Journal of Flood Risk Management最新文献

This study examines the impact of spatial resolution on urban pluvial flood modeling, emphasizing the role of high-resolution topographic data in flood inundation mapping. Using the physically-based H12 urban flood model, we simulated the December 6–8, 2015, pluvial flood event in a sub-watershed in Portland, Oregon. We compared the temporal evolution of inundation maps from a benchmark 1 m resolution model with coarser resolutions (2–50 m) and assessed accuracy using water depth error measures and grid-based inundation extent metrics. Our results indicate that accumulated inundated water volume increases significantly with coarser grid resolutions, leading to larger discrepancies in flood extent. Coarser grids generally overpredict inundation extents, except in the 2 m model. Accuracy metrics decline with resolution coarsening, with the hit rate (H) dropping below 0.7 and the critical success index (C) falling to 0.5 or lower beyond 7 m resolution. Multi-directional flow path analyses reveal that inundation extent expands with coarser resolutions, while computational efficiency improves. The primary source of accuracy degradation is the inability of coarse grids to capture key urban topographical details, such as road networks, which influence floodwater movement. While no single optimal resolution applies universally, grid resolutions must be fine enough to accurately represent major urban features critical to flood dynamics.

2D/1D dual drainage models are one of the most useful tools for studying urban pluvial flooding. However, the accuracy of these models depends on data quality and completeness. This study assesses the effects of sewer network data completeness on the results of the 2D/1D free distribution model Iber-SWMM. The research is conducted in two case studies: Differdange (Luxembourg) and Osuna (Spain), considering six different return period storms. Different scenarios of data completeness were generated by simplifying the original sewer network, based on two characteristics of the conduit segments: the Strahler Order number and the length. Each scenario was evaluated by comparing the maximum flood extent maps obtained. The results indicate that the lower the degree of data completeness, the higher the overestimation of the maximum flood extent. For 80% completeness, the False Alarm Ratio is less than 0.05, but it can increase exponentially to over 0.30 when network completeness drops to 20%. However, if the available information includes the most important conduits, such as the main collectors, errors are minimal. Furthermore, if the data on surface elements (inlets) is also complete, the accuracy of flood modeling is maintained compared to the complete data scenario. These results can contribute to the simplification of flood model setup in large urban areas, where not always complete sewer network data sets are available and information preprocessing can be complex and time-consuming, and the computation of the network in SWMM can become a bottleneck in the simulation.

Extreme heavy rainfall in Western Europe on 13–15 July 2021 caused severe flooding, notably in Germany's Rhineland-Palatinate and North Rhine-Westphalia. This study examines Flood Forecasting, Early Warning, and Disaster Response weaknesses during this event, focusing on the city of Stolberg. An interdisciplinary mixed-methods approach integrated meteorological, hydrological, and social science research. Data included river gages, precipitation measurements, warnings, and 300 documents, with 30 expert interviews. Weaknesses included imprecise meteorological forecasts due to dynamic weather, leading to general warnings without specific impact guidance. Limited flood forecasting hindered local preparation and response, exacerbated by an emergency response system unprepared for the event's scale. The top-down approach of Flood Forecasting and Early Warning conflicted with the bottom-up processes of Disaster Response, hampering effective crisis management. The study reveals critical weaknesses and calls for improved forecasting, integrated response plans, communication protocols, and crisis channels to enhance flood resilience. Future research should explore these issues in other extreme flood events and compare international Flood Forecasting, Early Warning, and Disaster Response systems.

