Pluvial Flooding Events Research Articles

A Coupled Human and Natural Systems (CHANS) framework integrated with reinforcement learning for urban flood mitigation

To address the challenge of escalating urban flood risk and the deficiency in effective flood emergency management, this study introduces a novel Coupled Human and Natural Systems (CHANS) modelling framework that employs hierarchical reinforcement learning to optimise mobile pump scheduling and placement for urban flood risk mitigation. The CHANS framework integrates hydrodynamic and agent-based models within a multi-GPU computing environment for high-resolution, real-time flood inundation modelling and risk assessment to enrich Reinforcement Learning (RL) training. In the application to Ninh Kieu District in Can Tho City, Vietnam, the new RL-enabled modelling framework is used to evaluate optimal mobile pumping strategies for concurrent pluvial flooding and post-flooding events against the traditional deployment approaches. Results demonstrate that RL-based strategies can significantly enhance flood risk reduction, outperforming traditional methods by achieving 2× and 4× improvements in the concurrent and post-flooding periods/events, respectively. Incorporating human factors and adapting to local conditions, the RL agent provides valuable insights into mobile pump scheduling and deployment strategies. Sensitivity analysis confirms the robustness of the CHANS modelling framework and underscores the role of RL in optimising mobile pump scheduling and placement where traditional rule-based strategies are challenging.

Assessment of pluvial flood events based on monitoring and modeling of an old urban storm drainage in the city center of Yangon, Myanmar

Pluvial flooding is a critical issue in cities worldwide, particularly in lowland areas with old and deteriorating drainage systems. The primary driver of pluvial flooding is extreme rainfall; other drivers include urbanization, inadequate drainage systems, improper solid-waste management, and the tidal backwater effect. However, the interplay between these drivers makes predicting pluvial floods difficult and complex. Previous studies in developing countries seldom used water-level data or simulation modeling to identify the causes of pluvial flooding. In this study, rainfall data and water-level variations in an open channel drain and a receiving river controlled by sluice gates were collected and evaluated in detail to investigate pluvial flooding events. To predict these events, we generated a hydrodynamic model using InfoWorks ICM and verified its results using water logger data and official field reports. Analysis shows that drainage-system failures due to solid blockage and receiving water-level variation contribute more to pluvial flood occurrence than heavy rainfall. Lastly, we discuss measures to mitigate pluvial flooding in Yangon, Myanmar. The proposed monitoring and modeling approach can suitably predict pluvial flooding occurrence and provide useful quantitative data for flood risk management.

Real time probabilistic inundation forecasts using a LSTM neural network

Probabilistic inundation forecasts that consider the uncertainty in rainfall predictions are crucial for flood early warning systems to effectively manage and reduce potential risks posed by pluvial flood events. Timely generation of such forecasts is challenging with physically-based numerical models due to computational demands. In contrast, data-driven models have a relatively low computational cost and can generate results quickly, making them a promising alternative to overcome this issue. This study proposes a long short-term memory (LSTM) neural network that can predict inundation progression over time at a high spatial resolution The network is trained on 1600 hydraulic simulations conducted using a 1D2D hydraulic model. With the trained network, probabilistic inundation forecasts are generated by combining the deterministic inundation predictions of 50 ensemble members of the rainfall forecast. The model is successfully tested for temporally varying rainfall events in a rural study area, and can generate accurate probabilistic inundation forecasts within seconds.

Pluvial flood risk assessment for 2021–2050 under climate change scenarios in the Metropolitan City of Venice

