Journal of Flood Risk Management最新文献

Flood risk communication is a core component of flood risk management, yet persistent challenges limit its effectiveness in supporting public understanding, preparedness and adaptive behaviour. Here, we examine flood risk maps as communication tools at the interface of scientific modelling, visual design, and human risk perception. Using a narrative and scoping approach, we synthesise conceptual, theoretical and empirical literature, with particular attention to risk perception theory, framing and map design. We show that flood risk maps often fail to communicate effectively due to poor readability, technical language, inconsistent colour conventions and cognitive biases that shape interpretation and response. Although probabilistic flood maps have been developed to better represent spatial variability and uncertainty in flood risk, they are frequently misunderstood without appropriate framing and contextual support. Our review highlights the critical role of framing choices, communication channels and trust in information sources in shaping how flood risk information is interpreted and acted upon. We further show that participatory mapping can enhance local relevance, understanding and trust by incorporating lived experience, but its application is constrained by issues of scalability, institutional capacity and potential bias. We, therefore, argue that flood risk maps are most effective when embedded within broader, multi-channel communication strategies.

This article develops a two-component mixture model combining the weighted Inverse Rayleigh (WIR) distribution and Gumbel Type-II distribution for the estimation and prediction of flood events. The study utilizes 29 years (1990–2018) of flood data from the Federal Flood Commission (FFC) of Pakistan for the Jhelum River, using two gauging stations (Mangla and Rasul) across two catchments (U/S and D/S). Two distinct approaches, Annual Maximum series (AMS) and Peak over threshold (POT), are used for the estimation of parameters of the mixture models in a Bayesian context. Bayesian analysis is performed using the Square Error Loss Function (SELF) and Quadratic Loss Function (QLF) with gamma and beta priors. Bayes estimators and their posterior risks for both the Weighted Inverse Rayleigh and Gumbel Type-II distributions are derived. For the Gumbel type-II distribution, both the shape and scale parameters are treated as random. A comprehensive simulation study is conducted to examine the behavior of derived Bayes estimators and their posterior risks. The study also compares various loss functions and aims to explore a well-fitted distribution. Additionally, it aims to determine return periods for accurate flood event predictions.

Flash flooding is amongst the most severe natural hazards, causing widespread socioeconomic impacts across both wet regions and drylands. In ungauged mountainous basins, effective risk warning based on hydrodynamic modelling is challenging due to sparse hydrological observations and complex terrain. Rainfall forecasts can enable timely alerts despite the computational demands of modelling, but their inherent uncertainties further complicate predictions. This study explores the potential of a Flash Flood Risk Index (FFRI), which integrates hydrodynamic simulation outputs with socio-economic exposure and vulnerability indicators, to provide actionable early risk signal under data-scarce conditions. The 2022 Datong flash flood in China is used as a case study. Grid-based hydrodynamic simulations were conducted across varying key parameters and rainfall scenarios to discuss model uncertainty. Model performance was evaluated using UAV-derived inundation extents, achieving high F1 scores (0.88–0.90), indicating reliable reproduction of flood extents. Simulated water depths and river discharges, however, exhibited substantial discrepancies, particularly in downstream convergence zones, which highlights the critical influence of parameter and rainfall uncertainty on hydrodynamic outputs. The FFRI proposed in the study mitigated these uncertainties, consistently identifying high-risk areas, especially at the administrative (village) scale across all scenarios. These findings demonstrate that, in data-limited basins, integrating hydrodynamic modelling with socio-economic indicators is a practical way to provide actionable risk signal, supporting early-warning and emergency response where traditional calibration and detailed observations are unavailable.

Physics-based flood hydrodynamic models are widely used for predicting inundation in urban basins with complex building layouts. While the treatment of urban buildings in these models has been extensively discussed, over-assumptions can introduce inaccuracies, uncertainties, and excessive computational effort, particularly under data-scarce conditions. This study proposes a simple yet effective method, the Building Coverage Ratio (BCR) scheme, to account for building effects in city-scale urban inundation modeling. The BCR scheme quantifies water abstraction to generate surface runoffs in densely built-up areas by dynamically adjusting drainage and infiltration volumes based on the proportion of building footprint in each grid cell. This approach improves the accuracy of urban flood predictions when high-resolution data is unavailable. Validated against a historical rainstorm event in Zhuhai, China, the BCR scheme demonstrated its capability to efficiently and accurately reproduce spatiotemporal inundation patterns. The method significantly improved street-level flooding simulations, which are often underestimated in traditional approaches that neglect building effects. Results show that simulation accuracy increases from 33% without treatment to over 85% when the BCR scheme was applied to 30 m-resolution Digital Elevation Model (DEM). As the method relies entirely on open-source datasets, it offers a practical and transferable solution for urban flood prediction in data-scarce regions.

