Development and Experimental Validation of a Reference-type Coherent Doppler Wind Lidar for Meteorological Operations

Bu, Zhichao; Dai, Yaru; Huang, Yuanyuan; Chen, Yubao; Li, Zhigang; Jin, Guohua; Chen, Xuliang

doi:10.5194/egusphere-2026-3088

Preprints

https://doi.org/10.5194/egusphere-2026-3088

Preprints

03 Aug 2026

| 03 Aug 2026

Development and Experimental Validation of a Reference-type Coherent Doppler Wind Lidar for Meteorological Operations

Zhichao Bu, Yaru Dai, Yuanyuan Huang, Yubao Chen, Zhigang Li, Guohua Jin, and Xuliang Chen

Abstract. In response to the urgent need for calibration technology in the application of wind lidar in the meteorological operations, this paper describes the development of a Reference-type Coherent Doppler Wind Lidar(RCDWL) for calibration. Based on a dual-channel composite detection mechanism combining continuous and pulsed modes, the RCDWL achieves key technical specifications including a low blind zone of 10 m, a high resolution of 10 m, and a maximum detection altitude of 3,000 m. To verify the performance and stability of the RCDWL, comparison experiments and data analyses were conducted using a radiosonde and a meteorological gradient tower, respectively. Comparison results with the radiosonde showed a wind speed systematic error of -0.05 m/s and a standard deviation of 1.11 m/s, while the wind direction systematic error was 1.34° with a standard deviation of 12.05°; the correlation coefficients were all higher than 0.97.Comparison results with the gradient tower showed a systematic error of 0.01 m/s and a standard deviation of 0.28 m/s for wind speed, and a systematic error of -0.09° and a standard deviation of 2.90° for wind direction, with correlation coefficients all exceeding 0.98. Based on these results, we used RCDWL as a reference source for a joint comparison and analysis of 9 lidars in meteorological operations. The results showed that the correlation coefficients for horizontal wind speed were greater than 0.98, with standard deviations less than 0.8 m/s; for wind direction, the correlation coefficients were greater than 0.97, with standard deviations less than 8°. Additionally, when using RCDWL as a reference standard source for the comparison and calibration of lidars prior to shipment, RCDWL effectively improves the product accuracy of lidars prior to shipment. These results fully validate the detection capabilities and stability of the RCDWL, establishing a stable and reliable reference benchmark for lidar calibration within the meteorological operations. The RCDWL fills a technical gap in wind lidar calibration reference sources and can be used for the comparison and calibration of Lidar equipment, holding great practical significance for enhancing the application level of wind lidars in various fields.

Received: 29 May 2026 – Discussion started: 03 Aug 2026

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 paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Zhichao Bu, Yaru Dai, Yuanyuan Huang, Yubao Chen, Zhigang Li, Guohua Jin, and Xuliang Chen

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-3088', Anonymous Referee #1, 16 Aug 2026

This paper describes the development of a reference-type coherent Doppler wind lidar for calibration of meteorological operations. This lidar system have a dual-channel composite detection mechanism combining continuous and pulsed modes. Then, the comparisons analysis are conducted using the radiosonde and meteorological gradient tower, respectively. In general, they obtained very good agreement. These suggest that the RCDWL have good performance with high detection accuracy, stability, and reliability. However, some major revisions are necessary before potential publication.
Major comments：
1, In this paper, the developed coherent Doppler lidar is defined as a reference-type lidar. However, in the introduction to the background and lidar technology, the question arises as to why this specific lidar qualifies as a reference-type lidar, whereas other wind lidars do not, despite the fact that all of them can similarly be compared and validated against gradient tower and radiosonde measurements.
2, In section 3.2, “a total of 87 valid synchronous comparison…”, so, is the number the effective groups after excluding adverse weather conditions? And what is the spatial resolution of radiosonde observation? How do the authors match the observations from lidar and radiosonde?
3, Is the coherent wind lidar used for the radiosonde comparison not the same as the one used for the meteorological gradient tower comparison? If that is the case, the lidar notations in Figures 6, 8, 9, and 10 should be made consistent.
4, The validation of the wind speed and wind direction retrieval is incomplete based on radiosonde observation. The main evidence is correlation coefficient, standard deviation, and systematic error. The vertical profiles of these validation should be added.
5, In figure 13, how to understand the wind direction values which are greater than 360 degree?
6, In table 2, are these results from the comparison between nine lidars and radiosonde? Or RCDWL?

