the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Oct 2025
Long-term measurements of wind and turbulence from a calibrated 1290 MHz radar wind profiler at the Eastern North Atlantic atmospheric observatory
Abstract. Turbulence in the marine boundary layer is closely coupled with the physical properties of marine boundary layer clouds. However, these turbulent motions are difficult to observe, resulting in very few observations of turbulence properties over the open oceans. In this work data collected by the 1290 MHz Radar Wind Profiler (RWP) part of the Atmospheric Radiation Measurement (ARM) Eastern North Atlantic (ENA) site are used to characterize winds and boundary layer turbulence. The RWP wind, precipitation long pulse and precipitation short pulse modes were calibrated using the surface laser disdrometer. Over the 10 years of data analyzed here, the RWP wind mode calibration constant, sensitivity, and dynamic range fluctuated between -84 to -44 dB, -15 to 9 dBZ, and 12 to 21 dBZ, respectively. The clear-air RWP echoes were used to derive structure function of the refractive index (Cn2) and dissipation rates of turbulence kinetic energy (ε). Both Cn2 and ε decreased from surface upwards with higher values in the winter months and lower values in the summer months consistent with previous studies. Despite the marine location, the ε was affected by island heating during southerly wind conditions. The results reported herein have implications for RWP operational research and climatological studies at the ARM ENA site.
- Preprint
(6972 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on egusphere-2025-4434', Adam Theisen, 06 Oct 2025
-
CC2: 'Reply on CC1', Adam Theisen, 06 Oct 2025
Correction: KAZRARSCL does use the PR mode.
Citation: https://doi.org/10.5194/egusphere-2025-4434-CC2
-
CC2: 'Reply on CC1', Adam Theisen, 06 Oct 2025
-
RC1: 'Comment on egusphere-2025-4434', Anonymous Referee #1, 04 Mar 2026
This manuscript describes how the Department of Energy ARM program’s 1290 MHz Radar Wind Profiler (RWP) deployed at the Eastern North Atlantic atmospheric observatory was calibrated. While the subject matter is worthy of publishing in the journal of Atmospheric Measurement Techniques, the manuscript contains many conceptual errors and does not use standard radar terminology that makes the manuscript hard to read and understand. Most of the errors this reviewer found should have been caught by co-authors and corrected before submission.
This manuscript found that due to degradation of the RWP hardware, 39 months (over 3 years) of the 10-year record are not suitable for atmospheric research (see Line 300). This is a significant finding. Yet this finding is not mentioned in the abstract.
While I would like to see a future version of this manuscript published, the current manuscript needs a lot of work to clarify the work that was done.
Note that this reviewer read the Abstract through Section 3 and did not read the last 10 pages.
The following specific comments are in sequential order.
- Line 31-44, Abstract. The abstract is misleading in that it does not state that the RWP’s performance degraded over the 10-year period. The abstract does not mention the calibration constants changed over 30dB in the first 5 years, causing some radar parts to be replaced. One of the manuscript’s objectives was to document the RWP performance, but this significant degradation of performance is not mentioned in the abstract.
- The range of values of ‘-84 to -44 dB, -15 to 9 dBZ, and 12 to 21 dBZ’ are values taken from different operating modes and do not have much meaning when combined like this. It is more important to tell the reader that the calibration constant increased from A to B and the minimum detectable reflectivity increased from X to Y over the first 5 years of operation.
- Line 92-94. In this introduction paragraph, inform the reader that all instruments were deployed by the US Department of Energy ARM program and all input data are available to the public.
- Line 96. Incomplete sentence.
- Line 96 and 97. The first sentence says the radar was operational from September 2014 through February 2025. The second sentence says there was a 14-month gap in operations. Both sentences can be true. Rewrite these sentences to reflect the 14-month gap in data collection.
- Line 99. If there was heavy corrosion to the antenna and phase shifters, was there any degradation to the data quality or height coverage of the radar. Inform the reader that there was decrease in height coverage of the radar and the main subject of this manuscript.
- Line 103-104. The manuscript needs to be consistent in describing the RWP ‘modes’ and the RWP ‘beam directions’. It is incorrect to say that the “wind mode has three beams” and then refer to the beams as the “vertically pointing mode” and the “oblique mode”. The manuscript needs to be clear that the ‘mode’ refers to the operating parameters as listed in Table 1, and the individual beams refer to specific beam pointing directions. This correction is needed in several places in the manuscript.
- Line 104. There is no quantity with the name “Doppler spectra”. The term Doppler refers to a velocity or a frequency. Also, the radar produces a power spectra. Thus, the RWP produces a Doppler velocity power spectrum at each range gate. Please change the manuscript so that the first reference to the RWP spectra is called ‘Doppler velocity power spectra’. After this introduction, the manuscript can revert to a common name of spectra.
