My major concern is still the applicability of the filters proposed by the authors to different study areas. The algorithm bases on the fact that salinity is constant below 1500 m and so, the salinity drift can be computed and corrected. While this might be true for certain regions it is not true for all of them and the authors need to be clear on the fact that the user should know if the salinity varies below that level within the desired study area.
Giving an option to the user to choose the reference level when using the filter, if possible, would be great, even if the default can still be the 1500 m.
I still think that some figures are not very useful. Particularly the ones that show only region delimitation like 2 and 8. Keeping only one of them would be a better option in my opinion.
Lines 4-5: The authors say in the abstract “This work shows a methodology to filter profiles within a given polygon using the odd-even algorithm, this allows analysis of a study area, regardless of size, shape or location”. As I said, this is not precise, here the authors should make clear that the user need to ensure that in the study area salinity doesn’t change below 1500 (or let the user to choose the reference salinity)
In my previous review I made this comment: “I find that the aim of this study is interesting and the tool the authors provide is useful. However, it presents one main problem that is the applicability of the tool in any polygon of any area chosen by the user. I don’t think that the tool can be used globally and so, its limits need to be specified by the authors.”
And the authors responded: “Suggestion accepted (lines 208-210)”
….But I don’t see the answer to my concern in these lines. Is this a mistake in the line numbering?
Line 6: sentence too long and unclear. Please rewrite
110-115: And where can this analysis be seen? The authors talk about this later in the paper and I think that they should focus in the later occurrence of this topic and delete it from here (or just mention it quickly). By reading this paragraph, the reads wonders what I asked at the beginning of this comment, which I made before reading the whole paper.
117: This paragraph fits better in the section “web application”
Lines 148-150: The authors said: “Taking into consideration that at depths greater than 1500 m, the variations in salinity and temperatures are imperceptible, the cluster analysis was performed with the salinity data measured at depths greater than 1500 m”
Again, this is not true. I had asked the authors to show this is true or otherwise specify very clearly in the paper that the selected region must be checked by the user in order to ensure that below 1500 m the salinity differences within a given region are not important enough, and that the algorithm can be applied.
Figure 3: The difference between the two yellow and the two purple colors is imperceptible. Please chose more different tones of color or two different colors. Also, Argo data seem deeper (colder) than WOA, which is unlikely since Argo only reaches 2000 m and WOA data contains deeper data, so I think that the legend is the other way around (yellow for Argo and purple for WOA), please check this.
157: “To manually avoid indicating the number of k centroids” change for “To avoid indicating the number of k centroids manually”
180: I suggest to avoid referring to the user as “the researcher” since many Argo data user are not researchers. “ the user” would be maybe a better option.
205: I still think that to much attention is being paid to the PIP which is a very simple task. The number of vertices is really irrelevant when choosing data inside a closed contour in Matlab or Phyton. However, I agree that being able to select an irregular polygon is useful for the web interface.