Review of Sheridan et al. 2023: Offshore low-level jet observations and model representation using lidar buoy data off the California coast

First of all, I want to congratulate you on a very nice piece of work, that is really well written and is an important scientific contribution to the wind energy community. The figures are clear and easy to interpret and new creative ways of presenting the results are used, providing great insights in the data. It was a joy reading a manuscript that seems ready for publication as is.

The authors study observations of the wind profiles at two sites off the California coast, assessing them for low-level jets (LLJs). The seasonality and diurnal cycle of the appearance of the LLJs is investigated as well as key characteristics such as LLJ core height, core speed, and duration of LLJ events. Further, the authors validate three different models (the global ERA5 reanalysis and two regional models with different planetary boundary layer schemes) in terms of accurately resolving LLJs.

I only have some minor comments that I, humbly, think can enhance the manuscript further:

L35: It would be helpful for the reader to add a map showing the exact locations of Humboldt and Morro Bay, including names of sites (such as Cape Mendocino and Point Sur that you mention in the text) on the map. Perhaps also photos of the buoys and a visual presentation of the data availability from Humboldt and Morro Bay could be included in this Figure.

L47-48: 300 – 400 m is still in a layer above the top of the rotor of an offshore wind turbine. It would be good to give an example of the size (hub height and rotor diameter) of a modern offshore wind turbine, such as the NREL 15 MW turbine you refer to later on.

L54-55: The New European Wind Atlas (e.g., Dörenkämper et al. 2020) would also be a nice reference here

L66: To add more information on the offshore wind profile and LLJs and how they are affected by the selection of the PBL scheme, I would suggest adding a couple of sentences summarizing the results from Svensson et al. (2016)

L94: Here it could just be clarified that: 17 October 2020 to 30 September 2021 (i.e., almost a year) for Morro Bay and 17 October 2020 to 27 December 2020 and 24 May 2021 to 30 September 2021 (i.e., six months in total) for Humboldt.

L106: Clarification: intervals of 20 m between 40 m and 240 m a.s.l. (i.e., 11 height levels)

L108: Here the information that data was collected as 10 min averages should be added. Also, a few words on the quality control of the data would be nice.

L159 and throughout the manuscript: Using comma signs as thousand separators could increase the readability of the large numbers (such as 24,878 observations on L159)

L170-171: As the distributions of LLJ event duration is highly skewed (as also shown in Figure 8), the median duration should also be presented, accompanying the mean value.

L183-184: You refer to Figure 3a and Figure 3b, but there are no panels in Figure 3.

L185: Commenting on the diurnal cycle presented in Figure 3, I think you should also mention that the comparison between the two sites is possibly biased because of the data outage at Humboldt. It would be great if you could add confidence intervals around the lines in Figure 3 to motivate how certain you can be if there actually is a diurnal cycle in LLJ occurrence.

Figure 6: In the caption, if you could write “bulk wind shear below the jet core”, that would remind the reader and help interpreting the plot.

L242: Is it the closest grid points that have been used? Please specify the distance between the grid points used and Humboldt and Morro Bay, respectively.

L247: Also, Kalverla et al. (2020) should be cited here (see Figure 6 in that paper)

L252-256: In addition to the success rates, the frequency bias should be included.

Figure 13: As you are commenting on the prior LLJ event in the text, I would love to see the time series starting already at 07:00 UTC to also include this event in the plot. Perhaps you could also mark the LLJ core for each time step where an LLJ is present in the panels?

L375: Using the rotor equivalent wind speed (REWS) instead of only the hub height wind speed to calculate the estimated power production would really strengthen the work. For reference see e.g., St. Pé et al. (2018) and Murphy et al. (2020)

L409: Finally, adding a paragraph discussing general similarities and differences when comparing your results with LLJ studies from e.g., the US Atlantic coast, the North Sea, and the Baltic Sea, would put your work into a broader context and would be very valuable for future reference.