Specific comments:
1, What do VCM and VAD stand for? Their full names should be given, accompanied by the relevant references..
2, The formatting of the references is not uniform. For instance, both round brackets (parentheses) and square brackets appear in the introduction.
3, line 135：“…1.54 V(V＞10m/s)…”.
4, In Figure 4 and Figure 5, the unit of color bar should be given.
5, In Figure 10a, the annotation reads "100 m." Please verify this

Citation: https://doi.org/10.5194/egusphere-2026-3088-RC1
- AC1: 'Reply on RC1', Zhigang Li, 26 Sep 2026
  
  We would like to express our sincere gratitude for your valuable comments on this manuscript. We acknowledge that there were certain shortcomings in both the language and the presentation of our findings. We have carefully considered your suggestions, provided a detailed point-by-point response, and revised the relevant sections of the manuscript in accordance with your comments and recommendations.
  
  Q1: The term “reference lidar” does not denote a lidar with a specific or unique function. In China, an increasing number of wind lidars are being deployed in operational meteorological observation networks. To effectively validate and calibrate these instruments prior to deployment, while satisfying the accuracy requirements of operational meteorological applications, we developed a wind lidar characterized by a low blind zone and high measurement accuracy. Its performance was comprehensively verified through extensive intercomparison experiments. As this instrument provides the reference measurements for simultaneous comparisons with the lidars under test, it is termed a “reference lidar” for meteorological operations. In the manuscript, we have provided a more precise description of this ambiguous term.
  
  Q2: The simultaneous comparison with the radiosonde observations was conducted over a four-month period. The 87 valid profiles refer to the observations retained after excluding cases affected by rainfall, dense fog, and other adverse weather conditions. The vertical sampling interval of the radiosonde was not fixed but was generally on the order of tens of meters.
  For temporal matching, the lidar profile acquired closest to the radiosonde launch time was selected. The lidar had a temporal resolution of 10 min and a vertical resolution of 60 m. For vertical matching, the height of each lidar range gate was used as the reference, and the radiosonde wind measurement nearest to that height was identified. When the height difference was less than or equal to 60 m, the lidar and radiosonde measurements were treated as a valid matched pair and included in the comparison. Otherwise, the corresponding measurements at that height were excluded from the comparison dataset. We have added a
  relevant description in Section 3.2 of the manuscript.
  
  Q3: The coherent wind lidar used for the radiosonde comparison is the same one with the one used for the meteorological gradient tower comparison. The results presented in Figures 6, 8, 9, and 10 were obtained from experimental measurements conducted at the Shenzhen Meteorological Bureau, China. Owing to an inconsistency in the original instrument labeling, different model designations were used in these figures. We have corrected this issue and standardized the instrument label as “RCDWL” throughout the revised manuscript.
  
  Q4: In Section 4.1, we have added the vertical-profile results from one of the comparisons with the radiosonde observations, together with a corresponding description and discussion.
  
  Q5: Wind direction is a periodic circular variable, for which 0° and 360° represent the same direction. Therefore, wind direction cannot be treated directly as a linear variable when calculating differences or constructing scatter plots. A direct comparison of the original winddirection values may produce artificial discontinuities near the 0°/360° boundary. For example, if two observations are 355° and 5°, their actual directional difference is only 10°, whereas direct subtraction yields an apparent difference of 350︒. Consequently, the corresponding points appear widely separated in a scatter plot, falsely suggesting poor agreement.
  To address this issue, we applied the following boundary-correction method during data processing. When one wind-direction value was greater than 270° and the corresponding value from the other instrument was less than 90°, the two measurements were considered to straddle the 0°/360° boundary. In such cases, 360° was added to the value below 90° before calculating the difference. This procedure eliminates artificially large errors caused by the circular boundary, improves the clarity of the graphical presentation, and more accurately reflects the agreement between the two datasets. In Section 3.2 of the manuscript, we provide a brief description of this processing.
  