Pluvial flood is a natural hazard occurring from extreme rainfall events that affect millions of people around the world, causing damages to their properties and lives. The magnitude of projected climate risks indicates the urgency of putting in place actions to increase climate resilience. Through this study, we develop a Machine Learning (ML) model to predict pluvial flood risk under Representative Concentration Pathways (RCP) 4.5 and 8.5 for future scenarios of precipitation for the period 2021–2050, considering different triggering factors and precipitation patterns. The analysis is focused on the case study area of the Metropolitan City of Venice (MCV) and considers 212 historical pluvial flood events occurred in the timeframe 1995–2020. The methodology developed implements spatio-temporal constraints in the ML model to improve pluvial flood risk prediction under future scenarios of climate change. Accordingly, a cross-validation approach was applied to frame a model able to predict pluvial flood at any time and space. This was complemented with historical pluvial flood data and the selection of nine triggering factors representative of territorial features that contribute to pluvial flood events. Logistic Regression was the most reliable model, with the highest AUC score, providing robust result both in the validation and test set. Maximum cumulative rainfall of 14 days was the most important feature contributing to pluvial flood occurrence. The final output is represented by a suite of risk maps of the flood-prone areas in the MCV for each quarter of the year for the period 1995–2020 based on historical data, and risk maps for each quarter of the period 2021–2050 under RCP4.5 and 8.5 of future precipitation scenarios. Overall, the results underline a consistent increase in extreme events (i.e., very high and extremely high risk of pluvial flooding) under the more catastrophic scenario RCP8.5 for future decades compared to the baseline.

The utility of using Volunteered Geographic Information (VGI) for evaluating pluvial flood models

Pluvial floods are increasingly threatening urban environments worldwide due to human-induced climate change. High-resolution, state-of-the-art pluvial flood models are urgently needed to inform climate change adaptation and disaster risk reduction measures but are generally not empirically tested because of the rarity of local high-intensity precipitation events and the lack of monitoring capabilities. Volunteered Geographic Information (VGI) collected by professionals, non-professionals and citizens and made available on the internet can be used to monitor the dynamic extent of a pluvial flood during and after an extreme rain event but is sometimes considered to be unreliable. In this paper, we explore the general utility of VGI to evaluate the performance of pluvial flood models and gain new insights to improve these models. As background for our research, we use the capital city of Budapest, which recently suffered three heavy rainfall events in just five years (2015, 2017 and 2020). For each pluvial flood event, we collected photographic evidence from different online media sources and estimated the associated water depths at various locations in the city from the image context. These were compared with the results of a 2D pluvial flood model that has been shown to provide comparable results to other state-of-the-art inundation models and is easily transferred to other urban areas due to its reliance on open data sources. We introduce a general methodology for comparing VGI with model data by probing different spatial resolutions. Our findings highlight untapped potential and fundamental challenges in using VGI for model evaluation. It is proposed that VGI may become an essential tool and improve the confidence in model-based risk assessments for climate change adaptation and disaster risk reduction.

An Evaluation Framework for Urban Pluvial Flooding Based on Open-Access Data

Identifying the location and estimating the magnitude of urban pluvial flooding events is essential to assess their impacts, particularly in areas where data are unavailable. The present work focused on developing and exemplifying a tool to evaluate urban pluvial flooding based on open-access information. The tool has three separate submodules: (1) sewer network generation and design; (2) hydrodynamic model development; (3) urban pluvial flood evaluation. Application of the first two modules in two catchments and comparison of these results with real data indicated that the tool was able to generate systems with realistic layouts and hydraulic properties. Hydrodynamic models derived from this data were able to simulate realistic flow dynamics. The third module was evaluated for one of the study cases. The results of this indicated that the current approach could be used to identify flood areas and associated flood depths during different rainfall scenarios. The outcomes of this study could be used in a wide variety of contexts. For example, it could provide information in areas with data scarcity or uncertainty or serve as a tool for prospective planning, design, and decision making.

ABM-based emergency evacuation modelling during urban pluvial floods: A “7.20” pluvial flood event study in Zhengzhou, Henan Province

ABM-based emergency evacuation modelling during urban pluvial floods: A “7.20” pluvial flood event study in Zhengzhou, Henan Province

Modeling Temporal Accessibility of an Urban Road Network during an Extreme Pluvial Flood Event

Modeling Temporal Accessibility of an Urban Road Network during an Extreme Pluvial Flood Event

