the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Characterizing Coastal Turbulence and Wind Speed Gradients Using an Anemometer Array to Improve Offshore Wind Energy Assessment
Abstract. As interest and investment in offshore wind farms increase, it becomes essential to better understand how interactions at the air-sea interface impact wind and turbulence in the marine boundary layer. Central to this effort is understanding the role of waves, especially on the relatively shallow continental shelves where most wind farms are located and waves are shaped by bathymetry. To address this, we analysed hundreds of hours of high frequency wind speeds measured by an anemometer array on a coastal tower from 14 to 26 m above the mean water surface. Wind speed gradients diverge substantially from established values, and the boundary layer was found non-constant ~60 % of the time. This means that the assumptions required for applying Monin Obukhov Similarity Theory cannot be met, so wind speeds aloft cannot be accurately predicted using this canonical methodology. Wind speed gradients were highest during alongshore winds and lowest during onshore. This occurred because waves refract and shoal parallel to shore which increases surface roughness independent of wind. Onshore winds cross the waves and create strong wind–wave coupling, while alongshore winds move along the waves, weakening the coupling and reducing effective roughness. Changes in the roughness then alters the wind speed which propagated through the boundary layer impacting winds aloft. The results of this study should be used to inform wind farm siting to optimize energy yield.
- Preprint
(1383 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 12 Mar 2026)
-
RC1: 'Comment on wes-2026-21', Anonymous Referee #1, 02 Mar 2026
reply
The comment was uploaded in the form of a supplement: https://wes.copernicus.org/preprints/wes-2026-21/wes-2026-21-RC1-supplement.pdfReplyCitation: https://doi.org/
10.5194/wes-2026-21-RC1 -
RC2: 'Comment on wes-2026-21', Anonymous Referee #2, 03 Mar 2026
reply
The comment was uploaded in the form of a supplement: https://wes.copernicus.org/preprints/wes-2026-21/wes-2026-21-RC2-supplement.pdf
-
RC3: 'Comment on wes-2026-21', Anonymous Referee #3, 07 Mar 2026
reply
The manuscript deals with an analysis of a profile of sonic anemometer observations at a coastal mast. Wind profiles as well as flux profiles and stability correction parameterizations are investigated.
Except for the conclusion, that the MO theory should be used with caution in the coastal area – which is well-known – I do not see any real findings that from an applied point of view can be used to improve wind energy assessment.
Â
Some comments:
- Line 13: You should clarify what you precisely mean by “boundary layer”, likely you mean profiles within the boundary layer, but it could also be its height or any other parameters such as i.e. temperature.
- Line 18. “reducing effective roughness”. Please add a reference in support of this statement
- Line 25. Be more specific how the effect of waves and swells can be accounted for and how they are dealt with in this study. I think you have measurements of both.
- Line 51: “fair weather”, be specific.
- Line 60:It is the absolute value of L that represents the depth – not the magnitude. Please note that the fluxes during stable conditions are not equal but have different signs.
- Line 68: which is the assigned height for Uo. It is given in the manuscript as the surface. It cannot be z=0, and at z=zo the wind speed is zero. There is a problem with Eq.(6), that by the way also misses psi(zo/L).
- Line 69. In order to derive U(10N) from Eq.(4) and (5) the roughness length is needed (or you need observations at two heights). Please clarify.
- Line 117: How were tides and swells accounted for? Tides can substantially change the distance between the mast and the shoreline and thus influence the flux foot print.
- Line 124: More details on the measurements and data processing are needed- lengths and direction of the booms, was the Schotanus correction considered and the consequences if not, how was flow deformation by the lattice tower taken into account….
- 2: Please add color bars as in Fig. 3 for the wind speed ranges.
- Line 170: Discuss and quantify the statement “This limitation does not critically impact the objectives of our study which does not require precise localization of the flux source”. The deviation from standard MO theory over flat terrain of flux profiles you find in the sonic profile observations are (partly) caused by differeces in the flux foot print so the precise location is important!
- Line 171: “provide context to the results”. Please be more specific.
- Line 172: How did you get the roughness length and boundary-layer height?
- Line 185:”We also note that the flux footprint includes a horizontal crosswind component which isn’t quantified here but further increases the heterogeneity.” Please clarify and discuss the effect and here is a good opportunity to discuss tides and swells as well.
- Line 194: At what height is the wind velocity estimated?
- Line 198: What is precisely meant by “account for the impact…..” . To account for the impact a parameterization is needed, and u-star/U can hardly be considered as a parameterization.
- Line 230: “Results indicate that the flux layer does not vary with height”. But it has just been showed, that u-star varies with height! And what results are you referring to? Please clarify.
- Line 233: “most of the time” : be specific
- Line 234 I suggest to help the reader to better understand what p actually signify in this context. A value of 0.05 is the traditional value for significance, but what does a p value of 0.25 actually indicate?
- Line 247-248. The numbers (69%, 50% and 57%) are not in accordance with the measurements (bars) in Fig. 9, and the fitted distribution is a very poor fit.
- Line 267: Not correct, increasing u-star increases L (absolute value) and thus moves the atmospheric conditions towards neutrality.
- Line 292: I note that there is no L in Eq.(3) on the right hand side so I wonder how you can estimate (z/L) in the psi function from Eq. (3). Please clarify.
Â
Citation: https://doi.org/10.5194/wes-2026-21-RC3
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 185 | 90 | 12 | 287 | 12 | 20 |
- HTML: 185
- PDF: 90
- XML: 12
- Total: 287
- BibTeX: 12
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1