Dear Editor and Esteemed Reviewers,

I would like to sincerely thank you for your thorough review and the valuable feedback provided on my submitted manuscript. I truly appreciate the time and effort you have dedicated to evaluating the paper. Every suggestion has been carefully considered, and I have made revisions accordingly, which are clearly marked in the revised manuscript. Below, I provide a detailed response to each of the reviewer’s comments.

We feel we have comprehensively addressed the concerns raised, and will leave how to progress our paper to your judgement. I am looking forward to hearing your positive response soon.

Sincerely.

1. The combination of machine learning and AHP methods for flood risk assessment is already quite common. However, using only a single machine learning algorithm tends to result in poor interpretability of flood risk, leading to uncertainty in the model’s flood risk results. Additionally, further research is needed to efficiently and accurately select the optimal machine learning algorithm for the region.

Automatic machine learning tries to automatize the steps of feature extraction, model and algorithm selection, parameter optimization, and so on so that it needs no human assistance and avoids man-made bias. This approach only requires the configuration of different run times, allowing the algorithm to explore a wider array of model and parameter combinations within the allocated time, ultimately leading to the identification of the best-performing model.

This paper selects the Auto-Sklearn framework to address the binary classification problem of flooded point identification and to calculate flood risk. By utilizing the characteristics of automated machine learning, the efficiency of machine learning is improved, and the importance ranking of flood hazard factors is obtained. The next step is to use the AHP to calculate the relevant weights by combining the flood risk results with exposure, vulnerability, and resilience indicators. AHP aims to quantify the decision-making process by scoring weights according to their level of importance, ultimately yielding the flood risk results.

2. Line 57 - Replace the sentence in line 57 with: " Through continuous improvement and development of machine learning algorithms,ensemble methods have been addressed the limitations of traditional machine learning models. "

3. Line 76 - The meaning expressed in line 76 is inaccurate; a more precise phrasing would be: “The effectiveness of machine learning "automatically improves with experience," and a key challenge in the research is how to integrate the data processing capabilities and feature selection strengths of hybrid models with ensemble models.”

4. Line 178 - The reason for choosing uneducated individuals as one of the indicators is that they are often part of socially vulnerable groups. People with lower levels of education may not fully understand warning information, disaster prevention measures, or have access to sufficient disaster preparedness resources, which increases their vulnerability during disasters. Additionally, those with higher education levels typically have access to more information channels, while individuals with lower education levels may be unfamiliar with new technologies or information channels (such as mobile apps and internet alerts). At the same time, lower-educated groups may live in areas with less developed infrastructure, making it difficult for them to receive timely social aid and support. Uneducated individuals may also find it harder to regain economic independence after a disaster, as they may lack access to technical training or knowledge updates, leading to slower recovery. Although income level is an important factor in assessing vulnerability, average income distribution can sometimes obscure individual differences. For example, a region may have a high average income level, but low-income groups (such as uneducated individuals) can still be in a highly vulnerable state. Additionally, income level may not directly reflect an individual's awareness of disaster, knowledge reserves, or ability to take action.

5. Line 179- Urbanization rate refers to the proportion of the urban population to the total permanent population in a given region, and it reflects the level of urbanization in that area. This indicator has an inverse relationship with flood vulnerability. Generally speaking, the higher the urbanization rate of a region, the higher the level of social development and the capacity for protection, which can reduce flood vulnerability to some extent.

6. Line 187- The flood inventory map in this paper was developed using inundation data from the Global Flood Database and flood disaster data from the EM-DAT database, supplemented by satellite and Google image interpretation and verified against existing historical flood records. The actual flood-affected areas were delineated based on flood traces from the inundation datasets and image interpretations. For this study, 278 flood inundation points were randomly selected within the inundation data range during the study period, and the location of each point was used as the basis for subsequent statistical analysis of flood events.

7. Line 188- For the collected data related to flood disaster risk, differences in sources and formats have resulted in variations in spatial resolution, dimensions, and magnitudes across the datasets. When using these data as indicators to assess risk, it is necessary to standardize the spatial and statistical data. This involves two key aspects: unification of spatial scale and normalization of numerical range.

(1) Unification of spatial scale means aligning data within the same coordinate range and resolution. The research data is standardized through projection transformation, converting all datasets into the same geographic and projected coordinate systems. The Kriging interpolation method is used to spatially process all discrete data. Finally, if the spatial data has different resolutions, resampling is performed to standardize all data to the same resolution, which in this study is unified to 30m×30m.

