Dear Anonymous Referee #1,

Thank you for taking the time to review our manuscript and provide feedback. We respectfully address your comments as follows:

1. "[The paper] does not present a systematic methodology to address the research question and relies solely on basic statistical analysis based on the collected data."

We respectfully disagree with this assessment. Our study presents a clear and systematic methodological strategy to explore the relationship between health outcomes and key explanatory variables – namely, community resilience indicators and Disaster Risk Management (DRM) cycle stages – while controlling for potential confounders. This approach is consistent with established empirical practices aimed at reducing omitted variable bias in observational studies. Moreover, we contest the characterization of our statistical analysis as "basic." While our primary estimation strategy is based on Ordinary Least Squares (OLS), this choice was made deliberately, as it is well-suited to the data structure and research question. Also, our OLS approach is comprehensive in a targeted manner given the research question. Our methodology includes construct validation, descriptive analysis, and multivariate regression modeling. We also explored more advanced estimators; however, due to sample size constraints, those models yielded unstable results, which we clearly acknowledge in the manuscript. We maintain that the strategy adopted is both statistically sound and appropriate for the empirical setting.

2. "Given the substantial differences among the seven countries—across social, economic, and infrastructural dimensions—the use of such simplified statistical methods lacks the robustness required to account for these complex factors."

We acknowledge that cross-country heterogeneity poses an important analytical challenge. To mitigate this, we adopted several strategies. In addition to including controls for demographic and hazard/exposure-related factors, we adjusted standard errors using the classification developed by Chapagain et al. (2024), which captures patterns in resilience across different community profiles. While it is not possible to completely eliminate unobserved heterogeneity, these steps represent a careful and methodologically sound effort to account for underlying structural differences. Our study also relies on the Flood Resilience Measurement for Communities (FRMC), a well-established framework designed to generate empirical insights into resilience in diverse settings (Keating et al. 2017; Hochrainer-Stigler et al. 2020; Laurien et al. 2020; Keating et al. 2025). The FRMC uses a consistent set of indicators across all study sites, ensuring comparability, but allows flexibility in how data are gathered to reflect local realities. Data collection is carried out in partnership with local stakeholders, ensuring that the perspectives and experiences of communities are embedded in both the process and outcomes (Hochrainer-Stigler et al. 2021). This combination of standardization and adaptability supports meaningful cross-site comparisons. Chapagain et al. further demonstrated that the FRMC successfully identifies distinct community types through statistical clustering, confirming its ability to capture variations in flood resilience across different contexts.

In light of these considerations, we kindly request that the editor take this clarification into account when assessing the manuscript for further consideration.

References

Chapagain, D., S. Hochrainer-Stigler, S. Velev, A. Keating, J.H. Hyun, N. Rubenstein, and R. Mechler. 2024. A taxonomy-based understanding of community flood resilience. Ecology and Society 29(4). https://doi.org/10.5751/ES-15654-290436

Hochrainer-Stigler, Stefan, Finn Laurien, Stefan Velev, Adriana Keating, and Reinhard Mechler. 2020. “Standardized Disaster and Climate Resilience Grading: A Global Scale Empirical Analysis of Community Flood Resilience.” Journal of Environmental Management 276 (December):111332. https://doi.org/10.1016/j.jenvman.2020.111332.

Hochrainer-Stigler, S., S. Velev, F. Laurien, K. Campbell, J. Czajkowski, A. Keating, and R. Mechler. 2021. Differences in the dynamics of community disaster resilience across the globe. Scientific Reports 11(1): Article 17625.

Keating, A., K. Campbell, M. Szoenyi, C. Mcquistan, D. Nash, and M. Burer. 2017. Development and testing of a community flood resilience measurement tool. Natural Hazards and Earth System Sciences 17(1): 77–101. 

 

Keating, A., S. Hochrainer-Stiger, R. Mechler, F. Laurien, N. Rubenstein, T. Deubelli, S. Velev, M. Szoenyi, and D. Nash. 2025. Reflections on the large-scale application of a community resilience measurement framework across the globe. Climate Services 38. https://doi.org/10.1016/j.cliser.2025.100562. 

Laurien, Finn, Stefan Hochrainer-Stigler, Adriana Keating, Karen Campbell, Reinhard Mechler, and Jeffrey Czajkowski. 2020. “A Typology of Community Flood Resilience.” Regional Environmental Change 20 (1): 24. https://doi.org/10.1007/s10113-020-01593-x.