- Line 104. Moments are not collected, they are calculated from the spectra.
- Line 106-107. The sentence starting with “Between each wind mode …” is poorly written. Verbs do not match nouns. For example, “modes are operated” does not make sense because the modes do not operate. The RWP operates in different modes. Please re-write this sentence.
- Line 109. The full cycle time of approximately 53 seconds does not match the times listed in Table 1. In Table 1, the wind, long pulse, and short pulse modes require 27.6 seconds (3x9.2 seconds), 4.5 and 4 seconds, respectively. This adds up to less than 37 seconds. The text in the manuscript needs to match the values in the tables.
- Line 109. Again, the manuscript needs to be clear when discussing modes and beams. There are 3 modes described in Table 1.
- Table 1. The units for Inter Pulse Period (IPP) are wrong. They should be microseconds and not milliseconds.
- Line 114. The RWP does not have a “vertically-pointing wind mode”. This inaccurate description of the RWP operating modes and the beam pointing directions is very confusing to the reader. This error occurs is several locals in the manuscript.
- Line 117-124. The paragraph describing the KAZR radar and the ARSCL data product contains several conceptual errors, contains a misrepresentation, and has several grammar errors that this paragraph needs to be re-written. Specifically:
- Line 121. The KAZR radar is not calibrated with a corner reflector! This is a major error that raises questions about other errors that this reviewer is not catching.
- Line 121-122. As written, the sentence referring to the work of Clothiaux et al. 2000, implies that the current work applied the merging and noise-filtering of Clothiaux et al. This is not true, and is a misrepresentation of the work not performed in this manuscript. The merging and noise-filtering described in Clotheaux et al. was performed by ARM and is available to the public via the KAZR-ARSCL data product.
- Line 137. Wow, the laser disdrometer is not very accurate if it underestimates drops with “mean hourly rain rates exceeding 2.5 mm h-1”. Change the manuscript to have the correct units.
- Lines 150-160, first paragraph of section 3.1 Noise Filtering and binning. This paragraph does not describe the filtering process with enough clarity to be repeated by an independent researcher. The paragraph contains terms that are inconsistent with engineering principles. Examples of non-sensical statements include:
- Line 151-152. How are “daily profiles” of SNR binned to a “uniform time resolution”? This does not make sense to me.
- Line 152-153. How is the “noise floor” estimated from the SNR? The noise floor cannot be estimated from SNR measurements that contain atmospheric signals.
- Line 153. What are the visual inspection criteria used to determine the noise floor?
- Line 154. What is the “domain”? Does the domain span some or all range gates? Is there a temporal component to the domain?
- Line 157. What is the frequency response of the high-pass filter? Are the daily mean profiles being used or are the minute data being used in the high-pass filter?
- Line 158. What is a ‘pixel filter’?
- Line 158. What does it mean to have “noise pixels’ when the measurements are SNR?
- Figure 2. What is the definition of “noise floor” when the measurements are signal-to-noise ratio? How is the noise power estimated from just the SNR? Is this figure showing some estimate of the minimum valid SNR level, but incorrectly referring to this estimate as a “noise floor”.
- Figure 2. Is the “noise floor” estimated for every range gate? The minimum SNR will be different at low range gates than at far range gates. Include in the manuscript the range gate dependance of the “noise floor”.
- Line 165-166. The sentence on line 165 states that ground-clutter pixels are removed. The sentence on line 166 states, “…the artifact can be suppressed without removing actual measurements.” How can these two sentences be consistent? The text implies that data is removed, yet, data is not actually removed. That is an interesting concept. Please correct the manuscript.
- Line 167 and Table 2. What are the operating parameters during epoch A?
- Line 151 and 174. Line 151 says that the daily profiles were used to estimate the noise floor, yet line 174 says that the profiles were binned to 1-minute resolution. Please clarify the manuscript to remove this ambiguity.
- Lines109 and 176-177. Line 109 says that the two precipitation modes occur between the wind mode. Lines 176-177 says that the wind mode could occur once or twice in a minute, and the two precipitation modes occur 3 or 4 times in a minute. How do these sequencies match the 53 second cycle time listed in Line 109? Please make the changes in the manuscript.
- Figure 3a and Figs. 4a and 4c. There is an increase in SNR (Fig. 3a) and reflectivity (Figs. 4a and 4c) at the lowest range gates. Is this increase real or an artifact?
- Line 184. Where the 915 MHz RWPs used in the Tridon et al. and Williams et al. work manufactured by the same company that built the 1290 MHz RWP deployed at ENA? The manuscript should mention the manufactures and discuss how the operating parameters impact the collected data.
- The use of “minimum detectable signal” is another example of not using radar terminology correctly. The manuscript describes how the minimum detectable reflectivity is estimated, not the minimum detectable signal. Please change the manuscript.
