Journal of Flood Risk Management最新文献

{"title":"Assessing the Riverine Flood Forecast Skill of GloFAS and Google Flood Hub With Impact Data and Discharge Observations to Support Early Actions in Mali","authors":"Els R. Kuipers,&nbsp;Valentijn W. Oldenburg,&nbsp;Phuoc Phung,&nbsp;Edwin H. Sutanudjaja,&nbsp;Andrea Ficchì,&nbsp;Peter Salamon,&nbsp;Marc J. C. van den Homberg","doi":"10.1111/jfr3.70247","DOIUrl":"https://doi.org/10.1111/jfr3.70247","url":null,"abstract":"<p>Riverine floods are among the most destructive and frequent natural hazards in Mali. To reduce their impacts, the Mali Red Cross implemented an Early Action Protocol (EAP) to enable anticipatory actions through pre-defined triggers and forecast information. Currently, the protocol relies on upstream water levels from the National Directorate of Hydraulics (DNH) to predict downstream flooding. However, this model does not consider meteorological forcing, limiting lead times to a maximum of 4 days. Recent advancements in global flood forecasting systems present opportunities to enhance Mali's EAP by leveraging increasingly skilful medium-range weather forecasts as inputs of both physics-based models, as in the Global Flood Awareness System (GloFAS), and AI-based models, as in Google Flood Hub (GFH). We compare GloFAS v4.0 and v3.0, GFH, and Mali's current trigger model using discharge observations and district-level impact data derived from multiple sources and text-mined news-articles. Model performance was assessed for a range of lead times and discharge thresholds. GloFAS and GFH demonstrate sufficient skill for early action beyond 4-day lead time in frequently flooded regions and have a larger spatial coverage compared to the current trigger model, suggesting early action plans could operate with 7-day lead time and span a larger area. Overall, this study highlights the potential and challenges of flood forecasting for anticipatory action in flood-prone, data-scarce regions. In particular, we (i) assess the usability of two different ground truths (observed discharge and impact data) for forecast validation; (ii) assess the possibility of extending Mali's current trigger model in lead time and spatial coverage and (iii) evaluate user-oriented forecast skill across models and contexts.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70247","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994056","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparative Analysis of Urban Flood Mapping Using Sentinel-1 SAR Imagery and Machine Learning","authors":"Md Abdullah Al Mehedi,&nbsp;Virginia Smith,&nbsp;Peleg Kremer","doi":"10.1111/jfr3.70246","DOIUrl":"https://doi.org/10.1111/jfr3.70246","url":null,"abstract":"<p>Flood mapping using remote sensing data has the potential to play a crucial role in monitoring and assessing flood events, particularly in urban environments, where complex infrastructure creates significant challenges. While Synthetic Aperture Radar (SAR) has promising potential for flood mapping, current methods struggle in the urban landscape due to the presence of heterogeneous surfaces and complex infrastructure, which can create double-bounce effects and alter reflectance. This research advances urban flood mapping through a comprehensive comparative analysis of traditional change detection and Machine Learning (ML) methods, including the MultiSenseRF multi-source feature-fusion workflow. The study analyzes five geographically distinct urban areas—Houston, Texas, USA, Dhaka, Bangladesh, Iwaki, Japan, Lumberton, North Carolina, USA, and Beira, Mozambique—during flood events to evaluate method performance across diverse urban settings. Traditional change detection methods and ML techniques were evaluated alongside integrated approaches. Combining temporal change-detection signals with landscape features using ML improved flood-mapping performance, with average F1 scores of about 0.75 and a range of 0.58 in dense urban areas to 0.93 in peri-urban regions. These findings provide a robust framework for improving urban flood mapping accuracy and have significant implications for urban flood management.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70246","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148993974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Linking Landscapes, Flood Risks and Communities: Flood Susceptibility Assessment of Connemara and Aaran Islands, Galway, Ireland","authors":"Beulah Evelyn Lazarus,&nbsp;Mahdieh Raji,&nbsp;Salem Gharbia","doi":"10.1111/jfr3.70253","DOIUrl":"https://doi.org/10.1111/jfr3.70253","url":null,"abstract":"<p>Atlantic-facing coastal communities of western Ireland face recurrent flood hazards driven by intense precipitation, storm surges, and tidal influences. Existing national flood mapping efforts provide valuable hydrodynamic simulations but rarely integrate multi-criteria decision analysis with data-driven machine learning to capture local-scale flood susceptibilities. This study provides a first integrated Analytical Hierarchy Process (AHP) and Machine Learning (ML) for Atlantic-facing of Connemara and Aaran Islands, County Galway. The integrated AHP-ML approach outperforms standalone ML models by integrating transparent, expert-driven weights that reflect local geomorphology and data constraints often missed by non-linear ML algorithms. Nine flood influencing parameters: rainfall, elevation, slope, curvature, land cover, distance to rivers, soil type, and normalised differential vegetation indexand four ML models (Logistic Regression, Random Forest, XGBoost, and Support Vector Machines) are considered for the study. The model performances were qualified using the area under the receiver operating characteristic (AUC), precision and recall. The