References:

Dörenkämper , Martin, et al. "The Making of the New European Wind Atlas – Part 2: Production and Evaluation." Geoscientific Model Development 13.10 (2020): 5079-5102, doi: 10.5194/gmd-13-5079-2020)

Kalverla, Peter C., et al. "Quality of wind characteristics in recent wind atlases over the North Sea." Quarterly Journal of the Royal Meteorological Society 146.728 (2020): 1498-1515, doi: 10.1002/qj.3748

Murphy, Patrick, Julie K. Lundquist, and Paul Fleming. "How wind speed shear and directional veer affect the power production of a megawatt-scale operational wind turbine." Wind Energy Science 5.3 (2020): 1169-1190, doi: 10.5194/wes-5-1169-2020

St. Pé, Alexandra, et al. "Classifying rotor‐layer wind to reduce offshore available power uncertainty." Wind Energy 21.7 (2018): 461-473, doi: 10.1002/we.2159

Svensson, Nina, et al. "Stable atmospheric conditions over the Baltic Sea: model evaluation and climatology." Boreal Environment Research, ISSN 1239-6095, E-ISSN 1797-2469, Vol. 21, s. 387-404 (2016)

Summary Comments:

            This study presents a nice comparison of observations from a temporary LiDAR buoy deployment off of the California coast, to the NREL 20-year Wind Resource Dataset (CA20-Ext), the 2023 National Offshore Wind Dataset (NOW-23), and the ERA5 reanalysis product. This work is important as we look forward to offshore wind development on the West Coast, as model errors in low-level jet (LLJ) representation can affect wind turbine energy generation. This study successful compares three different models to offshore lidar observations and identifies the strengths and weaknesses of each.

            I do have a few minor comments that I believe will help the clarity and flow of the paper. Some general comments are that I do think that the figures could benefit from panel titles that have the lidar buoy location/model name. Even though it is explained in the figure caption, I think it would be nice to also add titles for quick referencing. I also noticed that the word "respectively" was very used often in the analysis sections (specifically sections 4.1 and 4.2), as well as a few long run-on sentences which made the paper more difficult to read. The rest of my comments are more specific and are listed below.

            Overall, I think this is an excellent study and I am excited to see the results of another offshore lidar deployment. It is difficult to get profiler measurements offshore, and the field needs studies like this one, that can showcase the discrepancies between model predictions and observational data for wind energy applications. I really appreciated how the authors compared how the model bias in LLJ prediction could impact wind energy generation forecasts. I recommend this paper goes to publication.

Major Comments:

Lines 30-35: This paper is missing a figure that shows entire study region, including the location of the 2 lidar buoys (at Morro Bay and Humboldt) as well as the proposed 5 lease areas. I think this would be very helpful in providing some context for those that aren't familiar with the study area.

Lines 85-90, 188: Again, referencing a map of the study region here would help your discussion.

Lines 153-156: Why did you choose a 2 ms^-1 windspeed drop off threshold in your LLJ identification? Is this based off a percentage of the mean wind speed that is typically seen in the region? I think explaining why the 2 ms^-1 works well in this offshore environment will be helpful in making this study more applicable in other areas.

Lines 159-165: Did you consider any other variables when identifying the LLJs? You mention that there is no restriction set on the vertical distance of the 2ms^-1 fall off, but did you set a minimum or maximum wind speed threshold? If not, why? What is the mean windspeed at 140m at these buoy locations? Is the jet windspeed maximum generally higher than the mean windspeeds? I think in order to help make this LLJ identification algorithm more relevant to other offshore locations, it would be helpful to explain your specific choices and thresholds in a little more detail.

Minor Comments:

Line 16: You mention that these cold season LLJs generally occur below 250m and that the warm season LLJs are generally higher. It might be nice to mention the height the warm season LLJs are observed at for context (e.g. higher altitude California coastal jet (typically at heights of 300–400 m) influenced by the North Pacific High). 

Line 18: I think you could cut out a couple words in this sentence to simplify it (it is a long sentence). "The lidar buoy observations also support the validation of LLJ representation in atmospheric models that are essential for assessing the potential yield of offshore wind farms"

Line 25: It might be helpful for the reader to briefly define the term "false alarm" here.

Lines 38, 145-155: Another study you could take a look at is McCabe and Freedman 2023 (https://doi.org/10.1175/WAF-D-22-0119.1). That paper discusses another shear based approach to identifying LLJs that occur during sea breeze events on the east coast.

Lines 40-45: You mention the impact of the LLJ on wake effects and turbine fatigue, but you don’t mention how the LLJ may be able to increase wind speeds across the rotor plane, therefore increasing potential energy production until the end of the paper. I think it might be helpful to add a sentence on that here.

