Dear authors,

thank you for a valuable overview of the challenges of complex terrain. Overall I quite liked the paper, and have just a few minor comments.

Two comments to start with: you seem to lump operational forecasting together with wind resource assessment, which is not getting so clear from the introduction. It's fine to discuss both, but it was my feeling from the start that you were mainly discussing WRA. And is there no other noteworthy issue than icing, since you have that particular effect quite prominent? Lightning or dust, e.g.? 

Line 24+: "Also, some regions...have to build in hilly..." The assessment here sounds a bit negative. Often you also see wind power erected on hills to get the speed-up effect on hill tops, which is an active choice. 

L102: Hopefully, to use more effort has a positive effect, not just a potential effect?

Figure 2: Looks weird. What is the high-frequency floor at 4 m/s, why is it not changing at all during the two days, and from the plot there is no way that both time series have the same mean (within 4%). The diurnal "mountain" goes to 10m/s, flattening the steep slopes to a rectangle gives some 4-5 hours duration per day, and the floor is pretty much identical. With those simple geometric assumptions, I get the red line to be over 5m/s in average. 

L222: Drones can cover an area, or is that not what you are thinking of here?

L242: Should you also include Direct Numerical Simulation (DNS) in this list?

L248: ICON-D2 is to my knowledge not a LES code? 

L306+: There are attempts to make climate models more user friendly under the Copernicus initiative from the EU ("C3S").

L454: On the difficulties of ground-based lidars you probably have a reference?

L506: please add references to both test centres. 

L644: The rationale for the RIX was that from the slope of 30%, the flow is non-attached, and is outside the operational envelope of WAsP. It would be nice to also mention both critical slopes in the comparison.

Typos and other editorial issues (please only answer to those if you disagree):

Abstract lines 2+3, please check the sentence structure. 

L81: "At this stage of a wind energy" - project?

L142: check sentence structure.

L187-190: "This can mean that they measure in inhomogeneous flows" sounds maybe not quite right here. And the measurement volumes are only a part of the problem, the way the lidar averages over a circle is also to blame (as you also illustrate in Figure 4). 

L197: vary -> very?

L216: There seems to be a reference missing at "()"?

L289: ...require the wind industry _to_ develop... ?

L421: IEA Users_TCP ?

L430: focus_on

Page 17, footnote 1: "Task 19 document" is not a good reference, sounds more like a placeholder during the writing process...

L494: to be minimise -> to minimise ?

L565: ...from one regime to the other, errors... (add comma?)

L705: Task 43_Digitalisation_ (adding the short qualifiers to the task names makes them easier to remember).

L750: The Acknowledgement is quite unspecific, usually they require project numbers.

And you often seem to have an extra blank before your references, but maybe that's just some layouting which will go away in the final paper trim. E.g. L90, L100, L173, L333, L339, L406, L600

WES-2022-11 | Review article

General comments:

The article sets out a fairly comprehensive set of research challenges for wind energy in “atmospherically complex” locations.

My main feedback is that this article should be seen as adding to similar “challenge” articles in recent years. For example, van Kuik et al (2016) (not cited in the article under review) and Veers et al (2019)  (not cited in the article under review). While the two aforementioned published challenges articles go beyond atmospheric science aspects, they do feature many atmospheric challenges. I would like the authors first to recognise these aforementioned challenges papers  and state how their article goes beyond what is stated in those publications. 

I think the article is rather long. I can see the purpose of the article, which really is focussing on complex terrain locations, but I lack a succinct summary and set of recommendations. Perhaps Table 1 can be made more central to the article and have additional columns with headings such as “impact” and “priority” following the R&D column. Then the main text can be reduced and focussed on discussion points around the table. 

One last comment, while terrain complexity and icing are mapped (Figures 1 and 3), I think the “complex flow” and “complex weather and climate” are quite overlapping and very general terms. It would be difficult to map these. Where is weather and climate simple, one could ask? Therefore it is almost a corollary; that where there is complexity there is also challenge. 

Specific comments:

P 1

Abstract: Please state briefly what is the new conclusion emerging from this review beyond what can be found in earlier “challenges” articles.

P2 

L20 What is the definition of mountainous in Fig 1 and main text.

L25 Mention offshore resource can also be used if “simple” terrain is exhausted or not available.

L30 Insert “...and sudden *associated* changes in wind speed…” 

P3 

L57 “Complex weather and climate” seems a bit loose. Is there “simple weather and climate”? See general comment above.

L70 “relating to hilly, mountainous or forested locations”, is that the same as complex terrain discussed on same page? If so, can that term be used?



P5 

L110  Fig 2  I completely agree that the diurnal cycle is important to capture. But this is not a new result. Furthermore time series data is not required to capture that. Knowing variance of wind speed, via for example, k (shape parameter) together with A (scale parameter) of a Weibull distribution of wind speed, would capture this too effect too.

P9 

L195 I think a reference to Bingöl would be fitting here:

Bingöl, F.: Complex terrain and wind lidars, Dissertation, Risoe National Laboratory for Sustainable Energy, Technical 545 University of Denmark, Roskilde, 2009. Bingöl, F., Mann, J., and Foussekis, D.: Conically scanning lidar error in complex terrain, Meteorologische Zeitschrift, 18, 189–195, 2009. 

P14

L350 “human”->”person” perhaps?

L357 About the word “maximise” and “minimise” in the sentence. I do not agree with the sentence. Layouts should rather maximise the ratio of production to costs. It is not the same.

P16 

L430 “focuson”->”focus on”.

P18

L474 Figure 6, it is not completely clear what this figure is illustrating. Is it needed?

P20

L545 Please state how expenses can be reduced?

P21

L562 Consider changing “inherent uncertainty” with “chaotic nature” or similar.

P22 

L619 It is not surprising there is no agree-upon definition of “complex terrain” because the definition will probably depend on the use case or model in questions. I would argue that when speaking of complexity, one needs to qualify the measure being used, e.g. what threshold slope value is being used etc. 

Several places “RiX”->”RIX”, it stands for Ruggedness INdex.

P24 

L669 “interaction of the terrain and *flow*” is perhaps better in this context.