probabilistic benchmarking validation was also performed against the GloFAS 100-year flood hazard product, DEM-based Sea Level Rise (+1 m, +2 m, +5 m and +10 m) scenarios, and OPW flood records. The AHP results indicate that 1.36% and 33.13% of the study area falls within ‘very high’ (16.92 km²) and ‘high’ (410 Km²) flood susceptibility zones. A spatially separated evaluation of the ML models, showed highest predictive accuracy for XGBoost (73%) closely followed by SVM (68%), Random Forest (64%) and Logistic Regression (63%). Discrimination and calibration were moderate (ROC-AUC≈0.70, PR-AUC≈0.70, PR-AUC≈0.75, Brier ≈0.206) for the best-performing XGBoost Model. The socio-demographic data overlaid on the final susceptibility classes reveals that a total population of 11,318 people (with approximately 24% elderly and 19% children age groups) were vulnerable to high flood susceptibility. The study findings demonstrate the need for integrating complementary modelling approaches to enhance flood susceptibility assessments in mid-latitude Irish coastal and remote regions.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70253","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148942397","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Vernacular Flood Resilience: Constructive and Spatial Responses to Hydro-Climatic Hazards in Seven Water-Based Settlements","authors":"Tania Feldzer,&nbsp;Salvator-John Liotta,&nbsp;Emmanuel Dufrasnes","doi":"10.1111/jfr3.70250","DOIUrl":"https://doi.org/10.1111/jfr3.70250","url":null,"abstract":"<p>Flood resilience research has predominantly focused on technical protection, governance frameworks, and probabilistic modeling, leaving largely unexamined the spatial, constructive, and ecological responses developed by communities that have inhabited aquatic environments over centuries. This paper examines seven vernacular water-based settlements across five hydro-climatic contexts: Kampong Phluk (Cambodia), Buenavista Sitionuevo (Colombia), Uros (Peru), Al Tahla/Ma'dan (Iraq), Bajau Laut (Philippines/Malaysia), Cai Beo (Vietnam), and Nzulezo (Ghana), through a multiple-case comparative design founded on replication logic. Data sources combine peer-reviewed literature, satellite imagery, analytical architectural drawings, field observation, and primary interviews conducted with inhabitants and academics in direct relation with each territory. Analysis applies a homogeneous framework across 14 IPCC AR6-derived hazards and five exposure categories, with urban, architectural, and nature-based responses comparatively assessed across the seven cases. Through systematic architectural decomposition of each case, from territorial organization to constructive detail and maintenance cycle, six recurring patterns are identified: streetless spatial organization, unfixed cadastral organization and territorial mobility, hyperstatic and repairable stilt structures, kit-house logic, material cascade and resource renewal, and voluntary fallow or end-of-life abandonment of occupied substrates. These patterns operate simultaneously across material, morphological, and territorial scales and are present in combination in the most resilient cases. Water drought, salinization, and erosion remain unaddressed across the majority of the cases observed, constituting the primary directions for contemporary architectural intervention. The study suggests that the transferability of vernacular flood resilience patterns relies on architecture, continuous maintenance, cyclical material renewal, and collective resource governance as interdependent co-conditions for long-term coexistence with water.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70250","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148848969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An Integrated Framework for Flood Risk Management: Addressing Local Challenges and Resilience in Rize, Türkiye","authors":"Ayse Yildiz,&nbsp;Gokhan Sirmali,&nbsp;Rajib Shaw,&nbsp;Lee Bosher","doi":"10.1111/jfr3.70252","DOIUrl":"https://doi.org/10.1111/jfr3.70252","url":null,"abstract":"<p>In recent years, increasing flood events, driven by human activities and climate change, have caused significant loss of life and property around the globe. Consequently, effective flood risk management strategies have become crucial in response to rising flood risks. This study examines how flood risk management (FRM) is implemented in Rize, Türkiye, with particular attention to the relationship between governance and decision-making processes and community-based FRM practices. By applying a key informant interviewing qualitative research design, 51 senior civil servants, appointees, experts, and managers involved in flood risk management in Rize, Türkiye, were interviewed. The findings identify a persistent implementation gap between formal governance arrangements and community-based FRM, particularly in relation to policy and enforcement, leadership and institutional coordination, practical education, community action planning, and locally feasible preparedness. The results highlight the necessity of a comprehensive, well-coordinated strategy that connects institutional management systems and community-driven initiatives. Based on these findings, the study develops an integrated, implementation-oriented framework that synthesises the relationships between governance and decision-making and community-based FRM. The contribution of the framework lies in specifying how weaknesses in the linkages between key components can undermine FRM implementation in practice. While the framework is derived from a single-region case and requires further validation, it offers a diagnostic structure that may be adapted in comparable flood-prone settings.