Overall, the quality of the paper is good and the work is interesting (although not entirely novel, subject to academic trends). Some case studies received more attention than others, leading me to believe that maybe the paper could have benefitted from a more narrow focus on case studies (1-3), to allow space for more in-depth exploration of contextual issues. Nonetheless, it's refreshing to see a practical take on the issue of cascading/multi-hazard risks – a body of knowledge that seems to be populated by frameworks and concepts nobody uses. However, some edits would be required before publication. 

1. Does the paper address relevant scientific questions within the scope of ESD? Yes 
2. Does the paper present novel concepts, ideas, tools and data? Not novel, but the perspective is interesting. 
3. Are substantial conclusions reached? Partly – the limitations of available data should be highlighted more clearly. I understand the emphasis on stakeholders' perspectives, but they provide a flimsy ground to make sweeping claims (see point 4). 
4. Are the scientific methods and assumptions valid and clearly outlined? Partly. The issue of terminology is acknowledged vastly too late (section 4.1), and the approach to handing stakeholders a glossary clearly turned out to be counterproductive. The author's initial approach risked reinforcing gaps between expert discourse and practice, rather than exploring and assessing differences in interpretation. To make scientific knowledge useful, aligning scientific and local knowledge should be a priority. This should be highlighted earlier in the paper. Secondly, the readers would benefit from understanding what countries were involved in regions described. Considering the geographic, political, and socio-economic diversity of the Danube Region in particular, it is difficult to interpret interview responses in the absence of contexts (especially when the authors make claims regarding "fragmented risk management systems which limit effective responses"). Fragmented systems where? Fragmented how/why? Evidenced by what/whom? 
5. Are the results sufficient to support interpretations and conclusions? Partly, but more in-depth discussion is required. For example, the missing references to socio-economic conditions in the context of vulnerability is surprising – particularly when the data suggests this specifically (e.g., Danube and the "maintenance funding constraints" underpinning infrastructural vulnerability). These are not some random system changes, but rather reflect politico-economic decision making that (usually for the benefit of the free markets) seeks to cut public funding for services. This applies to social vulnerability as well: people do not choose to live on unsafe land, in poor quality housing, or choose not to build back better. They do so because they tend to have limited financial means (by-product of economic arrangements that are, again, not without intention). It is as if the authors go out of their way to avoid mentioning economic issues, rather choosing to point to "systemic transformations" which leave the reader to decide what the issue is. However, this is an academic article, not a Rorschach's test. 
6. Is the description of experiments and calculations sufficiently complete and precise to allow their reproduction? Yes (although the paper would benefit from including the codes used as an annex). 
7. Do the authors give credit etc. Yes 
8. Does the title clearly reflect the contents? Yes, although the bridging appears to come in as an afterthought (the issue of section 4.1). 
9. Concise and complete summary? Yes. 
10. Overall presentation: well structured and clear? Yes 
11. Is the language fluent and precise? One minor issue –  the more commonly acknowledged term for "long/fast duration events" is slow/fast onset event/disaster. Also in the beginning of the paper, it perhaps would be best to start with risks are, rather than "risk is". 
12. Are mathematical formulae, symbols, abbreviations, and units correctly defined and used? N/A, however the usage of vulnerability concept remains rather inconsistent. Firstly it notes vulnerability as a process, and the goes onto box 1 to provide a neat categorisation for the purposes of discussion. Yet, social vulnerability (for example) is not easy to separate from economic vulnerability, and economic vulnerability tends to produce physical vulnerability (e.g. quality of housing). Whilst tidy for analysis, the point of the paper is to emphasise how messy and interconnected these issues are. The authors should revisit the results section with this in mind (without forgetting the issue of economy!). 
13. Should any part of the paper be clarified, reduced, combined, eliminated? See above. Also, the inclusion of all case studies has been done for the sake of the project, but for the sake of the paper and detail, focusing on 1-3 may be more beneficial (if feasible). 
14. Are the number of quality references appropriate? Yes. 
15. Is the amount and quality of supplementary material appropriate? Would benefit from the inclusion of codes 

Other minor thoughts: p7, droughts are generally a water management problem as much as they are an environmental one https://www.mdpi.com/2071-1050/11/2/308. Does not hurt to mention this. 

P12 and discussions regarding levees: infrastructural measures themselves can create a false sense of security and therefore becoming counterproductive to DRR . Could be an interesting issue to explore. 

General Comment

Overall, this is a well-written, polished, clear and important piece drawing on stakeholder insights from various case-studies (“pilot regions”) to discuss multi-risk contexts and their complexity in those regions. Given the need for both highly context-sensitive DRR and DRM processes on the one hand, and a generalizable approach towards better understanding and dealing with an increasingly complex disaster risk context, the article provides important input for both the academic study and the practical handling of multi-hazard risks. An aspect that the authors may further develop is how they introduce the various stakeholders and the respective risks they face. As they point out, risks are multi-faceted and complex, and affect people in different ways. However, this differentiation of who is affected by what kind of risk, why that risk matters and for whom, could be highlighted a bit further (see comments below for specifics). Accordingly, I recommend the article be published after some minor revisions. 

Individual Scientific Comments

From line 244 (3. Results), onward: It would be interesting to understand better who is actually dealing with what kind of hazard/risk. When the authors mention the Danube region (line 274), which is arguably an immense space (compared to e.g., Veneto), it would be interesting to know where and how the hazard interactions are complex to manage, and for whom. This also matters, since NGOs, academics, city administrators, and local municipalities may all have different conceptions of what makes a risk, and why that risk matters. While this is most obvious in the Danube region (given its size), similar things can be said about the North Sea, and the Canary Islands. While it makes sense that the authors only focus on a specific risk for a specific number of stakeholders in one publication, it would be useful to contextualize those risks with the various actors involved, rather than leaving the hazard “speak for itself”.

On a more general level (and granted, the authors somehow mention this in the discussion/limitations section of the article), it may be interesting to reflect a bit more on the generic value of assessing and comparing highly heterogeneous set of stakeholders and regions to one another. Not that it may not have epistemic value to do so – but it may be interesting to highlight a bit more, why the information garnered from these interviews is valuable for DRM/DRR development beyond the fact that it is complicated and context-sensitive (a fact, that the article expresses nicely). 

Technical/Editorial Comments

-    Should “in progress” be cited? (line 135)

-    Double period (line 456)