Bridge owners and regulatory agencies have a duty to assess risks derived from hydraulic actions including scour, uplift, drag, debris impact, deck displacement, and other consequences that can lead to a loss in the load carrying capacity of a bridge. In the UK, the CS469 (Management of scour and other hydraulic actions at highway structures) is the standard for the assessment of hydraulic actions to highway bridges. The methodology in CS469 for the calculation of the hydraulic characteristics of the flow at critical cross-sections within the channel and the bridge crossing, although simplistic by design to minimize computational effort, is intrinsically inaccurate since it makes use of unrealistic (i.e., non-physically based) approximations. This results in estimations of risk and vulnerability levels that could include high levels of uncertainty. In this paper, we propose to bypass these approximated hydraulic calculations by harnessing the computational power of 2D hydraulic models, which would not require any additional field data collection than needed for the original CS469 method. We recommend a fully 2D HEC-RAS model with the inclusion of bridges as 1D elements within the flow areas and only requiring publicly available data or data obtained from existing assessments in order to future-proof the approaches and adhere to an open-source/open-access philosophy, but also imposing only a marginal increase in cost for bridge management teams. Results from the two models—2D HEC-RAS and the existing approach in CS469, are compared for a number of real-world bridges. The comparisons show that the estimations by HEC-RAS are substantially higher for water depth (up to 138%) and lower for flow velocity (down by 58%). When these values are applied to the estimation of hydraulic vulnerability and scour risk, the differences are significant. Scour depths with the use of HEC-RAS models are typically much lower (up to 3.9 m, and on average 1.7 m) than with simplified hydraulic equations, and this translates into lower (yet, more appropriate) scour risk levels. Hydraulic vulnerability to submergence of the assessed bridges is also assessed very differently, typically higher by the 2D model method. Overall, the results show that 2D numerical hydraulic simulations present a much more accurate estimation than existing methods, better balancing risks deriving from scour and hydrodynamic actions and with comparable effort and data requirements. The model displays consistency across an exhaustive set of simulations for a range of variables and bridges, showing limited variability and proneness to errors, whilst values estimated by CS469 are in most cases significantly different. Future versions of CS469 and similar documents should prioritize this methodology to provide a more accurate and realistic risk estimation.

Flood protection assets such as dams are increasingly seen as part of a larger system, but the complexity of dam management, unclear communication, and operational misunderstandings in operational protocols can lead to unnecessary downstream flooding. This paper investigates how human factors, such as dam operators' communication and roles and responsibilities, influence flood flows. Using HEC-RAS, the study varies initial reservoir volume and pre-release duration in four New Zealand catchments, with potential for adaptation in other areas. The results found: (i) dams designed to provide flood storage had stronger correlations between the duration of pre-releases and outflow reductions, (ii) dams with large storage capacities and fewer release mechanisms had stronger correlations between the initial reservoir volume and outflow reductions, (iii) a dam's ability to appropriately control flow is governed by the presence and implementation of clear operating procedures shown by the dam mandated to provide flood storage having the highest consistency in flow reduction and a 6% difference in maximum dam outflow between best and worst-case operations, and (iv) mismanagement of outflows can increase downstream flooding; in one catchment, the outflow was 38% above the inflow. The results are widely applicable given the increasing importance of flood control mechanisms and operational protocols.

Floods are among the most destructive natural disasters, presenting significant challenges due to their unpredictability and complex behavior. This study develops a robust flood prediction framework for the Chakdara monitoring station on the Swat River, Pakistan, by integrating traditional statistical methods with advanced machine learning (ML) models. Four statistical distributions—Log-Normal, Gumbel, General Extreme Value (GEV), and Log-Pearson Type III (LP-III)—were evaluated for flood frequency analysis. Among these, the LP-III distribution demonstrated the best performance with an R² value of 0.78. To enhance prediction accuracy, two ML models—Artificial Neural Network (ANN) and multilayer neural network (MLNN)—were employed. The MLNN model outperformed all others, achieving R² values of 0.96 for training and 0.93 for testing, confirming its high reliability for streamflow prediction. Furthermore, the trained MLNN was adapted to future climate conditions using downscaled and bias-corrected CMIP6 projections under SSP245 and SSP585 scenarios. This allowed for reliable discharge forecasting under changing precipitation and temperature trends. The proposed hybrid approach not only improves the accuracy of flood predictions but also supports long-term planning for flood risk mitigation. These findings provide essential insights for policymakers, engineers, and disaster management agencies to design adaptive infrastructure and implement proactive flood management strategies in the Swat River basin.