Building level flood exposure analysis using a hydrodynamic model

The advent of detailed hydrodynamic model simulations of urban flooding has not been matched by improved capabilities in flood exposure analysis which rely on validation against observed data. This work introduces a generic, building-level flood exposure analysis tool applying high resolution flood data and building geometries derived from hydrodynamic simulations performed with the 2D hydrodynamic flood modelling software CityCAT. Validation data were obtained from a survey of affected residents following a large pluvial flood event in Newcastle upon Tyne, UK. Sensitivity testing was carried out for different hydrodynamic model and exposure tool settings and between 68% and 75% of the surveyed buildings were correctly modelled as either flooded or not flooded. The tool tends to underrepresent flooding with a better performance in identifying true negatives (i.e. no flooding observed with no flooding modelled) compared to true positives. As higher true positive rates were accompanied by higher false positive rates, no single scenario could be identified as the optimal solution. However, the results suggest a greater sensitivity of the results to the classification scheme than to the buffer distance applied. Overall, if applied to high resolution flood depth maps, the method is efficient and suitable for application to large urban areas for flood risk management and insurance analysis purposes.

Spatio-temporal cross-validation to predict pluvial flood events in the Metropolitan City of Venice

Due to a combination of climate change and urbanization, the instances of pluvial flooding are expected to increase in the next decades posing raising threats to properties, people and productive assets. Predicting and mapping pluvial flood-prone areas is becoming a crucial step in flood mitigation and early warnings, as well as climate change adaptation strategies, to be incorporate in urban planning. Most commonly applied machine learning (ML) procedures for pluvial flood risk assessment, neglect to account for spatio-temporal constraints, leading to overoptimistic models that underestimate the prediction error. In this paper, we propose a novel ML-based methodology for pluvial flood risk prediction in the Metropolitan City of Venice which, introducing a features selection process and spatio-temporal cross-validation, permits to reduce overfitting of the resulting ML models. Spatio-temporal characteristics of floods are derived from a dataset of 60 historical events occurred in the area between 1995 and 2020. Logistic Regression (LR), Neural Networks (NN) and Random Forest (RF) models are applied to identify and prioritize sub-areas that are more likely to be affected by pluvial flood risk, considering the daily precipitation amount and 12 different triggering factors. The models were validated using Random Cross-Validation (R-CV) and Leave Location and Time Out cross-validation (LLTO-CV), that split data in training and validation set considering both time and space. In addition, a forward features selection procedure was applied to identify the features, among the triggering factors, that better face spatio-temporal overfitting in pluvial flood prediction based on the Area Under the Curve (AUC) score. Results suggest that Logistic Regression and LLTO-CV represent the most reliable model to predict pluvial flood events in new spatio-temporal conditions, while, among the triggering factors, distance to river and distance to road resulted the prominent ones.

Wastewater System Inflow/Infiltration and Residential Pluvial Flood Damage Mitigation in Canada

Pluvial flooding in urban areas is one of the most significant drivers of disaster loss in Canada. Damages during pluvial flood events are associated with overwhelmed urban drainage (stormwater and wastewater) systems. During the period from 2013 to 2021, Canadian property and casualty insurers reported approximately CAD 2 billion in personal property (residential) pluvial sewer backup claims during flood catastrophes. There has been growing interest in managing pluvial urban flood risk, notably through newly funded national programs focused on climate change adaptation. These programs have included the development of new guidelines and standards focused on managing the underlying factors contributing to urban and basement flooding. Inflow and infiltration (I/I) has received limited attention in the pluvial flood literature, however. Informed by significant engagement with practitioners in Canada, this paper provides a review of the issue of I/I into wastewater systems and its relation to pluvial flooding. The paper will address concerns related to private property engagement in I/I and urban pluvial flood reduction programs. Both improved technical standards and administrative support are needed to ensure that wastewater infrastructure is less susceptible to I/I over its lifecycle.

Planning Nature Based Solutions against urban pluvial flooding in heritage cities: A spatial multi criteria approach for the city of Florence (Italy)

Study regionCity of Florence, central Italy Study focusAiming at defining a nature-based strategy to mitigate pluvial flood risk, a two-tiered methodology is proposed. Firstly, the areas prone to pluvial flood are identified by a spatial multi-criteria analysis that combines five criteria (imperviousness, slope, hydrologic soil groups, density of sewer system and social vulnerability) to build a Pluvial Flood Index (PFI). The PFI is validated by the comparison with historical pluvial flood events in the city and it is used for the identification of high priority areas for intervention. Then, this information is merged with the analysis of urban planning and NBS design constraints to identify the suitable areas for NBS installation. New hydrological insights for the regionResults allow the definition of a NBS implementation strategy against pluvial flooding in the city of Florence, identifying for each pluvial flood hotspot the set of measures that can be implemented to solve the problem. The proposed approach represents a flexible assessment technique that can be reproduced in other urban context and provide useful support for NBS adoption in urban flood risk management for heritage cities.