(2) Normalization of the numerical range can be achieved using a normalization process. Through a linear transformation, the values of the data are mapped to the range [0, 1], thus eliminating the influence of differing dimensions among the data indicators. In this study, the Min-Max Normalization method is used for normalization, and the formula is as follows:

    x^'=(x-min⁡(x))/(max⁡(x)-min⁡(x))    

8. Line 194 - Through experiments, it was found that the effect of using a 1:1 ratio for the training and testing datasets is better than the 1:2 and 1:3 ratios.

9. Line 201- Each label consists of 278 points representing entire dataset.

10. Line 201- The overview map of the study area has been revised to display the spatial distribution of flooded and non-flooded points.

11. Line 231- The text here contains a definition error, which has been corrected in the main body of the paper. Thank you to the reviewer for pointing this out.

12. Line 290- The definition of consistency has been added to the paper: In a pairwise comparison matrix, the decision-maker's judgments must exhibit logical coherence and transitivity. This means that if option A is considered more important than option B, and option B is considered more important than option C, consistency requires that option A must also be judged more important than option C.

13. The λ in Eq. 11 should be λmax as defined in Eq. 10, and this has been corrected in the paper.

14. Line 299 - This sentence has been revised: Where average Random Consistency Index (RI) represents the average random consistency which depends only on the order of the judgment matrix. The RI values for judgment matrices of order 1 to 10 are shown in Table 3.

15. Line 316 - These indicators are provided for the testing data. In response to the reviewer’s suggestion, we will consider removing the accuracy metric and retaining precision, recall, F1-score, and Kappa.

16. Line 319 - The probability thresholds for accuracy, precision, recall, and F1-score range from [0, 1], while the Kappa coefficient ranges from [-1, 1].

17. Line 327 - We used 5-fold cross-validation to assess overfitting by comparing the performance of the training and testing sets. The experimental results indicate that the performance of the training set and testing set is relatively close, suggesting that the model does not exhibit overfitting.

18. Line 368 - This sentence has been revised: SHAP is an explanation method based on game theory and belongs to post-hoc model interpretation methods. SHAP values provide a unified framework that decomposes the model’s output to quantify each feature's contribution to the prediction result.

19.  Figure 8 - Using AutoML and AHP, the flood hazard, exposure, vulnerability, and resilience of the YRDUA were calculated. Based on model-determined weights, flood hazard level (a), flood exposure level (b), flood vulnerability level (c), flood resilience level (d), and flood risk spatial distribution (e) were derived through natural breaks classification in ArcGIS software, resulting in a flood risk zoning map for the Yangtze River Delta region.

20. Line 450 - The differences in flood risk among cities in the Yangtze River Delta over the past few decades are primarily due to a complex interplay of various factors, including geographic and climatic conditions, urbanization processes, socio-economic factors, ecological changes, and historical flood events. The topography and precipitation patterns of different cities affect their capacity for rainwater drainage and accumulation, while urbanization leads to an increase in impervious surfaces and variations in infrastructure development, impacting flood management capabilities. Additionally, differences in population density, economic development levels, and flood management policies can exacerbate flood risk. Furthermore, the increasing frequency of extreme weather events due to climate change further elevates flood risk. These factors determine the varying levels of flood risk among cities within the Yangtze River Delta.

Dear Editor and Esteemed Reviewers,

I sincerely thank you for the thorough review and valuable feedback you provided on my manuscript submission. I greatly appreciate the time and effort you dedicated to evaluating this paper. I have carefully considered each suggestion, and below, I provide detailed responses to the key reviewer comments. For some specific details of the paper, I will make the necessary revisions within the manuscript.

Sincerely.

1. Line 13- Thank you for your correction. The original text has been revised to replace "remains" with "remain."

2. Line 19- The presentation of CatBoost has been completed, and the original sentence has been revised to:“Results indicate that, among those of different assessment models, the accuracy, precision, F1-score, and kappa coefficient of the Categorical Boosting (CatBoost) model for flooded point identification are the highest.”

3. The missing spaces before citations mentioned by the reviewer have been corrected, and all citations throughout the text have been thoroughly checked.

4. The point has been removed. Thank you for your correction.

5. Thank you for your valuable suggestion. We have carefully revised lines 108-114 as follows:

"The comparative analysis of superimposed flooded points data shows a strong alignment between the distribution of flooded points in the study area and the high to medium-high risk areas, highlighting the reliability and applicability of the proposed model. The remainder of this paper is structured as follows: Section 2 introduces the study area, data sources, and methodology; Section 3 presents the results and analysis; Section 4 discusses the findings and their implications; and Section 5 concludes the study with key insights and recommendations."

6. Thank you for your constructive feedback on Figure 2. I have updated the color palette to avoid any confusion with rivers and lakes, and have also inserted an inset map to illustrate the position of the YRDUA relative to the entire China. The Digital Elevation Model (DEM) units of measure, which are indeed meters above sea level (m asl), have been clearly stated. The revised figure is attached in the submission documents for your review. I appreciate your detailed suggestions and hope these revisions meet your expectations.