Response to Reviewer #2

Thank you for an interesting read. We thank the referee for carefully reading our paper and for the thoughtful comments and suggestions. Below, we respond to each point individually. 

1.    Abstract: Please add Data collection and analysis. Main findings and contribution. 

We added the following sentences to the abstract:

“Data was collected using household surveys, community focus groups, key informant interviews, and secondary sources. We applied a quasi-experimental regression design, controlling for demographic and hazard exposure variables, to estimate the effect of resilience and DRM stages on health outcomes.

[…] 

This study finds that preparedness is the most consistent predictor of positive health outcomes, while some forms of resilience may not translate into reduced mortality. By combining resilience capacity, DRM stages and health indicators, this paper contributes to bridge a gap in the literature and offers policy-relevant insights for improving community-level disaster response.”

2.    The DRM cycle has some limitations and shortcomings as well. Please mention those as well. See https://doi.org/10.1016/j.uclim.2021.100893. See section 2.2 on the paper. 

We have introduced a new section (Section 2.2) focusing on the stages of the Disaster Risk Management (DRM) cycle and incorporated the study by Rana et al. (2021), which critically examines its limitations:

“However, the DRM cycle also has well-documented limitations. Scholars have criticized its continuous cyclic nature and broad phase definitions, which can hinder application across diverse urban settings and complicate its integration into climate change and resilience discourse (Rana et al., 2021). While acknowledging its imperfections, the DRM cycle continues to be used due to its convenience and robustness (Alexander, 2018). While debates continue on how to adapt it for more effective management – taking into account time, resources, preferences, capacities, needs, and institutional changes – its practical benefits continue to support its broad use (Baas et al., 2008).”

3.    Please add operational definitions of terms used.  

Thank you for this helpful suggestion. Although the reviewer did not specify which terms required operational definitions, we selected one that we believe lacked clarity in this regard: the DRM Cycle Stages. Additionally, operational definitions for all variables are provided in Table 1. We have now included a concise operational definition for the DRM cycle variables in the revised manuscript as follows:  

“Each DRM stage – Corrective Risk Reduction, Prospective Risk Reduction, Preparedness, Response, and Recovery – is represented by a composite score derived using Principal Component Analysis (PCA). For each stage, we grouped a set of FRMC indicators based on expert assessment from the Zurich Flood Resilience Alliance, reflecting theoretical and practical relevance. The first principal component extracted from each group is used to quantify the respective DRM stage. The Corrective Risk Reduction score is based on the following indicators: risk reduction investments; flood exposure awareness; asset protection knowledge; governance awareness; priority managed units; natural habitat restoration; household flood protection; large-scale flood protection; community representative bodies; social inclusiveness; and integrated flood management planning. The Preparedness score includes indicators on business continuity; household income continuity strategy; evacuation and safety knowledge; first aid knowledge; early warning systems (EWS); flood emergency infrastructure; community participation in flood-related activities; external flood response and recovery services; inter-community flood coordination; and national forecasting policy and plan. The Prospective Risk Reduction score draws from indicators on conservation budget; education commitment during floods; future flood risk awareness; environmental management awareness; natural capital condition; priority natural units; natural resource conservation; community disaster risk management planning; and local leadership. The Recovery score includes household asset recovery; community disaster fund; provision of education; flood energy supply; and community safety. Lastly, the Response score is constructed using indicators related to disaster response budget; water and sanitation awareness; flood healthcare access; transportation interruption; communication interruption; flood emergency food supply; flood safe water; flood waste contamination; and community structures for mutual assistance.”

4.    Can be deleted "The paper is structured as follows: Section 2 outlines the conceptual frameworks, focusing on the FRMC five-capitals model and the DRM cycle stages. Section 3 reviews literature on factors influencing flood-related mortality and morbidity, including demographic characteristics and flood exposure/hazard. Section 4 describes the FRMC data used for analysis. Section 5 details the methodology applied. Section 6 presents the study results, while Section 7 discusses the key findings and limitations of the study."

We have incorporated this change in the revised version.

5.    The FRMC section is too long. Can be shortened. Add DRM Cycle para here as well.

We have significantly shortened the data section, streamlined the description of the FRMC, and incorporated a paragraph on the DRM cycle as suggested. 