China has complex topography, diverse flood mechanisms, and high population exposure, making it highly vulnerable to flooding, highlighting the need for robust national-scale hazard assessments to identify flood-prone regions. However, most existing hazard studies are limited to regional scales or rely on empirical indicator-based methods that overlook flood dynamics. While some global-scale studies use physics-based modeling, they offer little insight into China and rarely consider reservoir operations. This study advances national-scale flood hazard mapping for China using the hydrodynamic Global Flood Model, CaMa-Flood (v4.2). Simulations driven by ERA5-Reanalysis runoff showed stronger agreement with observed streamflow than ERA5-Land. Flood frequency analysis identified the nonparametric Kernel Density Estimator as the most suitable approach. The resulting 0.05° flood hazard maps reveal that nearly half of mainland China faces some level of 1-in-100-year flood hazard, with 26% in the high to very high category. Incorporating reservoir operations reduced the number of national high hazard areas by up to 31%, underscoring their vital role in mitigation. The derived hazard, population exposure, and GDP-based analysis provide a data-driven foundation for national and provincial flood risk management, offering a scalable framework for robust hazard assessment and improved exposure and flood risk evaluation.

The Sudd wetland in South Sudan extends over 90,000 km². Large-scale flood events in recent years (2019–2022) are said to have led to the displacement of an estimated 1.8 million people in total. However, these estimates are approximate and to date there has not been a systematic analysis of population exposure to flooding in the Sudd region. This study seeks to address this gap by using global flood modeling, satellite observations of flood extent, and global gridded population datasets to analyze population exposure. Recognizing the inevitable limitations of these datasets, we intersect all the available global flood mapping and population datasets. The results indicate that 0.8–2.9 million people are currently exposed to the 100-year return period flood extent, depending on the flood model and population dataset used. Aggregated results of the model agreement intercomparison indicate that all five global models agree on key flood-prone areas within and around the Sudd, which is further corroborated with satellite flood observations. Intercomparison of the population density among the four georeferenced population products demonstrates that WorldPop and GHSL-Pop population distributions better represent the patterns of the Sudd rural settlements that are typically in forms of clusters. The uncertainty in exposure estimates is attributable to variations in both flood outlines and geospatial population estimates. These findings provide hitherto unavailable insights into flood exposure in South Sudan, to inform flood management decisions and disaster reduction responses in the Sudd region. This study demonstrates the global significance of model intercomparison as best practice for any flood exposure analysis to underpin policy and decision-making in Africa and other data-scarce regions.

Flooding remains one of the most critical natural hazards threatening livelihoods, infrastructure, and ecological systems in Ethiopia's highland landscapes. This study presents a rigorously integrated, multi-criteria flood risk assessment that combines the Analytical Hierarchy Process (AHP) with GIS-based spatial modeling to delineate, classify, and prioritize flood-prone zones within the Gimba sub-watershed. Eight biophysical flood-generating factors stream density, rainfall, slope, elevation, land use/land cover (LULC), soil type, geology, and groundwater depth were systematically evaluated to derive a standardized and weighted Flood Hazard Index (FHI). Standardization was performed through min-max normalization to ensure comparability across variables, while weighting was achieved using pairwise comparison matrices in AHP, allowing expert judgments to quantify the relative influence of each factor on flood susceptibility. The resulting FHI was subsequently integrated with exposure layers, including population density, built-up intensity, and road network distribution, to compute a comprehensive Flood Risk Index (FRI). Model accuracy was assessed through spatial overlay with historical flood inventories and Receiver Operating Characteristic (ROC) curve analysis, generating an Area Under the Curve (AUC) value of 0.885, which confirms very strong predictive capability. Results reveal that approximately 34.6% of the Gimba sub-watershed falls within high or very high flood hazard classes, with pronounced hotspots in downstream floodplains, mid-slope flash-flood corridors, and zones undergoing severe land degradation. A targeted mitigation suitability analysis further shows that 32.1% of these high-risk zones are highly appropriate for nature-based solutions such as agroforestry, check-dam installation, terracing, and catchment-scale reforestation. Importantly, participatory engagement through key informant interviews (KIIs) and focus group discussions (FGDs) played a direct role in refining hazard classifications and prioritizing intervention sites, ensuring that the spatial outputs aligned with community experience, local knowledge, and on-the-ground feasibility. This co-production of knowledge enhanced both the accuracy and social legitimacy of the proposed measures. Overall, the study provides a transparent, adaptable, and policy-relevant framework for evidence-based flood risk management. By integrating biophysical hazard metrics, terrain suitability, and stakeholder insights, the approach supports Ethiopia's Climate Resilient Green Economy (CRGE) strategy and national Disaster Risk Management (DRM) initiatives. Moreover, the methodology demonstrates strong scalability and transferability, offering a robust decision-support tool that can be applied across other data-scarce, hazard-prone watersheds to strengthen climate-resilient landscape planning.

