industry to prioritize cost-effective wind resource assessment [1]. Floating Doppler Wind Lidars (FDWLs) have become the standard tool for this purpose, as they can measure 10-minute mean horizontal wind speed and wind direction with high accuracy compared to reference anemometers. However, FDWLs face challenges in accurately measuring wind turbulence, a critical parameter for turbine design and control, because of two opposing error sources: on one hand, the spatial and temporal averaging inherent to lidar measurements, which underestimates turbulence, and, on the other hand, wave-induced platform motion, which introduces apparent turbulence and leads to overestimation [2].

Recently, Salcedo-Bosch et al. [3] presented a novel methodology to compensate for both sources of FDWL turbulence measurement error using measurement simulations over Mann-model-generated three-dimensional turbulence boxes. The methodology simulates measurements from a FDWL and an ideal sonic anemometer over turbulent wind fields generated to match the experienced atmospheric state by selecting appropriate Mann model parameters [4]: turbulence length scale (LMM), eddy lifetime parameter (Γ), and turbulent energy dissipation rate (ae^2/3) [3]. By comparing the turbulence estimates from the two instruments, a correction scaling factor R is derived and used to compensate FDWL measurement errors.

In this work, we assess the FDWL turbulence compensation method over a three-month period using data from the IJmuiden campaign in the North Sea, where a FDWL was deployed alongside a meteorological mast equipped with anemometers at multiple heights serving as reference measurements. The results show that the compensation method effectively corrects the error sources at all measurement heights (25 m, 56 m, and 87 m a.s.l.), with FDWL turbulence measurements closely matching those of the anemometers, achieving R² > 0.85, RMSE < 0.13 m/s (a 30% improvement), and mean bias < 0.02 m/s (an 80% improvement) compared to uncorrected measurements.

REFERENCES

[1]  M. Taylor, P. Ralon, and S. Al-Zoghoul, “Renewable power generation costs in 2021,” Int. Renew. Energy Agency IRENA, Abu Dhabi, UAE, Tech. Rep., 2022.

[2] A. Peña, G. G. Yankova, and V. Mallini, “On the lidar-turbulence paradox and possible countermeasures,” Wind Energy Science, vol. 10, no. 1, pp. 83–102, 2025.

[3] Salcedo, A.; Rocadenbosch, F.; Peña, A.; Mann, J.; Lolli, S. “Understanding the impact of turbulence on floating lidar measurements.” IEEE transactions on geoscience and remote sensing, 2025, vol. 63, article 5704014.

[4] Jakob Mann, “Wind field simulation,” Probabilistic Engineering Mechanics, vol. 13, no. 4, pp. 269–282, 1998.

ACKNOWLEDGEMENTS

This research is part of the project PID2024-155592OB-C21, funded by MInisterio de Ciencia, Innovación y Universidades (MICIU)/Agencia Estatal de Investigación (AEI)/10.13039/501100011033 and ERDF/EU