Review of Wind Models at a Local Scale: Advantages and Disadvantages

Round 1

Reviewer 1 Report

The manuscript in review surveys the ongoing efforts to produce finer local forecasts of wind speed using coarser results from global circulation models. It presents no original methods or results. It compares the performance of two broad classes of methods, known as Dynamic Downscaling and Statistical Downscaling.  As demonstrated in the works of Edward Lorenz from the 1970s, there exist fundamental limitations to predictability of atmospheric dynamics, which here manifest themselves as limitations to the accuracy of downscaling by any method.  The manuscript concludes that the issue of proper downscaling is still very much open-ended.

The manuscript uses too vague of a language to describe the problems of downscaling and the approaches explored so far. Multiple concepts and methods (Gram-Schmidt orthogonalization, Weibull distribution, etc.)  get only cursory mentioning, with no elaboration on their suitability and effectiveness. More clarity and specifics would greatly benefit the reader.

Hopefully, the review will be of some interest to the wind forecasting community.        

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper reviews the methods and functions used for statistical downscaling methods when conducting local scale wind prediction. It concludes the advantages and disadvantages of different method from simple to sophisticated methods. It also proposed a series of future works needs to be done. My major concern is that all the literatures are focused on local scales, however, the author didn’t mention the scales in each literature. As we know, simulation of wind with scales varying from hundreds of meters to ~10 km may have significant difference. Some high-resolution simulation may also need to incorporate more land features. It would be better if the author can do some further review to make sure the SD methods were applied to a “small scale” with consistent resolution. Therefore, this reviewer would suggest a minor revision due to some knowledge gaps of several definitions and methods. Please see my comments below:

Line 29: developing à developed

Line 31: Is the transfer function here similar with the techniques (regionalization techniques, scale reduction techniques, and/or downscaling techniques) mentioned previously? Please clarify.

In addition, it would be better if the authors can provide some literatures using these techniques mentioned here.

Line 36: can the authors provide more detailed description of the definition of “wind data transference”?

Line 41: please give some examples of “large-scale circulation models and the prediction from these for both large and regional scales”.

Line 50: Here the authors summarized the advantages and disadvantages of Statistical Downscaling and that of Dynamic Downscaling. This reviewer would like to suggest adding some introduction of the functions/roles/significance of DD and SD when studying small scale wind predictions.

Line 57: inconsistency in scales. Line 28 mentions the wind over hundreds of meters scales, Line 57, the authors indicate a wind transfer function on a few km scales. Please make the scale consistent.

There are some logical gaps when stating the definitions and purposes of using different techniques.

Line 62: on the other -> on the other hand

Line 73: differences -> different

Line 134: what is the definition of Weibull expression? A brief introduction of Weibull expression may be helpful.

Line 136: wind probability density functions (PDF)

Line 143: why it is not “simple” to use Kernel interpolation?

Line 199: the authors already discussed the PDF in the 3.3 session, here in 3.4, the authors talked about the PDF again. Is it necessary to separate this part from previous session?

Line 282: can the authors provide information of parameters of Weibull distribution?

Line 293 and 299: again, please list the parameters of Weibull distribution.

Line 332: Conclusions: may briefly summarize the advantages and disadvantages found for SD method.

Line 342: previous statements, the authors proposed the local scale to be hundreds of meters and km scales, however, here the authors indicate 10 km scale. Please be consistent in terms of the local scales.

Author Response

Please see the attachment

Author Response File: Author Response.pdf