Dear authors,

thank you for your valuable submission. I enjoyed reading your manuscript, which I believe is worth being published in the journal Wind Energy Science. I only have a few minor comments that will hopefully improve the final draft:

• Abstract: The sentence “The difference in loads between unstable and stable conditions was found to be 98% for the tower and 20% for the blades” is too generic even for an abstract and is likely impossible to be understood without reading the rest of the manuscript. Also, the sentence “loads from the two sources” is unclear. What are the two sources? I'd suggest rethinking the abstract to make it more self-standing.
• Page 2, line 27: you write "On the contrary, atmospheric turbulence is often thermally generated, is inhomogeneous in the vertical direction and follows non-Gaussian statistics at the smallest scales". Although I agree with you, please include a couple of references supporting your claims.
• Page 2 line 54: “due to inaccuracies in the aerodynamic modelling and the absence of certain controller features in the simulation environment”. The aerodynamic modeling has inherent uncertainties, but I don’t think the authors pinned the discrepancies in the results to problems with aerodynamic modeling. The lack of the industrial controller certainly contributed to the offsets and the following TORQUE 2024 publication (Doubrawa et al., 2024), although limited to a single case, showed a notable improvement to the metrics when a rotor position controller was adopted. You already cite the work from Doubrawa et al., and you might want to add a note in the intro about it.
• Table 1: bottom two rows are identical, probably a typo perpendicular vs parallel?
• The font size across the different figures of the manuscript is inconsistent. Please improve. (Font size in Figures 1, 2, and 5 is too large.)
• Page 6, line 120: “Indeed, when computing the spectra for all 10 min periods within a given bin, we found the spread to be large, encompassing at least three decades.” What does this mean? Three decades? Or three standard deviations?
• Page 8 line 156: typo, as “expected”
• Page 9, figure 3: I suggest to move the legend to the top left plot, top right corner. It took me a second to find it, and currently it covers some of the data.
• Page 12 line 213: I don’t think you are “reminding” the reader. I expected that results were going to be nondimensionalized, but I don’t think you had told me just yet. Maybe “highlighting” is a better choice?
• Page 14 line 259: “It is difficult to replicate these in the simulation environment.” I’d suggest a small rewording. Modeling these effects is easy, the problem is the exact quantification of the imbalances.
• Page 14 line 262: “While we present an analysis of load measurements, Sathe et al. (2013) showed the effect of atmospheric stability on fatigue loads in a simulation environment. However, our results appear to be remarkably similar which indicates a fundamental relationship between atmospheric stability and fatigue loads on a pitch controlled horizontal axis wind turbine.”. Why however? Your results are similar to the ones from 2013? I’m confused by this sentence. Also, should it maybe appear after the results are discussed, and not before?

Thank you again for the submission and congratulations for a very nice contribution to the community.

See attached document