A Southeastern United States Warm Season Precipitation Climatology Using Unsupervised Learning

Round 1

Reviewer 1 Report

Very good work done by the authors, the methodology has been clearly presented as well as the discussions of the results and their conclusions. The relevance of the results obtained is indisputable and it is expected that they can be transferred to society.

I would just subtract, I would assess if there are any monthly or quarterly trends that were significant that would surely come through in further analysis.

Author Response

Thank you for your thoughtful review.  Given our short 3-month time period we felt that quarterly patterns would not provide additional insight.  Future work could look at monthly variability, but we anticipate the patterns to remain fairly stable by month outside of the randomness of tropical cyclone landfalls, which is why we elected to use the full warm season.  We added a sentence in the future work section to address the inclusion of monthly patterns in future work.

Reviewer 2 Report

It’s a good piece of work but not complete in the current version (see the #2 below). Therefore, a major revision is needed.

(#1) The use of KPCA-based clustering method for nonlinear situation is a good try.

(#2) In the Discussion section, it is also insightful that the authors related the three clusters to ENSO and TC activities. However, what kinds of “average” synoptic situation (atmospheric general circulation) are related to the three clusters are missing. I suggest adding three composite synoptic weather maps by averaging the cluster members in each cluster in the same section, as Figs. 11, 12 and 13, as well as a discussion of the new figures.

Author Response

Please see attached.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have addressed my concerns in this revision by adding and discussing the new figure 9.  Thanks.