the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 03 Apr 2025
Evaluating the Impact of Motion Compensation on Turbulence Intensity Measurements from Continuous-Wave and Pulsed Floating Lidars
Abstract. Floating Lidar Systems (FLS) play a crucial role in offshore wind resource assessment, offering a cost-effective and flexible alternative to traditional meteorological masts. While wind speed and direction measurements from FLS demonstrate high accuracy without further in-depth correction required, platform motions introduce systematic overestimation of turbulence intensity (TI). This motion-induced bias requires compensation techniques to ensure reliable TI measurements. This study evaluates the impact of a deterministic motion compensation algorithm on TI measurements from two FLS of the same type, equipped with different lidar types: a continuous-wave (cw) lidar and a pulsed lidar. The analysis compares raw and motion-compensated TI data against reference measurements from a fixed cw lidar and a met mast cup anemometer.
A comprehensive evaluation is conducted using multiple performance metrics, including Regression Analysis, Mean Bias Error (MBE), Mean Relative Bias Error (MRBE), Root Mean Square Error (RMSE), Relative Root Mean Square Error (RRMSE), Representative TI Error, and Quantile-based distribution analysis. The results show that the applied motion compensation significantly reduces the overestimation of TI, with the pulsed lidar exhibiting the most substantial relative improvement across various metrics. The cw lidar, while also benefiting from motion compensation, demonstrates a closer alignment with the fixed lidar in terms of absolute bias reduction.
Despite these improvements, residual discrepancies remain, attributed to differences in measurement principles, remaining motion effects, lidar-specific characteristics and sensitivities. Our findings confirm that deterministic motion compensation can enhance the reliability of FLS-derived TI measurements, bringing them closer to those obtained from a fixed lidar system. Future work should focus on refining compensation algorithms by incorporating lidar-specific sensitivities, improving sensor time synchronization, and exploring machine learning-based enhancements for an even better agreement with a met mast reference.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Wind Energy Science.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.- Preprint
(16974 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2025-45', Anonymous Referee #1, 10 Apr 2025
The authors evaluate the effectiveness of a deterministic motion compensation method in improving the measurement of turbulence intensity (TI) using pulsed and continuous-wave floating lidar systems (FLS). They compare TI measurements obtained from both pulsed and continuous-wave FLSs against reference measurements from anemometers and a fixed lidar installed on the FINO3 met mast at multiple heights. The accuracy of the TI measurements is assessed before and after applying the compensation method, demonstrating its effectiveness. The study employs several statistical indicators aligned with both industry and academic standards, providing a thorough analysis of the results. The findings show that the compensation method achieves best-practice accuracy in most tested scenarios, validating its utility.
However, some important aspects of the study require further development to justify its publication. First, while the introduction briefly mentions existing motion compensation methods for FLSs, the results of this study are not compared against those methods. Such a comparison would be essential to demonstrate the novelty and performance of the proposed approach. Second, the deterministic motion compensation method is not described in sufficient detail, which limits the reproducibility of the results. I recommend addressing these issues prior to publication. Specific comments are provided below to support improvements to the manuscript.
Major Comments
- Redundancy in Structure: The manuscript could benefit from reducing repetitive explanations, particularly in the outline sections (e.g., Lines 117–128).
- Methodology Motivation: The paper lacks a clear motivation and novelty statement regarding the proposed motion compensation method. It is not evident how this approach differs from or improves upon existing techniques.
- Description of Lidar Retrieval Methods: The VAD (Velocity Azimuth Display) and DBS (Doppler Beam Swinging) methods should be more thoroughly described. Consider including detailed explanations in an appendix if necessary.
- Motion compensation method definition: The manuscript should thoroughly define/formulate the motion compensation method used.
- Comparison with Existing Literature: The results should be compared to existing studies in the field, such as those by Kelberlau et al. (2020, 2023) and Gutiérrez-Antuñano et al. (2018), which report statistical indicators with similar magnitudes. Such comparisons are essential to contextualize the study’s contributions.