7. Thank you for your valuable suggestions regarding Table 1 and the description of basic geographic information data. I have taken your feedback into consideration and improved the dataset descriptions by adding more detailed information. Additionally, I have revised the appearance of the table to include very tiny horizontal lines after each data name, which will help distinguish each data name and source more clearly. I appreciate your detailed review and helpful recommendations.

8. Lines 168-183 have been rewritten to provide a brief explanation of the 19 secondary indicators. The revisions are as follows:

The hazard indicators consisted of six indices: Average annual precipitation (PREC), Annual Cumulative Heavy Rainfall Duration (DURA), Digital Elevation Model (DEM), SLOPE, Drainage Density (DD), and Normalized Difference Vegetation Index (NDVI). Rainfall is the primary factor leading to flooding, particularly extreme rainstorms caused by climate change. According to the Meteorological Bureau's definition, a heavy rainstorm event is characterized by rainfall of 50mm or more within 24 hours. DURA is defined as the total number of days with heavy rainstorm events occurring at all meteorological stations within the study area each year. The more days heavy rainstorms accumulate and the longer their duration, the greater the likelihood of flooding and other disaster events. DEM and SLOPE are important topographical indicators. Areas with low DEM and SLOPE values are generally more susceptible to flood threats. DD refers to the area of rivers or lakes per unit of land surface area and is a crucial indicator of a watershed's structural characteristics. It determines the watershed's capacity to withstand flooding. The higher the DD, the greater the likelihood of flooding and the higher the potential flood risk. Vegetation plays a role in water storage, retention, and infiltration. The lower the vegetation coverage, the weaker the buffering capacity, making it more likely for surface water to accumulate. The NDVI index measures the relative abundance of green vegetation, with values ranging from -1 to 1. The higher the value, the greater the vegetation coverage, and the lower the risk of flooding.

Land area (AREA), Population Density (DPOP), GDP Density (DGDP), and Building Density (DBUI) were selected as exposure indicators to assess the degree of vulnerability of both the natural environment and social systems to flooding. The land area for each administrative unit at the prefecture-level city is calculated individually. A larger land area corresponds to a greater extent exposed to flooding. DPOP and DGDP represent the concentration of population and assets, respectively. Areas with higher DPOP and DGDP are more susceptible to potential threats from pluvial flooding. DBUI, the ratio of total building area to total land area in a region, indicates the building density. A higher DBUI reflects greater exposure to flooding.

Vulnerability indicators focus more on the social aspects of flood disasters. This study selects four vulnerability indicators: Proportion of Child Population (PPOP_CHI), Proportion of Elderly Population (PPOP_ELD), Proportion of Uneducated Population (PPOP_UEDU), and Urbanization Rate (UR). Age is a key feature of social vulnerability, and both the population aged 0-14 and those over 65 are considered vulnerable groups, as these age groups are more susceptible to flood damage. The uneducated population generally has a weaker awareness of disaster risks and lower self-protection capacity, which makes this group more vulnerable to flooding. The urbanization rate refers to the proportion of the urban population in the total resident population of a region. This indicator is inversely related to flood vulnerability. In general, a higher urbanization rate indicates greater social development and stronger protective capacities, which can reduce vulnerability to flooding to some extent.

The resilience indicators selected in this study include Gross Domestic Product (GDP) per capita, Unemployment Rate (UEMP), Number of Doctors (DOCS), Number of Medical Institutions (INSTS), and Number of Hospital Beds (BEDS). GDP per capita is the ratio of a region's GDP to its total resident population over a specified period, reflecting the region's economic condition. A higher GDP per capita indicates a more developed economy, which is associated with a greater capacity to recover quickly after a flooding event. The Unemployment Rate (UEMP) measures the proportion of the idle labor force, indirectly reflecting the stability of urban development. A high unemployment rate signals economic instability, which weakens the capacity to cope with floods and extends the time required for post-disaster recovery, thus impeding disaster response efforts. The indicators of DOCS, INSTS, and BEDS provide insights into a region’s healthcare and medical support capabilities. Areas with stronger healthcare systems are better positioned to manage flood risks and recover more effectively from such disasters.