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70252","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148849081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Flood Insurance in Dutch Unembanked Areas: Quantifying the Interplay Between Climate Change, Urban Development, and Adaptation Strategies","authors":"Cees Oerlemans,&nbsp;Bas Kolen,&nbsp;Martine van den Boomen,&nbsp;Matthijs Kok","doi":"10.1111/jfr3.70231","DOIUrl":"https://doi.org/10.1111/jfr3.70231","url":null,"abstract":"<p>Many coastal and riverine communities lack dedicated flood protection and face growing flood risk from climate change and urban development. In such areas, flood insurance is one of few mechanisms to manage financial exposure, yet how adaptation measures, participation structures, and infrastructure policy jointly shape long-term insurability is poorly understood. We develop an integrated framework that combines hydrodynamic modelling, damage assessments, and premium estimates to quantify flood risk and insurance viability under coupled climate and socioeconomic change. Applied to unembanked areas in the Rotterdam Metropolitan Region, the Netherlands, we evaluate sea-level rise scenarios (0–2 m), urban development projections and adaptation strategies. Currently, flood insurance is unavailable for unembanked residents in the Netherlands. We therefore investigate whether such a scheme could be viable, what premiums it would imply, and how adaptation would shape its long-term sustainability. Under full participation, risk-based annual premiums covering expected damages and cost of capital would average approximately €50 per household today, but without adaptation could rise to nearly €3000 under high sea-level rise. Beyond 1.0 m sea level rise, the dominant risk driver is no longer gradual climate change itself but its second-order effect on infrastructure policy: storm surge barrier closure thresholds that must rise with the sea. We find that building-level adaptation reduces aggregate damages by up to 60%, but under risk-based participation structures increases individual premiums by shrinking the pool faster than it reduces losses. Our results indicate that flood insurance for unembanked areas could be viable under current conditions, though sustaining manageable premiums over time depends on combining building-level and regional adaptation with risk transfer mechanisms before climate impacts accelerate. The framework is transferable to other coastal and riverine settings facing similar challenges, though flood risk and premium trajectories will vary with local hydrodynamic conditions and how flood damages are allocated among insurers, governments, and households.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70231","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148753449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Rainfall Frequency Analysis to Estimate Probable Maximum Precipitation Using Stochastic Storm Transposition With Radar–Rain Gauge Data in Japan","authors":"Wasitha Dilshan,&nbsp;Yusuke Hiraga,&nbsp;Daniel B. Wright","doi":"10.1111/jfr3.70251","DOIUrl":"https://doi.org/10.1111/jfr3.70251","url":null,"abstract":"<p>Reliable design precipitation estimation is vital for flood risk management and infrastructure design. In Japan, the Depth-Area-Duration (DAD) method is widely used for operational Probable Maximum Precipitation (PMP) estimation, but it lacks a probabilistic link to rainfall frequency and does not yield the spatiotemporal rainfall fields needed for hydrological modeling. This study applies Stochastic Storm Transposition (SST), the first application of SST in Japan, to 35 years of high-resolution Radar/Rain gauge-Analyzed Precipitation (RA) data to generate probabilistic design precipitation estimates for the Arakawa and Akagawa watersheds. SST-based precipitation depths are consistently lower than DAD-based PMP estimates and operational design values adopted by the Ministry of Land, Infrastructure, Transport and Tourism (MLIT), used operationally as conservative, roughly 1000-year benchmarks, while aligning more closely with recent dynamical model-based PMP studies. At an annual exceedance probability (AEP) of 10⁻⁵, the SST-based 48-h estimate for Arakawa is 9% lower than the DAD-based value and 16% lower than the MLIT design value; for Akagawa, the SST-based 12-h estimate is 20% and 23% lower, respectively. SST-based results also align closely with recent dynamical model-based estimates, differing by 3% for Akagawa at AEP = 10⁻⁵ and showing consistent behavior for Arakawa at AEP = 2 × 10⁻³.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70251","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148752426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Direct Policy Search for a Risk-Based Levee Design Framework With Scenario Reduction","authors":"Jingya Wang,&nbsp;David R. Johnson","doi":"10.1111/jfr3.70248","DOIUrl":"https://doi.org/10.1111/jfr3.70248","url":null,"abstract":"<p>Risk-based management of coastal flood risk using realistic, process-based models poses computational challenges, particularly when addressing uncertainties like sea-level rise. This study explores the use of direct policy search (DPS) as a state-dependent, adaptive decision-making approach for upgrading levee systems. DPS is applied to the Larose to Golden Meadow Hurricane Protection Project (LGM), a single-polder levee system in Louisiana. By incorporating state variables that reflect changing system conditions, DPS identifies adaptive signposts for levee upgrades, enabling periodic adjustments to address uncertainties. This allows for adaptive changes in upgrade frequency and the identification of optimal timing for system management. To handle the computational burden of evaluating multiple future scenarios, scenario reduction techniques are developed. DPS-generated policies are evaluated against a static intertemporal optimization benchmark, comparing efficient tradeoffs between the present values of system costs and expected losses and the time series of levee height adjustments. The analysis highlights how DPS-based adaptive strategies outperform static designs by improving hypervolumes by over 11%, regardless of upgrade frequency, while reducing average marginal flood damage by 22%. When tuning the DPS formulation by varying the frequency of allowed levee system upgrades, we find that increasing the frequency of upgrades does not always lead to additional benefits. Furthermore, most levee reaches show a clear preference for early upgrades over deferred action.