The risk posed globally by pluvial flooding to people and properties is growing due to urbanisation, infrastructure development and intensification of rainfall due to climate change. Whilst tools to model pluvial flood hazard have also advanced, there remains a knowledge gap around whether design storms used in modelling adequately represent the temporal distribution of rainfall within the extreme convective storms which drive flooding. In the UK, the industry standard design storm considers rainfall events to always have a singular, central intensity peak. Study of UK extreme rainfall observations suggests that loading of rainfall towards the start or end of events is in fact more common. This study highlights the sensitivity of pluvial flood extent, hazard and timing to the shape of the design rainfall profile for two urban catchments in northern England. We demonstrate that for events with the same accumulated rainfall depth, there is up to a 25% increase in total flood-affected area with a back-loaded compared to a front-loaded profile. Failing to account for the variability in event profile shapes observed in real events may result in substantial inaccuracies in the design of flood risk management solutions, leading to both underestimation and overestimation of the required measures.

Rainfall-driven urban surface water flooding is one of the most common natural disasters that lead to traffic disruption, economic loss, and even casualties. Assessing its hazards is critical not only for flood management but also for urban and territorial planning. Physics-based models can simulate hydrological and hydraulic processes to predict floods; however, they are computationally expensive for large-scale and high-resolution simulations. This study presents a U-Net-based deep learning method for assessing the hazard levels of urban surface water flooding. The approach adopts three methods to improve the baseline U-net model: (1) Squeeze-and-Excitation Blocks that enhance feature representation; (2) Focal Loss, a loss function that mitigates the influence of data imbalance; and (3) Random Cutout, a data augmentation method that prevents overfitting. Catchment data are used as input to train the deep learning model against flood hazard targets under three different levels of annual exceedance events. The results showed that the models are capable of identifying mid and high hazards. The proposed three methods mutually constrained each other and can reduce the influence of data imbalance. The proposed model demonstrates potential for practical flood management through rapid and accurate identification of high-risk areas.

Hubot, K., H. Goitom, G. Aregay, and T. Yisfa. 2025. “Prediction of Future Land Use and Land Cover Change Impact on Peak Flood: In Case of Babur Watershed, Tekeze Basin, Ethiopia.” Journal of Flood Risk Management 18, no. 2: e70077. https://doi.org/10.1111/jfr3.70077.

In the published article, there was an error in the original title: “Tekeze Basin” should have been removed and replaced with “northern”.

The correct title is: “Prediction of future land use and land cover change impact on peak flood: in case of Babur watershed, northern Ethiopia”.

The descriptions for Tekeze basin found on page 3, lines 1-8, line 24, page 4, line 3, page 16, line 30, and references belongs to this description should be removed from this article.

In the published version of this article, Figure 1 (Ethiopian River basin map) contained generalized boundaries that could lead to misinterpretation. To improve accuracy, this figure has been replaced with an administrative map of Ethiopia, which provides clearer and more precise spatial context for the study area (Babur watershed, northern Ethiopia).

We apologize for this error.

In this study, flood mitigation measures, both structural and nature-based ones, are proposed, and their social and economic performances are quantified using a multi-criteria decision analysis framework. The effectiveness of the selected measures is evaluated through numerical modeling. As a case study, flood inundation areas and flood hazard maps are determined in the Ağva District of Istanbul using high-resolution LIDAR data in a 2D hydrodynamic model for different flood scenarios. The numerical model is calibrated against measured discharges at the river gauging station in the river basin. For the social and economic assessment, a total of seven criteria are assessed: number of affected inhabitants, number of affected cultural items, number of public institutions under flood, expected annual damage, investment cost of measures, annual maintenance cost, and benefit–cost ratio. The overall score of each flood mitigation measure alternative is computed, and their performance is compared for different strategies. For our case study, the implementation of a levee exhibits the highest economic and social performance for flood risk management.