Dear Marcello, Mario and Rui,

Thanks a lot for your responses to the 2 reviews. As you have seen, both reviewers have severe concerns with your part 1 manuscript. However, I agree with your view that including part 1 as theory chapter into the part 2 paper would lead to a very long manuscript. I hope that, in the end, you can convince the reviewers that part 1 justifies an own paper.

Further, each piece of your 2 companion papers should stand by itself, in the sense that a reader who is only interested in the theory should not be forced to read part 2 and vice versa. I understand that requires a delicate balance, but I ask you to consider this aspect in your revision.

Further, I ask you to have a careful look on conciseness and clarity during your revision. For example, the second sentence of the abstract contains "... the connections, interdependencies and interactions between exposed elements ...". I wonder whether the 3 terms connections, interdependencies, interactions mean 3 different things. If yes, then it is okay to use them (but then explain what the differences are), if not, then use only 1 (or 2) terms. Given that the title of part 1 would possibly attract many readers with different expertise and background knowledge, utmost conceness and clarity are needed. Further, I feel that you tend to use very long paragraphs. I would split long paragraphs into shorter ones, where each paragraph discusses a specific issue or message. Again, I feel that this would fecilitate understanding.

Please go ahead with revising the manuscript as you have proposed in your responses. I will try to motivate the 2 reviewers to assess also the updated version.

Best regards, Bruno Merz

Dear Marcello, Mario and Rui,

the two reviewers of the first version have also reviewed the revised version. As they came to conflicting assessments, I have invited a third reviewer to assess the revised manuscripts. Based on these three reviews, the manuscripts (part I and II) need still major revision before they can be published. The most important concerns (relating to both parts) are:

- Can both parts stand on their own?
Reviewers #2 and #3 both think that there is not enough substance for two articles.

- Is the content really that novel as you claim?
Reviewers #1 and #3 argue that the novelty is oversold, and that earlier important work has not be cited.

- What exactly is the scientific contribution of the two manuscripts?
For example, reviewer #1 asks what does graph theory give us that other analyses do not, and argues that one needs to clearly articulate what it specifically offers that is different from other holistic risk assessment approaches. Similarly, reviewer #2 argues that it is not yet clear what the manuscripts add to the literature.

- Conciseness in terms of concepts and definitions
Reviewers #1 and #3 feel that concepts and definitions may be mixed up or not well explained. Reviewer #3 proposes to add equations for some basic concepts/network measures to have more concise definitions.

Finally, I want to stress that I agree with the reviewers. Reading the two parts leaves me with the feeling that the manuscripts on the one hand claims to provide a novel method, which potentially allows holistic risk assessments, whereas on the other hand, it cannot see how this method could achieve that claim. Besides working on the clarity and conciseness, I recommend to moderate the tone and really strive to reflect on the advantages and possible downsides of the method. (I assume that the method will not solve all problems.)

Best regards, Bruno Merz

Dear colleagues,

2 former reviewers have again assessed your manuscript. Although they see value in the 2 papers (Part I and II) and they acknowledge improvements in relation to the earlier versions, they both recommend rejection. Both argue that the contents and novelty of the 2 pieces do not suffice to publish 2 papers. Having read, again, through the 2 manuscripts, I agree with them and propose to merge the 2 parts.
One can shape the merged paper in different directions: Either put the focus on Part I and use Part II as real-world case study (which seems to me the preferred option), or use Part I as extended literature review of Part II (as reviewer #1 suggested).
Merging the 2 parts could result in a rather long paper, but with merging quite some contents can be deleted which is now repeated in Part II to be self-standing. Further, some of the figures can probably be merged or be provided as supplement.
Best regards, Bruno Merz

Dear colleagues,

Thanks you very much for the revision and the merging of the 2 manuscripts. Referee #1 has again evaluated the manuscript and acknowledges that it has been further improved. He/she has still 3 major comments, and I ask you to carefully consider these coments in a hopefully final round of revisions.

In addition, I ask you to check where the manuscript could be shortened. For example, chapter 1 is rather wordy and I think it could be shortened without losing contents, e.g. by deleting the quote of Aristotle or Lhomme et al. (2013) ["... showed that the “city has to be considered as an entity composed by different elements and not merely as a set of concrete buildings...”].
Another example: The first paragraph of section 2.2.1 ["... The mathematical properties of a graph can be studied using Graph Theory (Biggs et al., 1976), which as mentioned above could provide a framework to assess risk from a holistic and systemic viewpoint. This section summarizes some of the main concepts of Graph Theory on which the proposed methodology, presented in Section 3, is based..."] could probably be written as "... The mathematical properties of a graph can be studied using Graph Theory (Biggs et al., 1976); its main concepts are summarized as follows....".
These are just 2 examples, and I think that a rigorous check could significantly shorten the paper.

Best reagrds, Bruno Merz

Dear colleagues,

I am pleased to inform you that the revised version of your manuscript NHESS-2013-26 can now be accepted for publication in our journal Natural Hazards and Earth System Sciences (NHESS). Thank you very much for this final revision.

Best regards, Bruno Merz