General Comments

With the progress in the floating turbines technology, an increasing number of offshore projects are going to be implemented in coastal deep waters; offshore areas so far disregarded, are receiving now increasing attention like the US East coast.

In this context, this paper is timely focusing on the major meteorological phenomenon in coastal areas i,e, the breezes. It presents a methodology to detect the three types of sea breezes and their characteristic features, such as the calm zone associated with pure sea breezes and coastal jets associated with corkscrew sea breezes and discuss those feature in a wind energy prospective

Specific Comments

This paper is well structured and written but, in my opinion, the authors,  should expand the discussion of the impact of the SB from the wind energy perspective. In fact, the authors show that there are calms and divergence zone that impact on single turbine production in different breeze types (pure and corkscrew and backdoor).

They found that “the power production associated with a 10 megawatts offshore wind turbine would produce approximately 3 to 4 times more electrical power during a corkscrew sea breeze event than the other two types of sea breezes”. But there is more than this. There is  the issue of finding the right layout of a wind farm or of wind farms clusters with respect to wakes; a wind farm might be split by a calm zone in at least two areas with different wind directions. In this case, the wake losses of the whole wind farm might be less and the production more.

@pag 10 the authors write " In addition, the location of the calm zone varies by cases, although most calm zones develop relatively close to the coastline "  Here, my comment is that an analysis of the variability of the distance from the cost and the amplitude of the calm zone are variables s for sure of interest for projects developers.

The paper is well written and easy to read. It considers an important topic for wind energy meteorology and better understanding of sea-breezes is  a welcome addition to the field. However I have a few comments considering the methodology. I believe that there are some critical aspects of methodology that have not been properly described. 

Major comments.

• P3L84-L85. The simulations are one month long. Was there any kind of nudging performed during the simulations? If not, please explain why there was no nudging performed because one month is quite a long time and the model can “run away” from the real atmospheric conditions.
• P6L129. Mean wind direction at 10 m is calculated. Wind direction is a circular variable and therefore I find it hard to interpret “mean wind direction”. Please explain in more detail how you calculated mean wind direction in complicated meteorological situations and why such an approach is feasible, especially, taking into account the fact that the averaging is done over coastal quadrants, where sea breeze front can be present and therefore opposite wind directions can be next to each other. For instance, the average of W and E direction is S wind. 
• Figure 3. I do not understand Figure 3. It is supposed to demonstrate the differences in prevailing wind between different types of sea-breeze. But the classification is based on the relationship between the prevailing wind and the shoreline. In these schematics the shoreline is not indicated. The brown and blue color is especially confusing here, because it is reminiscent of land/sea border in maps. I also have a problem that authors haven’t defined how they interpret the direction of shoreline in the actual map. Is the north wind supposed to be purely offshore (coast-perpendicular) wind in this study? But the coastline is not oriented W-E, it has a complicated shape. Authors should clearly describe how they interpret the shoreline direction in this study and which prevailing wind directions correspond to “pure”, “corkscrew” and “backdoor” directions and why.
• Figure 6. I am not sure if showing the composites here is the best way how to represent the findings. Authors admit it themselves: “Even though the composite 10 wind speed over the calm zone is between 2 and 4 m/s, it falls primarily between 0 and 1 m/s for each individual case”. I am wondering if showing a representative single case would not be better to illustrate the properties of sea breeze. I am wondering whether the problem is the fact that the evolution of sea-breeze depends less on the “absolute” timing and more on the hours elapsed after sunrise. Maybe if the composite was done by averaging timeframes relative to the time after sunrise, the composites would be better. I imagine that sunrise time changes quite a lot during the year at those latitudes.
• P9L200-202 “ This could be partially associated with the increase in the land-sea thermal contrast. As the land-sea temperature difference becomes more strongly positive, there is greater potential for corkscrew sea breeze development over pure sea breeze development along the U.S. Northeast coast.” I am confused about such an assertion. If the difference between pure and corkscrew sea breezes comes from the difference in prevailing wind direction ( some authors use “geostrophic wind” here), how does the sea-land temperature difference influence prevailing wind direction?

Minor comments:

• Figure 2. It would be easier to understand Figure 2 if the explanations for the abbreviations, such as WR: ”Wind regime” would be explained in the figure caption.
• Line 152. “SW-WR” – I assume that “CS-WR” is meant here.   