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70248","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148617103","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Volunteer-Led Natural Flood Management (NFM) in Calderdale, West Yorkshire: Community Perceptions, Perceived Effectiveness and Social Co-Benefits","authors":"Sophie Tankard,&nbsp;Megan Klaar,&nbsp;Richard Grayson","doi":"10.1111/jfr3.70245","DOIUrl":"https://doi.org/10.1111/jfr3.70245","url":null,"abstract":"<p>Climate change is projected to increase flood risk in the United Kingdom, with more frequent heavy rainfall affecting communities. Natural Flood Management (NFM) seeks to reduce flood peaks by restoring natural processes and is increasingly implemented by community and volunteer groups. Evidence on community engagement with volunteer-led NFM, and its wider social co-benefits, remains limited. We conducted an exploratory cross-sectional survey of residents and people working in Calderdale, West Yorkshire (<i>n</i> = 51) to examine views on effectiveness and the personal and community co-benefits of NFM interventions. This purposive, self-selected sample provides insight into individuals engaged with or aware of flood management activities, rather than representing the wider Calderdale population. Participants indicated high overall acceptance of NFM and identified environmental, personal and community benefits, despite ongoing uncertainty about the extent of local flood risk reduction. Volunteers also described improvements in mental health and a stronger connection to place. Consistent with the exploratory design and self-selected sample, flood risk reduction is interpreted independently of hydrological evidence and no claims of wider population generalisation are made. The discussion therefore emphasises implications for community-led NFM practice, including communication of co-benefits, sustaining volunteer engagement and conditions shaping transferability to other flood-risk contexts. Our findings suggest that, among respondents, volunteer-led NFM may contribute to community resilience and social learning, even where measurable hydrological reductions in flood risk are not evident.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70245","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148617105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Machine Learning Framework for Assessing Systemic Societal Vulnerability to Coastal Flooding","authors":"Adesola Akindejoye,&nbsp;Christophe Viavattene,&nbsp;Sally Priest,&nbsp;David Windridge","doi":"10.1111/jfr3.70249","DOIUrl":"https://doi.org/10.1111/jfr3.70249","url":null,"abstract":"<p>Vulnerability indices are increasingly used to understand why coastal urban communities experience differential flood impacts. However, they are criticised for subjective indicator selection, weighting and aggregation, which rely on researcher judgement and rarely integrate uncertainty. This study develops a machine-learning framework for comparing and ranking household-level vulnerability to coastal flooding in the Lekki Peninsula, Nigeria. The research employed a multi-source dataset, including household surveys, key informant interviews, field observations and secondary data. The expectation-maximisation (EM) algorithm identifies latent vulnerability clusters, while support vector machines (SVMs) model nonlinear relationships to predict household membership. A probability-weighted index and confidence threshold are integrated to quantify classification reliability. The EM-SVM model achieved 98% accuracy on the test dataset, with near-perfect separability, as indicated by an area under the curve of 0.99. Spatial analysis showed that 7.4% of households were highly vulnerable, 68.9% moderately vulnerable and 23.5% had low vulnerability. Five vulnerability profiles emerged; two exhibited relatively high education and, in one case, higher income, but remained moderately vulnerable due to dependence on flood-exposed healthcare and water infrastructure and housing types susceptible to flooding. One profile showed low vulnerability due to lower exposure of healthcare infrastructure and higher self-perceived coping capacity, indicating the role of critical infrastructure reliability and psychological preparedness. The findings reinforce the understanding that vulnerability is systemic and intersectional rather than determined by independent socio-economic attributes. Through this framework, the vulnerability approach reduces bias and offers a transparent, data-driven alternative to conventional approaches, thereby reinforcing the evidence base for equitable flood-risk planning and more targeted adaptation interventions.</p>","PeriodicalId":49294,"journal":{"name":"Journal of Flood Risk Management","volume":"19 3","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/jfr3.70249","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148616290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}