6.    Please add how factors (modifiers) can influence/contribute to flood morbidity. DOI: 10.7189/jogh.12.11007

We added an explanation of how climate-related effect modifiers contribute to flood mortality, drawing on Lan et al. (2022), in Section 3.2.

"In addition to flood intensity and return period, climatic variables may also act as effect modifiers that shape health outcomes following flood events. Certain environmental factors can significantly influence the relationship between floods and diarrheal morbidity. In an empirical study conducted in Sichuan Province, China, three effect modifiers were identified that amplify the impact of flooding on diarrheal outcomes: elevated air pressure, reduced diurnal temperature range, and higher ambient temperatures (Lan et al., 2022)"

7.    Some paragraphs are too small. Either elaborate, merge or delete. For example, "Research suggests that the relationship between flood-related morbidity and mortality and age, and gender is complex as exposure and vulnerability factors often interact, with their impact on health and mortality frequently depending on the context." See others as well. 

We have substantially revised several short paragraphs, either merging or expanding their content.

8.    Urban-rural linkages are very important and often overlooked. It can be further polished by the works of Jamshed and Birkmann, who linked urban-rural linkages with floods. Their work is seminal in this context. 

We thank the reviewer for the suggestions and have incorporated the following references accordingly:

Jamshed, A., Rana, I. A., Mirza, U. M., and Birkmann, J.: Assessing relationship between vulnerability and capacity: An empirical study on rural flooding in Pakistan, Int. J. Disaster Risk Reduct., 36, 101109, https://doi.org/10.1016/j.ijdrr.2019.101109, 2019.

Jamshed, A., Birkmann, J., Feldmeyer, D., and Rana, I. A.: A Conceptual Framework to Understand the Dynamics of Rural–Urban Linkages for Rural Flood Vulnerability, Sustainability, 12, 2894, https://doi.org/10.3390/su12072894, 2020a.

Jamshed, A., Birkmann, J., McMillan, J. M., Rana, I. A., and Lauer, H.: The Impact of Extreme Floods on Rural Communities: Evidence from Pakistan, in: Climate Change, Hazards and Adaptation Options: Handling the Impacts of a Changing Climate, edited by: Leal Filho, W., Nagy, G. J., Borga, M., Chávez Muñoz, P. D., and Magnuszewski, A., Springer International Publishing, Cham, 585–613, https://doi.org/10.1007/978-3-030-37425-9_30, 2020b.

Jamshed, A., Birkmann, J., McMillan, J. M., Rana, I. A., Feldmeyer, D., and Sauter, H.: How do rural-urban linkages change after an extreme flood event? Empirical evidence from rural communities in Pakistan, Sci. Total Environ., 750, 141462, https://doi.org/10.1016/j.scitotenv.2020.141462, 2021.

"Some studies also address this gap by showing how floods affect rural vulnerability through disruptions in these linkages. Rural areas, particularly in developing countries, depend heavily on cities for jobs, services, and information, while cities rely on rural areas for labor, food, and ecosystem services. However, these interdependencies are rarely considered in disaster risk frameworks (Jamshed et al., 2019, 2020a, b, 2021). Their conceptual framework demonstrates how floods alter the flows of people, goods, finances, and information between rural and urban areas. In some cases, floods increase rural dependence on cities – for example, through heightened migration, financial support from urban relatives, or greater reliance on urban markets and information. In other cases, dependence may shift toward nearby rural areas when access to cities is constrained due to damaged roads or inflated prices. These changes are shaped by a range of social, economic, infrastructural, spatial, and environmental factors, such as household education, income, city proximity, and exposure to flood sources. Ultimately, the nature of these shifting linkages directly affects how vulnerable rural communities are to future disasters – sometimes enhancing resilience, other times deepening risks. In sum, their work shows that urban-rural linkages are dynamic and central to understanding disaster impacts. Ignoring them risks overlooking critical drivers of vulnerability and resilience."

9.    Again, short paragraphs in section 7. 

We appreciate the careful reading of this section and have substantially revised it, merging several shorter paragraphs where appropriate.

10.    Comparison with similar studies is missing. Please compare the results for or against arguments. Add references in the discussion section.

We appreciate this important suggestion. Whenever possible, we have connected our results to existing literature. However, due to the novelty of our empirical strategy and dataset, it was not always feasible to provide direct comparisons for all findings.

11.    Some references have missing information. Please recheck all references.

Thank you. We had extensively revised the reference list for missing information.