The paper described a method to identify three types of sea breezes – pure, backdoor, and corkscrew – from WRF model simulations conducted over the US Northeast. The simulations cover only one year (September 2019 – August 2020) and used two-way nested domains of 6 and 2 km resolution. Statistics of the results are presented, as well as a sensitivity analysis to the values of the thresholds adopted to identify the sea breezes and a comparison of the typical wind power production to be expected for each sea breeze type.

The paper is well written and the figures are clear, but there are so many issues with the paper that, in my opinion, it should be rejected for publication in this journal, although it might be suitable somewhere else. 

I’ll focus on major issues first.

The first major issue is that it is not clear why this paper is relevant for wind energy. The authors focus on sea breezes along the US East Coast and only at the end of the Discussion section present two figures somewhat relevant to offshore wind to presumably show that the power output depends on the sea breeze type. How innovative or useful is this type of information? Why was the turbine placed in the center of domain 2? How would a developer or wind farm operator benefit from this information? I suggest that the authors choose a different journal, one perhaps focused on climatological or meteorological aspects.

The second issue is that the simulations cover only one year, therefore they are not long enough to produce meaningful statistics, climatologically speaking. The paper does not explain why the period was chosen. I would understand if the authors had collected observations over that period and wanted to validate the model results, but they did not, which is in fact my third issue. There is no model validation and we are left with no convincing evidence that the two-step method indeed finds sea breeze events correctly.

The fourth issue is methodological. The paper uses averages and means abundantly. I am particularly troubled by the use of average wind directions. Since the wind is a vector, the sum of vectors is not an average. What’s the average of a northerly (0°) and southerly (180°) wind? A wind from the east (90°)? It does not make any physical sense. As such, Figures 4, 6-8, and 11 are not acceptable because they show “average” wind vectors. In addition, even taking the averages over each sea breeze type at each hour is at least questionable. The authors make the implicit assumption that each sea breeze type evolves exactly the same at each hour and therefore taking the average at hour, say, 11 LT is meaningful. But this is not true, differences occur at 11 LT due to the season, due to the position of the sea breeze front, due to the background wind flow, to list a few. Aside from vectors, this is especially troublesome with convergence and divergence fields used to identify the average position of the sea breeze front, because averaging a positive and a zero or negative value at a grid cell, for example, could dilute the signal of the sea breeze location. The authors need to find alternative methods to characterize the statistics of the sea breezes, for example using median values or some pattern recognition techniques.

The last major issue is probably just a matter of explaining things better. There must be a sub-region or location of focus of the study, otherwise how can there be one sea breeze type for the entire domain 2? If the sea breeze is affecting New Jersey, it must be east-to-west, but in the northern shores of Long Island is it north-to-south? If it’s a pure sea breeze in New York City, could it be corkscrew somewhere else? I suspect that the issue is partly due to Figure 3, which I find very obscure. Where is the land? Is the prevailing wind the geostrophic wind? Where is north and south? Which way is the sea breeze flow?

Minor issues

1. Why were the 4 quadrants introduced? I don’t understand their purpose as they are not used. The text near line 130 talks about “mean … for each individual quadrant”, is this an area average over all grid points in the quadrant? Then a “dominant wind regime” for that day is obtained. What does dominant mean? How many hours out of 24? What if different quadrants had different sea breeze types?
2. Table 2: It suggests that 246 days had sea breezes in a year, which seems too many. Again, maybe the sea breeze types are not mutually exclusive, but then I do not understand how the averages are even calculated.
3. 185: here it seems that only 61 days were identified, but Table 2 is not consistent.
4. 205: again, I am confused about the averaging, are you averaging over all 3 types of pure sea breeze here? If so, it seems even more questionable to average over such a broad range of wind directions.
5. Figure 6: even given the fourth major issue above, it seems to me that the only location with a sea breeze here is Long Island.
6. Figure 8: The corkscrew sea breeze seems to be in New Jersey only.
7. Figure 9: because these figures are averages, the dynamic evolution is basically lost. There is no meaningful difference between the fields at 12, 13, and 14 for the pure sea breeze case, for example. Averaging out conv/div, the signal is diluted and the front is less distinguishable.
8. Line 260: I disagree that “Overall, the results indicate …” There is no evidence that the method works!
9. Line 292: which 10-MW turbine?
10. Line 293: why was this location chosen? Is it where a lease area is proposed? Please explain.
11. Line 319: Your methodology of averaging out everything washes out the details of the timing and evolution of the sea breezes, that is why your results are not consistent with past studies.