9. Thank you for your insightful question regarding the inclusion of three indicators related to the sanitary sector. These indicators—doctors, medical institutions, and hospital beds—were selected to comprehensively capture the region’s healthcare capacity, which plays a crucial role in resilience during and after disasters. While they all pertain to healthcare, each indicator reflects a distinct aspect: the availability of medical personnel (doctors), the infrastructure of healthcare facilities (medical institutions), and the capacity to accommodate patients (hospital beds). Together, they provide a balanced and multidimensional understanding of the sanitary sector's contribution to flood resilience. Additionally, I appreciate your thoughtful suggestion regarding the use of indicators in the sanitary sector. I agree that incorporating additional indicators such as civil protection forces, law enforcement, and firefighters could provide a more comprehensive assessment of the region's resilience capabilities. I will explore the availability of data on these indicators and consider integrating them during the revision phase of the manuscript. This approach will help enrich our understanding of regional resilience to flooding from multiple perspectives.

10. Thank you for your insightful question regarding the division of the training and testing datasets. Based on your feedback, I have revised this section as follows:

First, in Section 2.2, I provided a detailed explanation of the historical disaster data. The flood inundation data provided by the MODIS-based Global Flood Database (2000-2018) has been cropped to the YRDUA. I also introduced how the historical flood map for the study area was generated, and I have attached (a) a flood inundation map of the study area and (b) the spatial distribution of flooded and non-flooded points in the Yangtze River Delta urban agglomeration.

Next, a new subsection 2.3, titled "Extraction of Historical Flood Inundation Points," was added before the original section "2.3 Establishment of a Flood Risk Assessment Indicator System" to explain the division of the training and validation sets. The original Section 2.3 has been updated to "2.4 Establishment of a Flood Risk Assessment Indicator System," and the order of the subsequent sections has been adjusted accordingly.

11. Thank you for your questions regarding the models used in our study. Yes, "linear" refers to the linear regression model. As for the neural networks mentioned, we specifically employed a Multi-layer Perceptron (MLP). I will ensure that this is clearly explained and detailed in the manuscript. Thank you for bringing this to my attention.

12. Section 2.4: For the collected data related to flood disaster risk, differences in sources and formats have resulted in variations in spatial resolution, dimensions, and magnitudes across the datasets. When using these data as indicators to assess risk, it is necessary to standardize the spatial and statistical data. This involves two key aspects: unification of spatial scale and normalization of numerical range.

(1) Unification of spatial scale means aligning data within the same coordinate range and resolution. The research data is standardized through projection transformation, converting all datasets into the same geographic and projected coordinate systems. The Kriging interpolation method is used to spatially process all discrete data. Finally, if the spatial data has different resolutions, resampling is performed to standardize all data to the same resolution, which in this study is unified to 30m×30m.

(2) Normalization of the numerical range can be achieved using a normalization process. Through a linear transformation, the values of the data are mapped to the range [0, 1], thus eliminating the influence of differing dimensions among the data indicators. In this study, the Min-Max Normalization method is used for normalization, and the formula is as follows:

x^'=(x-min⁡(x))/(max⁡(x)-min⁡(x))

13. Line 231-232 The text here contains a definition error, which has been corrected in the main body of the paper. Thank you to the reviewer for pointing this out.

14. Thank you for your comment on Eq. 5. After reviewing several relevant papers, including Efficient LBP-GLCM texture analysis for asphalt pavement raveling detection using eXtreme Gradient Boost, A weighted metric scalarization approach for multiobjective BOHB hyperparameter optimization in LSTM model for sentiment analysis, and Quantum computing and machine learning for Arabic language sentiment classification in social media, I found that the term "F1-score" is typically used without a subscript in these works. Given this, I believe that it is not necessary to add a subscript in the equation. 

15. Thank you for your attention to detail in reviewing the manuscript. I have updated "judgements" to "judgments" in Table 2 as you suggested. I appreciate your guidance on this matter.

16.Line 299: This sentence has been revised: Where average Random Consistency Index (RI) represents the average random consistency which depends only on the order of the judgment matrix. The RI values for judgment matrices of order 1 to 10 are shown in Table 3.

17. Thank you for your valuable feedback on Section 3.1.1. I agree that including the results from the training phase could provide insightful comparisons. I will add the experimental results from the training set during the revision stage of the manuscript. I appreciate your suggestion and look forward to enhancing the content of the paper by comparing the differences between the training and testing set results.

18. Thank you for your insightful comment regarding Figure 5. I have added an explanation of the micro- and macro-average ROC curves in Section 2.4.3 of the manuscript. This should provide clarity on their meaning and significance in the context of our study. I appreciate your guidance on enhancing the comprehensibility of the paper.

19. Thank you for your valuable feedback regarding the organization of subsections in Section 3.1.2. Based on your suggestions, I will thoroughly review and ensure the logical coherence between the subsections. I plan to improve the overall readability and effectiveness of the section by integrating related content more closely and restructuring information as needed. Once again, I appreciate your constructive comments.

20. Figures 8 & 9: Thank you for your inquiry regarding the resolution of the rasters used in Figures 8 and 9. The raster data for these figures was obtained at a resolution of 30m x 30m.