Minor Comments
Sect. 2.
- Figure 1. Typically IMUs use an inertial reference system with respect to North-East-Down (or different). Please indicate those in the figure.
Sect. 3.
- Why only use wind speed data within the range 4-16 m/s?
- A comparison between mean horizontal wind speeds measured by the FLSs and the anemometers should be provided.
- Lines 315-319: This is not expected. Typically, CW lidars measure lower turbulence values than anemometers due to their inherent temporal and spatial averaging. For instance, see https://wes.copernicus.org/articles/10/83/2025/wes-10-83-2025.html and https://journals.ametsoc.org/view/journals/atot/28/7/jtech-d-10-05004_1.xml . Please, comment on that.
- Lines 321-322: Please comment on how the underlying measurement principles of the instruments generate the discrepancies. Do they cause the CW lidar TI over-estimation as well?
Citation: https://doi.org/10.5194/wes-2025-45-RC1 -
RC2: 'Comment on wes-2025-45', Anonymous Referee #2, 05 May 2025
General comment
The manuscript entitled “Evaluating the Impact of Motion Compensation on Turbulence Intensity Measurements from Continuous-Wave and Pulsed floating Lidars” presents a study of turbulence intensity estimations based on floating wind lidars. The topic of the study is interesting, since how to estimate offshore turbulence is still an open question in the wind energy research community.
However, even though the article is in general well written and structured, it is difficult to understand what the novelty in this manuscript is.
Overall,
- More details are needed regarding the floating lidar motion compensation. The article from Wolken-Mohlmann is referenced, but at least a brief description of the method is required. Furthermore, it is not explained clearly what the added value of the applied method is in comparison with already published motion correction methods. In the lines 78 – 80 it is written that the method is adopted “because of the transparency, robustness, as well as versatility of a physics-based correction model”. It is hinted this way that there is a lack of those features in the already applied approaches. I think that this part should be discussed more in the introduction since it will explain what the novelty of this study is.
- In the manuscript a series of different statistical parameters are estimated and the performance of the two wind lidars is assessed in relation those. The authors are very thorough in explaining the reasoning of testing those parameters. However, there is no conclusion regarding the result. For example, are the results good enough for a wind resource assessment? And what is the improvement in comparison to results of already published studies? In the discussion part it is stated that “some sources of residual error remain”. How important are those? and what is the further improvement that can be achieved for using machine leaning methods given the theoretical limitations of measuring atmospheric turbulence using vertical profiling wind lidars?
- Regarding the errors in the estimated TI values from the floating wind lidar, was it possible to correlate those with the significant height or frequency of the waves and/or to atmospheric stability?
- I suggest adding in an appendix or at least mention in the manuscript statistical results of the comparison of the mean wind speed estimations between the non-corrected / corrected and the reference sensors.
- The analysis of this study focusses on two specific lidar models, not necessarily two wind lidar techniques (i.e. cw and pulsed). For example, two different pulsed lidars could have different performance due to variations among others in the sampling rate, scanning angles, or optical components. I suggest clarifying in the manuscript that the results presented are relative to the ZX300 and the Windcube wind lidars.
Specific comments
- Line 42. I suggest reformulating the beginning of the sentence to clarify that the statement concerns the mean wind speed and direction.
- Line 46. I suggest replacing the word “fluctuations” with “errors”
- Line 62. It is written “However, they may struggle when faced with complex, non-linear motions … ” can you please elaborate more about what is meant with this statement?
- Line 92. What is the context of the thresholds proposed by Kelberlau et al 2023? Are they relevant to wind resource assessment?
- Line 149. I suggest removing the word “virtual” and add that both the VAD and the DBS methods assume a horizontal homogeneity of the wind conditions.
- Please add information of the heights where cup and sonic anemometers are installed (or refer to Table 2).
- Line 181. The deterministic motion compensation algorithm used in this study considers the line-of-sight velocities. How different is this method compared to the method applied by Yamaguchi and Ishihara 2016, Keleberlau et al 2020?