  Q6: In Table 3(Table2 in original manuscript) of Section 4.1, the comparison results for all nine lidars were obtained from simultaneous intercomparison measurements using the RCDWL as the reference instrument. We have clarified and expanded the corresponding description in Section 4.1 of the revised manuscript.
  
  Citation: https://doi.org/10.5194/egusphere-2026-3088-AC1
RC2:
'Comment on egusphere-2026-3088', Anonymous Referee #2, 17 Aug 2026
The authors have developed a reference-type coherent Doppler wind LiDAR (RCDWL) for calibration of meteorological LiDAR, which has strong practical value in application. The manuscript mentions a dual-channel composite detection mechanism based on a combination of continuous and pulsed modes. CDWL has achieved key technical indicators such as a low blind zone of 10 meters, high resolution of 10 meters, and a maximum detection height of 3000 meters. This job is quite interesting. And using radiosondes, meteorological gradient towers, and 9 types of LiDAR for comparative observation and analysis, the work is relatively solid. However, there is room for improvement in the writing of the manuscript and the standardization of the illustrations. Otherwise, it will lose the interest of many readers and cause confusion. I think it is necessary to make major revisions to the manuscript before publication. The specific revision suggestions are as follows:

In the abstract section: It is necessary to streamline the content and highlight the innovative work, rather than a large amount.

In the introduction section: A brief introduction was given to the development history of LiDAR and various calibration methods for LiDAR. From my perspective, I believe that the authors avoided the important and focused on the trivial, and did not provide any reference on the development process of LiDAR core technology. The implication is that in terms of the development of reference radar, what technology or methods did the author use to innovate and what problems were solved. The outline of the problem is not clear enough. Or what new methods are used in the calibration process? These did not provide clear statements. This is regrettable.

Authors need to reintroduce the development history of the core innovation section in the manuscript and add corresponding references.

From the current manuscript, it appears that there is insufficient expression of innovation and there is no clear discussion. If it is only focused on comparative verification, then I think this paper still needs further discussion.

Isn't it bad to give a conclusion directly at the end of the background introduction? This kind of writing is puzzling. Authors need to improve their writing skills.

In section 2 of the manuscript：Figure 1 (a) is not clear, the font is small, and the quality is poor. The schematic diagram in Figure 2 and the flowchart in Figure 3 look terrible, but from a normative perspective, there is a lot of room for improvement.

What is the purpose of placing the workflow diagram in Figure 3 here? Does it reflect your core skills and methods?

In the description of section 2, the process of data processing is provided, it was not seen what methods or techniques can be used to adjust or reduce errors when the speed accuracy and error are large. This further raises the question of whether this mechanism of LiDAR can achieve high accuracy?

The simulations in Figures 4 and 5 do not have specific descriptions, making it feel like they suddenly ended, which is terrible.

If possible, try to make the tables in the manuscript into three line tables to increase readability.

In section 3.2, the comparison results are very beautiful. If there are any errors or inaccuracies during the observation process, I would like to know how to adjust them? Why not display some observation results compared to altitude? Or statistical analysis of errors with altitude? At which altitudes is it possible to have significant errors and what are the reasons?

The authors' rough handling of wind directions exceeding 360 ° in Figure 6 (b) is shocking, please correct it.

Figures 8, 9, and 10 have poor quality. Please make changes to improve readability.

In section 3.3, the author only provided data for two points (10 meters and 200 meters) when analyzing the comparison between LiDAR and meteorological tower sensors. What about the rest? I think when analyzing (Figures 9 and 10), more data analysis can be provided to increase the reliability of the evidence.

What is 66 operational LiDARs (compared to the other 66 LiDARs?) in Figure 11 and the corresponding narrative section of the manuscript? What is its function? What do you want to express here?

