Dear authors,

Thank you for an elaborate response to the issues raised by the two reviewers and me. I tend to agree with your responses, although I cannot anticipate the full picture with all the proposed modifications. I invite you to submit a thoroughly revised manuscript. Given that the responses imply quite some major modifications, I would like the revised manuscript to be reviewed afresh, and be checked on internal consistency.

Pieter van der Zaag

In general I agree with the way in which the authors have addressed the comments by the two reviewers and the editor. The result is that large parts of the text have been modified, and that the argument has been made more straightforward, yet the essence of the original paper has remained.

In my view the paper is nearly ready for acceptance. However, I have still the following concerns that need to be addressed:

1. I do not find the new names of the concepts of “unbroken dependency” (was “intervened”) and “broken dependency” (was “continuous”) felicitous, and even counter intuitive. (Isn’t a broken dependency a dependency that has been broken, i.e. a dependency that no longer exist? So for me an utterly confusing combination of words). So I urge the authors to find a more straightforward nomenclature. When reading the revised document, I came up with “hidden” and “open” dependency. In fact the manuscript frequently equates unbroken dependency with hidden dependency in the text (6 times, on pages 2, 3, 9, 13 (2x) and 14), apparently because “hidden” clearly conveys its meaning.

2. The definitions of dependency must be formalised and unambiguously defined (Table 1 does not provide definitions, merely descriptions). I propose something as follows (using the definitions of discharge runoff as in Table 1; but please check whether my rendering is correct):

No dependency: local demand < local runoff OR local demand > actual discharge
Dependency: local runoff < local demand < actual discharge
Unbroken (hidden) dependency: local demand < actual discharge
Broken (open) dependency: local demand < natural discharge

3. Section 2.2.1 describes the method of expressing water availability, but it only deals with the spatial dimension. The temporal dimension is entirely missing. So the reader can only guess what temporal resolution is used. This must be included.

Related to this, I would add a sentence on page 7 lines 210-211, namely that these annual values may mask water scarcity during the dry season(s)

4. I find Fig 2 confusing and ambiguous with respect to applying the SBA concept: are there three SBASs (“upstream” “middle stream” and “downstream”) or are there four SBAs (OdSBAcz, OdSBApoa, OdSBAge, and OdSBApob)? I guess you mean the latter, if so omit the former. If not, clarify and avoid ambiguity.

5. I would combine the last sentence of section 2.2.2 (p. 7 lines 346-347) (“As illustrated …. in Fig.4.”) with lines 226-227 on the same page, adding a clarifying remark, for example thus:

“The water scarcity status was categorized as No scarcity (N) and Scarcity (S) using average annual water availability from 1981 to 2010. We now can define the earlier defined four different dependency categories in terms of scarcity an SBA can face: NNN, SNN, SNS, and SSS. This is shown in Fig.4.”

6. Fig 3: in the caption the concept of “water requirements" is used, but this is not defined in the text. Do you simply mean “WW”. Please be consistent and parsimonious with concepts! Also Falkenmark’s threshold value should have the unit of m3/cap/year.

7. The last paragraph of section 4.1 (p. 13, lines 400-410) could also refer to the no harm principle (article 7) of the UN Watercourses Convention of 1997.

8. The last two sentences of section 4.2 seem to contradict each other. Please clarify.

Kindly address the above and re-submit, together with an explanation how the above concerns have been addressed.