Line 101: Just to clarify, here you mention the buoy being equipped with the Leosphere WindCube 866 lidars, and in the past section (line 81), you mention the AXYS WindSentinel buoys are have Leosphere WindCube v2 lidar systems. Are these lidars the same?

Lines 183-184: You refer to figure 3a and 3b, but figure 3 only has one panel.

Lines 190-194: The second half of this sentence is a little confusing for the reader. I would suggest re-phrasing the sentence (especially the part after "directions associated with ... ") for clarity.

Line 226: I would consider reminding the reader that L (Obukhov length) is defined in equation 1.

Lines 288-289: For ease of reading, it might help to break this sentence into two sentences. You could simply add a period after the word "phenomena".

Line 325 (Figure 10): I think you need some sort of legend, or just an explanation in the figure caption, that clarifies what the numbers on the plots represent.

Lines 340-345: How does the LLJ core wind speed bias vary with height? Is it more pronounced at certain LLJ core heights than others?

Line 378: I would suggest putting the abbreviations you are using as the subscript in equations 2 and 3 for lowest (lo) and highest (hi) in parenthesis.

 Summary

The authors present novel offshore LLJ research off the coast of California in regions relevant to wind energy. This research includes not only observational LLJ characterization, but also various model comparisons. Overall, observations capture LLJ events occurring only a few percent over the observational campaign but note that the vertical extent of these observations will undercount LLJ events. The models tend to underestimate the number of LLJ events but overestimate the duration. Models do a good job of capturing the hourly timing, but struggle on a seasonal basis. Other LLJ characteristics have positive/negative biases depending on the location and model used.

Overall, I believe this research to be a great addition to the offshore wind energy literature, and I recommend this article to be accepted with minor revisions, which are listed below.

Major comments

• I feel the paper ends abruptly and that no proper conclusion is given. It would be beneficial for the paper to include, at the very least, future directions this research could/should go to help offshore low-level jet characterization.

Minor comments

• Line 45: It is stated that the California coastal LLJ is well-studied, yet only two references are included. The first reference, Parish 2000, is repeatedly mentioned and referenced from this point on. To the authors knowledge, are these the only two studies to investigate LLJs along the California coast? If so, I would recommend rewording this sentence to tamper down the claim that the California coastal LLJ is well-studied.
• Lines 88-89: Do the phrases “in 1100 m of water” and “in 625 m of water” refer to the depth of the ocean floor to the sea-level surface? This reads weird to me.
• Section 2.1: What is the temporal resolution of the lidars and ultrasonic anemometers, and was any temporal averaging done? I understand the specifics are in Severy et al. (2021) and Krishnamurthy et al. (2023), but I feel at the very least basic temporal resolution of these instruments should be discussed.
• Section 2.2: What is the advantage of using the TOGA COARE algorithm, compared, say, to alternative algorithms? I feel a little more discussion is warranted here.
• Section 2.4: While I am aware that quantifying LLJs are an unsettled topic of discussion in our field, I would be remiss not to request Debnath et al. (2021) to be included in this section. While Debnath et al. (2021) leverages Kalverla et al. (2019)’s work, the Debnath paper has been used extensively in LLJ discussions in the U.S.
• Lines 163-164: I appreciate the mention of the limitation of the observations in capturing the true extent of LLJ events.
• Figure 2b: It is interesting that the longest LLJ occurrence at Morro Bay occurs at the end of the period. I assume this is by chance, no?
• Line 242: When it is said that the data are resampled at the top of the hour, does this mean that only top of the hour observations are taken, or some mathematical operation is done (hourly average)?
• Figure 11: Neat way to visually compare LLJ core height differences!
• Lines 352-353: Is that duration time correct? 10:00 to 14:10 UTC is not 14.2 hours. Am I missing something here?
• Line 513: The paper for this reference has been officially published and should be updated accordingly.

References

(Debnath et al., 2021)

Debnath, M., Doubrawa, P., Optis, M., Hawbecker, P., and Bodini, N.: Extreme wind shear events in US offshore wind energy areas and the role of induced stratification, Wind Energy Science, 6, 1043–1059, https://doi.org/10.5194/wes-6-1043-2021, 2021.