"Bayesian uncertainty quantification of engineering models for wind-farm atmosphere interaction" provides an in-depth and thoroughly justified approach for an improved UQ method based on authors' previous work. The value is defined clearly, method is well-written and should allow for reproducibility, and credit and reference to other works is sufficiently provided. The paper is well-structured, and figures and tables are used well to provide clarity to points made.

I feel the paper does not require further work, but could be strengthened by attention to the following areas:

• Better introduction/description of statistical terms. Non-statistical wind science readers may benefit from a glossary of terms, referral to other sources at points where new terms are introduced, or a brief explanation where some of these terms are provided for the first time.
• Similarly, wind science readers with some statistical knowledge may well have only carried out frequentist UQ without being aware of the differential between Bayesian and frequentist statistics, or may not know why it is appropriate here. Giving more reference to this difference in approaches, or clearly outlining the link between Gaussian methods and Bayesian UQ may be useful for readers
• Subfigure labels are difficult to see, if these could be made clearer (if journal conventions allow)
• New line equations could be explicitly labelled e.g. a=b /newline = c should be equation 1a, 1b, as they can be considered a new statement (if journal conventions allow)