In the fourth section, Figure 12 presents a comparative presentation on December 18, 2023, which was not mentioned in the manuscript narrative section and is quite confusing. I think the authors did not take the manuscript writing seriously and it was very rough.

The authors did not provide latitude and longitude coordinates for the comparison locations of the experiments, which is confusing.

The authors' rough handling of data beyond 360 ° in Figures 13 (b) and 13 (d) is shocking. Please correct it.

The conclusion needs to be rephrased based on the revised content.

The writing and narration of the entire manuscript are poor and require overall improvement.
Citation: https://doi.org/10.5194/egusphere-2026-3088-RC2
- AC2: 'Reply on RC2', Zhigang Li, 26 Sep 2026
  
  We would like to express our sincere gratitude for your valuable comments on this manuscript. We acknowledge that there were certain shortcomings in both the language and the presentation of our findings. We have carefully considered your suggestions, provided a detailed point-by-point response, and revised the relevant sections of the manuscript in accordance with your comments and recommendations.
  
  Q1: In the abstract, we have streamlined the general background description and added a discussion of the key technologies employed in the reference coherent Doppler wind lidar, including its low-blind-zone design and the integration of two detection modes. We have also incorporated the analysis results obtained from multiple comparative experiments.
  
  Q2: In the revised literature review, we have placed greater emphasis on the core requirements of the self-developed lidar system, namely a low blind zone, high resolution, and high accuracy. The system employs continuous-wave/pulsed composite parallel detection technology, which effectively addresses the engineering challenge faced by single-antenna systems in balancing the near-field blind zone against far-field resolution [Zhu et al., 2012; Zhao et al., 2024]. Multi-beam spatially adaptive scanning and short-time adaptive weighted inversion algorithms are also employed to mitigate the effects of atmospheric turbulence and scanning-attitude errors on retrieval accuracy [Sathe et al., 2015; Shan et al., 2023]. Relevant references have also been added, and the corresponding revisions are marked in the manuscript.
  Regarding the system-accuracy calibration method, we adopted a conventional wind-lidar calibration approach using radiosonde and meteorological-tower measurements as reference data. Long-term synchronous observations were conducted, and a joint intercomparison dataset covering multiple altitude levels was established to evaluate wind-field measurement errors at different heights. On this basis, the overall calibration and performance of the wind lidar system were validated. The relevant description was revised based on the simulations presented in Section 2.3. Further details are provided in the final paragraph of Section 2.3.
  
  Q3: We have reorganized the sections describing the principal innovations of the manuscript, supplemented the discussion of the relevant technological developments, and added references to support the corresponding statements. The principal innovations include continuous-wave/pulsed composite parallel detection technology, systematic optimization of the blind zone of the pulsed coherent Doppler wind lidar through the numerical design of the optical receiver architecture, and high-accuracy wind-field retrieval technology. Figures 2 and 3 have also been revised to improve the clarity and readability of the technical details. Further information is provided in Sections 2.2 and 2.3 of the revised manuscript.
  
  Q4: We sincerely appreciate the reviewer’s constructive comments. We acknowledge that the original manuscript did not sufficiently identify or discuss the key innovations. In the revised manuscript, we have restructured the presentation of the innovations, strengthened the discussion of the technical principles and system design in the Introduction and system-development sections, and added relevant references to support the technical statements. We have also clearly distinguished the lidar development technologies from the field-validation experiments. In addition, further field-validation results have been included to verify the key technologies. For example, Section 3.3 now presents and discusses comparisons with meteorological-tower measurements at heights of 10 and 20 m, thereby providing experimental evidence supporting the effectiveness of the low-blind-zone design.
  
  Q5: In the original manuscript, the main findings were presented directly at the end of the background section, resulting in an abrupt logical transition that could cause confusion for readers. We have therefore rewritten the relevant paragraphs of the Introduction to improve the logical flow and readability.
  