- Line 185. The motion of the FLS does not distort the scanning geometry but rather rotates it in respect to the measured air volume.
- Lines 195 – 200. To mitigate errors associated with time offsets between the lidar and the motion sensor a methodology is proposed according to which as a time offset is selected the one that results in the smallest standard deviation of the 10-minute wind speed. If the time offsets are due to a drift between the two data acquisition systems, then I would guess that the time offset is increasing/decreasing linearly. Is this something that can be observed from the results of this analysis?
- Table 2. The measurement heights of the wind lidar are measured from the sea level?
- Page 17. Figure 6. I think that it would enhance the understanding of the results presented in this manuscript if this figure presented the statistics of the wind and wave conditions of the selected data set and not of the whole set of the acquired data.
- Line 303. What is the reason of presenting the results of the analysis at 71 m and 107 m at the appendix if there are not discussed in the main part of the manuscript?
- Line 315. According to the results presented in Fig 7 the TI estimation of the “fixed ZX wind lidar” measures higher TI than the mast. How is this explained? Did the authors apply a quality check on the lidar data?
- Line 319. Can you please elaborate more here regarding the several factors that have an impact on the regression analysis.
- Lines 347 – 354. Here it is written that in the case of the cw lidar the homodyne detection is a limitation that introduces errors in the case of the low wind speeds. However, in all the results that are presented the cw wind lidar is performing better than the pulsed wind lidar.
- Line 350. The homodyne detection is not the only way to detect the Doppler shift in cw wind lidars. Please reformulate to clarify that you refer to the ZX wind lidar.
- Lines 391 – 392. Please add a reference of the minimum and best practice performance thresholds.
- Lines 404 – 405. What is the direction of the booms of the cup anemometers on the mast?
- Lines 483 – 485. I think that the lower error that the motion-compensated floating wind lidar exhibits in comparison to the fixed is a numerical artifact rather a result that demonstrates the efficiency of the method. How is it possible that the performance of the motion correction method can provide results better than that of a fixed wind lidar?
- Line 532. It is written “The remaining scatter can be attributed to … the elevation of the fixed cw lidar”. Didn’t all wind lidars measure at the same height above the sea level?
- Lines 569 – 570. Isn’t this a repetition of the previous sentence?
Minor comments
- Throughout the manuscript there is a typo in the wind speed units. Roman fonts should be used. Please correct it.
- Line 229. Eq(3) Correct the subscript of TI.
- Line 331. Replace “indicate” with “indicates”
- Line 346. Replace “suggest” with “show”
- Line 422. Replace “betwee” with “between”
- Line 430. Replace “:” with “.”
- Line 451. Replace ”:” with “.”
- Lines 587 – 588: In the Acknowledgements it is written: “This research is part of our ongoing efforts in wind resource assessment, and we offer commercial measurement services for similar applications." Why is this relevant to the Acknowledgements?
Citation: https://doi.org/10.5194/wes-2025-45-RC2 -
EC1: 'Comment on wes-2025-45', Alfredo Peña, 08 May 2025
Dear authors,
I have closed the discussion phase of your manuscript. You can find the reviews from two different experts on your work. When responding to their comments, I think it is very important to make sure you clearly show/state the novelty of your work. Both reviewers, as well as myself, think the manuscript is interesting but cannot see the novelty on it. Right now it appears as you are demonstrating one approach to correct FLS turbulence measurements based on a previously developed method. It is not clear if this approach is different from others, or better. So there is a need to explain the difference between this and other approaches. Again, please address concisely and precisely the novelty of your work.
Although you probably need to add some more detailed description about the method you use (and its relation to others), your manuscript is too lengthy. So, you also need to find out redundant explanations and/or results. Some parts perhaps can be moved to appendices without detriment of the main sections.
Regards,
Alfredo
Citation: https://doi.org/10.5194/wes-2025-45-EC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 500 | 51 | 9 | 560 | 10 | 9 |
- HTML: 500
- PDF: 51
- XML: 9
- Total: 560
- BibTeX: 10
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1