- Figure 6. It is confusing to label the vertical axis as ‘Sensitivity’, when the value being plotted is the minimum detectable reflectivity. The figure would have more meaning to the reader if the vertical axis label was changed to minimum detectable reflectivity.
- Line 300. Identifying 39 months of data to be ‘bad’ is a significant result. This should be stated in the abstract.
Citation: https://doi.org/10.5194/egusphere-2025-4434-RC1 -
RC2: 'Comment on egusphere-2025-4434', Anonymous Referee #2, 08 Mar 2026
This manuscript describes the characterization of radar wind profiler (RWP) measurements at the ARM Eastern North Atlantic observatory (ENA). The dataset is then used to derive parameters that represent the structure function of the refractive index and dissipation rates of turbulence kinetic energy. The authors then discuss the climatology of these parameters and relate the findings to previous studies of turbulence in the marine boundary layer.
The description of the RWP characterization was thorough and followed methodologies from previous studies but was sometimes confusing, for example, referring to terms that were not introduced until later in the manuscript. It also seemed that there were details discussed that were not correct with regards to the ARM datasets, which could have been found in the technical documentation available on the ARM website. Several grammar errors and inconsistencies in the figures suggest the need for thorough proof reading in the revision.
The abstract and introduction do not explain clearly what the goal of the study is from a scientific perspective though it became more apparent toward the end. It would be helpful if you mentioned more specific findings in the abstract and a description of what you were expecting to learn going into this study, and from the climatology.
A well characterized RWP dataset would be valuable for other researchers, and I hope that you will make it publicly available. For example, you were able to identify the “good and bad” days. Having this quality flag and corrected data (if possible) would be very valuable.
Overall, this manuscript provides a thorough analysis of RWP data and making these issues aware to other potential users is valuable and worthy of publication. As mentioned, there are some inconsistencies and areas where the analysis is confusing, but these can be easily clarified, and I have pointed out some specifics below. I recommend this manuscript for publication once these revisions are complete.
Specific Comments
Fig 1b: It would be useful to show the location of the ENA site on the map.
Datasets: please state clearly the data level that you used for both the KAZR and RWP datasets and provide associated DOIs for data provenance and reproducibility purposes. For ARM data, a-level data is uncalibrated, b-level data is calibrated, and c-level is typically derived quantities using calibrated and corrected data. KAZR data at ENA was only recently calibrated to the b-level so it is most likely that the KAZR data in the ARSCL product that you are using is uncalibrated. Also, a corner reflector is not used for calibration. See the ARSCL technical report for information about KAZR calibration. DOE Tech. Memo. ARM VAP-002.1
Figure 2 shows the signal to noise ratio (SNR) for the RWP and 4 data epochs but the discussion around this figure and the development of epochs is related to sensitivity that is not discussed until much later in the manuscript. I suggest reorganizing the manuscript to introduce and analysis the specific parameters before defining the epochs and describing the calibration techniques.
Section 3.2 Calibration technique: Suggest adding some additional details regarding calibration method. It seems all the elements are there, but the order of the description does not make it easy to follow. Suggest describing the calibration method in words first rather than jumping into equation (1). For example, state that you calibrate the RWP by comparing the surface reflectivity derived from surface disdrometer measurements described in Sec. 2.3 with RWP reflectivity during raining periods. Then describe the calculations of the reflectivity from the RWP and disdrometer. Several times in this section you refer to following the references from papers by Williams and Tridon but don’t clearly describe the method in a logical way.
Lines 200-205: You state that 120 surface disdrometer measurements with >20 dBZ are needed on a given day – are these 120 continuous measurements?
Fig. 5: The ‘x’ symbols on the figure are very hard to see and distinguish from the larger, bolder 3-month average dots. Suggest making the ‘x’ symbols bolder in color.
Line 259-260: There are grammatical errors in this sentence.
Fig. 7: Line 290: You reference Figure 7b but it should be Figure 7d. Also in Fig 7 in the caption you refer to the KAZR ARSCL panel as ‘(e)’ but should be ‘(d)’.
Lines 299-300: Explain how you determined this -7.5 dBZ threshold.
Summary, lines 460-468: You mention comparing to ESMs, but this was only loosely mentioned in the summary, and it was unclear how you would compare the Cn2 and ε to model output. Can you expand on this comment?
Citation: https://doi.org/10.5194/egusphere-2025-4434-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 433 | 119 | 34 | 586 | 23 | 27 |
- HTML: 433
- PDF: 119
- XML: 34
- Total: 586
- BibTeX: 23
- EndNote: 27
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Nice work analyzing this long-term dataset from the ENA radar wind profiler. The detailed calibration approach and evaluation of instrument performance over nearly a decade provide valuable insight into data quality and long-term stability. There are a few minor points that may be beneficial to review for possible updates.