  Q6: Figure 1(a) has been reprocessed to improve its resolution, enlarge the text, and enhance its overall clarity. The schematic diagram in Figure 2 and the flowchart in Figure 3 have been redrawn, with their line styles and layouts optimized to ensure consistent graphical formatting and improve overall figure quality. The revised figures have been incorporated into the updated manuscript.
  
  Q7: Figure 3 is intended to provide an intuitive overview of the complete algorithmic workflow of the RCDWL system, from raw radial-velocity data acquisition to the final wind-field retrieval output. It enables readers to readily understand the step-by-step implementation of the multi-beam spatially adaptive scanning and short-time adaptive weighted inversion algorithms. In response to the reviewer’s comment, we have further optimized Figure 3 to improve its clarity and readability.
  
  Q8: The original version of Section 2 lacked a description of the error-mitigation procedures and therefore did not adequately explain the measures implemented to ensure high-accuracy measurements. In the revised manuscript, we have added a description of the data quality-control procedures to Section 2. During the preprocessing of the intercomparison data, samples collected under abnormal weather conditions, such as precipitation and dense fog, are first excluded. The 3σ gross-error rejection method is then applied to remove outliers in accordance with the Chinese national standard GB/T 4883-2008. These procedures reduce the influence of anomalous data on the wind-speed comparison results and improve the reliability of the intercomparison. The corresponding revisions are provided in Section 2 of the revised manuscript.
  Reference:“ Standardization Administration of China: Statistical interpretation of data — Detection and treatment of outliers in the sample from normal distribution, GB/T 4883-2008, China Standards Press, Beijing, China, 2008.”
  
  Q9: The description of the simulation process corresponding to Figures 4 and 5 in the original manuscript was relatively brief, lacking an explanation of the corresponding inversion strategy behind the simulation, and the simulation logic was therefore incomplete. We have supplemented the corresponding sections of the manuscript with the physical background and algorithmic details of the simulation. We have also added a transitional discussion between the simulation results and the long-term joint field comparison, making the overall logic of the manuscript more coherent.
  
  Q10: We have reformatted all tables in the manuscript using the three-line table format.
  
  Q11: The comprehensive comparison experiments conducted in this study were designed to verify the overall accuracy and system performance of the self-developed RCDWL for full-altitude wind-field retrieval, with an emphasis on evaluating the overall measurement errors of the system. First, anomalous observation samples caused by adverse weather conditions, such as precipitation and dense fog, were excluded to eliminate interference from extreme meteorological conditions. Subsequently, the 3σ gross-error rejection criterion was applied to remove random outliers, thereby ensuring the validity and reliability of the comparison data. Meanwhile, due to the limited number of matched samples at individual altitudes, stratified error analysis with altitude was not performed.
  
  Q12: We apologize that our wording caused a misunderstanding for the reviewer. Wind direction is a periodic circular variable, and 0° and 360° represent the same incoming direction; therefore, it cannot be directly treated as a linear variable for difference calculation and scatter plotting. If the original wind-direction values are directly compared, an artificial discontinuity will occur near the 0°/360° boundary. For example, if the wind directions measured by two observation systems are 355° and 5°, respectively, the actual difference is only 10°, whereas direct numerical subtraction yields 350°. Consequently, the data points will be distributed in diagonally opposite regions of the scatter plot, resulting in an apparently scattered distribution. In the data processing performed in this study, if one wind-direction value is greater than 270° while the other is less than 90°, the wind direction is considered to cross the 0°/360° boundary. A value of 360° is then added to the wind-direction value below 90° . This procedure eliminates the artificially large errors caused by the boundary discontinuity and improves the visualization.In Section 3.2 of the manuscript, we provide a brief description of this processing.
  
  Q13: We have redrawn and replaced Figures 9, 10, and 11(Figs. 8 ,9 and 10 in original manuscript), improved the image resolution, optimized the axes, legends, and font labels, and enhanced the readability of the figures. Additionally, we have added more images to Figures 9 and 10 to provide more comprehensive information. Please see Section 3.3 of the manuscript for details.
  