Urban flood disasters demand dynamic assessment of population risk, yet most evacuation models fail to capture the multifaceted, time-varying nature of such events. To address this gap, this study develops the modular urban flood-evacuation (MUFE) framework, an integrated approach to dynamic urban flood risk assessment. MUFE explicitly models pedestrian and vehicle behaviors under inundation and their interactions with evolving hydrodynamic conditions. The framework is tailored for island cities facing compounded threats from storm surges and sea-level rise, and is demonstrated on Haidian Island, Haikou, China. The research couples multi-scenario rainfall-tide hydrodynamic simulations (via InfoWorks ICM) with an agent-based model implemented in the open-source Mesa framework. Mesa enables the explicit representation of pedestrian and vehicle agents operating under perceive-decide-act cycles and local information constraints. This integration enables joint analysis of hydrodynamic hazard fields—shaped by dynamic tidal fluctuations—and the behavioral responses of pedestrians and vehicles. The framework is modular, allowing alternative hydrodynamic or data-driven flood models to be coupled for transparent comparison. Simulation results show that pedestrian hazard exposure varies markedly across space and time, shaped by adaptive evacuation behavior, shelter availability, and tide dynamics. Small vehicles are more susceptible to instability—even under shallow depths—while larger vehicles are more resilient but not immune under severe conditions. These patterns underscore the need for differentiated evacuation planning and targeted infrastructure resilience measures in island cities.

In light of the increasing adoption of sustainable urban drainage measures for flood control, driven by advances in the urban drainage field and, particularly, in downtown São Paulo, by the significant challenge associated with implementing large detention reservoirs due to the scarcity of open spaces within the watershed, this study investigates the hydrologic and hydraulic parameters that influence the rainfall–runoff simulation in urban areas using the PCSWMM model. Real-time telemetry data, including measurements from meteorological radar, rainfall gauges, and fluviometric stations, were employed for model calibration and validation. Historical images of São Paulo from the 1930s were used to assess the potential impact of promoting permeable surfaces within the basin on current design hydrographs and flood-prone area extents. Two future scenarios incorporating sustainable urban drainage strategies for flood mitigation were analyzed to evaluate their effectiveness, complemented by a sensitivity analysis to identify the parameters with the greatest influence on the simulated hydrographs. Results indicate that, even though sustainable measures significantly reduce flood-prone area extents, their performance is highly dependent on rainfall duration and return period. Consequently, an integrated approach combining conventional and sustainable strategies is recommended for managing design storms with longer durations and higher return periods.

In recent years, extreme rainfall events led to severe flooding in several European countries which caused extraordinary human and economic losses. Such events, which are projected to become more likely because of climate change, expose many citizens to extremely stressful situations that involve intense fear, shock, and loss. Previous studies clearly point toward a direct relation between such flood experiences and negative mental health outcomes of those affected. However, existing studies commonly focus on directly exposed populations only, preventing a direct comparison with a control group. This makes it more difficult to separate the effect of the flood event from other factors that potentially affect the mental health of the respondents. Here, we use survey data from 698 residents from the Marche region in Italy, which was affected by disastrous flooding in 2022. The survey focused not only on the directly affected population (n = 392) but also included a nonaffected control group (n = 306). We use the short version of the Kessler Distress Scale (K6) as a screener for severe mental distress. Results show that directly affected respondents exhibit a 13.1%–16.7% higher prevalence rate of indications of severe mental distress compared to the control group. The significant impact of the flood event on negative mental health outcomes is further confirmed by regression analyses, which show a direct influence of several flood stressors on severe mental distress, including physical health impacts and higher water levels on one's own building. Since mental illness is associated with high burdens for those affected and their families, as well as high socioeconomic costs, this aspect deserves more attention in postdisaster contexts. The results presented in this article can be used as a reference by responsible authorities for estimating the additional demand for psychological assistance needed in the aftermath of such severe events.