  Q14: This study includes seven comparison heights, resulting in a total of 14 scatterplot panels for wind speed and wind direction. Given the large number of plots, the original manuscript presented only two representative heights, 10 m and 200 m, which was indeed insufficient. In response to the reviewer’s comment, the revised manuscript now includes comparison results at four additional heights in Figs. 10 and 11 (Figs. 9 and 10 in original manuscript), thereby providing a more comprehensive set of cases for analysis. Quantitative statistical results have also been added to Section 3.3 of the manuscript.
  
  Q15: In China, all models of wind lidar must undergo standardized comparative verification at the Shenzhen Meteorological Gradient Tower before being introduced into operational meteorological applications. In this study, the accuracy metrics of the self-developed RCDWL were compared with those of 66 operational wind lidars that had already been verified at this site. The box plots in Figure 12(Figure11 in original manuscript ) show the quartile distributions of the accuracy metrics of these 66 operational lidars, thereby demonstrating the advantages of the RCDWL in wind measurement accuracy. In the revised manuscript, we have added background information on the site verification and revised the corresponding main text and figure caption to eliminate ambiguity. Please refer to Section 3.3 of the manuscript for details.
  
  Q16: We sincerely apologize for this omission. The original manuscript lacked an accompanying textual explanation of the time-series comparison results shown in Figure 13(Figure12 in original manuscript ), and the details of the figure were not sufficiently described. We have reorganized the relevant paragraphs in Section 4.1 and supplemented and refined the qualitative and quantitative description of the comparision results presented in Figure 13. We have also improved the description and analysis of the results presented in Table 2.
  
  Q17: We have added the geographical coordinates of the field comparison sites to the experimental section of the manuscript, including the sites used for the radiosonde comparison, meteorological-tower comparison, and joint lidar comparison.
  
  Q18: We apologize that our wording caused a misunderstanding for the reviewer. As suggested in Comment 12, we provide a brief description of this processing in Section3.2 of the manuscript.
  
  Q19: In response to the reviewer’s valuable comments, we have revised the Introduction, the discussion of the key system technologies, the simulation error analysis, the comparison results with radiosondes and the meteorological tower, and the joint comparison results in accordance with all revisions made throughout the manuscript. These changes ensure that the conclusions are fully consistent with the revised main text, figures, and tables. Details of the revisions are provided in the corresponding sections of the manuscript.
  
  Q20: We sincerely appreciate the reviewer’s constructive comments. The original manuscript did contain shortcomings in its writing and presentation. We have systematically revised the language and logical structure throughout the manuscript, standardized the terminology, improved the sentence structure, strengthened the transitions between sections, and corrected the correspondence between the figures, tables, and main text, thereby comprehensively improving the overall quality of the manuscript. The revisions can be found throughout the manuscript.
  
  Citation: https://doi.org/10.5194/egusphere-2026-3088-AC2

Zhichao Bu, Yaru Dai, Yuanyuan Huang, Yubao Chen, Zhigang Li, Guohua Jin, and Xuliang Chen

Viewed

Total article views: 367 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
242	90	35	367	24	19

HTML: 242
PDF: 90
XML: 35
Total: 367
BibTeX: 24
EndNote: 19

Views and downloads (calculated since 03 Aug 2026)

Month	HTML	PDF	XML	Total
Aug 2026	126	40	15	181
Sep 2026	73	33	15	121
Oct 2026	43	17	5	65

Cumulative views and downloads (calculated since 03 Aug 2026)

Month	HTML	PDF	XML	Total
Aug 2026	126	40	15	181
Sep 2026	73	33	15	121
Oct 2026	43	17	5	65

Viewed (geographical distribution)

Total article views: 303 (including HTML, PDF, and XML) Thereof 303 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 24 Oct 2026

Short summary

We developed a dual‑laser reference lidar that measures wind with high resolution from 10 m up to 3000 m. Validated against weather balloons and a tower, it calibrates other lidars to wind speed errors below 0.8 m/s and direction errors below 8°, filling a domestic gap and advancing wind lidar use in meteorology.


Total:	0
HTML:	0
PDF:	0
XML:	0