Extreme Sea Surges, Tsunamis and Pluvial Flooding Events during the Last ~1000 Years in the Semi-Arid Wetland, Coquimbo Chile

The coast of Chile has been exposed to marine submersion events from storm surges, tsunamis and flooding due to heavy rains. We present evidence of these events using sedimentary records that cover the last 1000 years in the Pachingo wetland. Two sediment cores were analyzed for granulometry, XRF, pollen, diatoms and TOC. Three extreme events produced by marine submersion and three by pluvial flooding during El Niño episodes were identified. Geochronology was determined using a conventional dating method using 14C, 210Pbxs and 137Cs). The older marine event (E1) was heavier, identified by a coarser grain size, high content of seashells, greater amount of gravel and the presence of two rip-up clasts, which seems to fit with the tsunami of 1420 Cal AD. The other two events (E3 and E5) may correspond to the 1922 (E3) tsunami and the 1984 (E5) storm waves, corroborated with a nearshore wave simulation model for this period (SWAM). On the other hand, the three flood events (E2, E4, E6) all occurred during episodes of El Niño in 1997 (E6), 1957 (E4) and 1600 (E6), represented by layers of fine-grain sands and wood charcoal remains.

Performance of nearest neighbour metrics for pluvial flood nowcasts in urban catchments

Urban flood warning systems need fast response times between the rainfall forecast and the flood alarm. Flood forecasts from physically based numerical models usually need much computation time. Flood forecasts based on databases from previous events or pre-simulated events can speed up the process of decision making. This work introduces and compares four distance metrics for temporal rainfall patterns used in a nearest neighbour based forecast system for dynamic water levels and velocities during pluvial floods. The system uses a database of 960 pre-calculated flood events. The performance of each metric is evaluated by analysis of time-series of flow variables. For the error quantification,a procedure to find a small number of representative locations in an urban catchment is described. A new approach to quantify the similarity of dynamic flow fields is introduced, which makes use of particle transport tracking. The four distance metrics are tested on forecast for four exemplary pluvial flood events resulting from rainfall of durations between 15 and 50 min and return-periods between 10 and 100 years. The best metric is based on the temporal precipitation pattern and takes the response time of the drainage system into account. If the pattern has a very short duration, a simpler characterisation-metric can be used, which takes the total volume, peak intensity and peak position into account.

U-FLOOD – Topographic deep learning for predicting urban pluvial flood water depth

This study investigates how deep-learning can be configured to optimise the prediction of 2D maximum water depth maps in urban pluvial flood events. A neural network model is trained to exploit patterns in hyetographs as well as in topographical data, with the specific aim of enabling fast predictions of flood depths for observed rain events and spatial locations that have not been included in the training dataset. A neural network architecture that is widely used for image segmentation (U-NET) is adapted for this purpose. Key novelties are a systematic investigation of which spatial inputs should be provided to the deep learning model, which hyper-parametrization optimizes predictive performance, and a systematic evaluation of prediction performance for locations and rain events that were not considered in training. We find that a spatial input dataset of only 5 variables that describe local terrain shape and imperviousness is optimal to generate predictions of water depth. Neural network architectures with between 97,000 and 260,000,000 parameters are tested, and a model with 28,000,000 parameters is found optimal. U-FLOOD is demonstrated to yield similar predictive performance as existing screening approaches, even though the assessment is performed for natural rain events and in locations unknown to the network, and flood predictions are generated within seconds. Improvements can likely be obtained by ensuring a balanced representation of temporal and spatial rainfall patterns in the training dataset, further improved spatial input datasets, and by linking U-FLOOD to dynamic sewer system models.

Urban Pluvial Flood Management Part 2: Global Perceptions and Priorities in Urban Stormwater Adaptation Management and Policy Alternatives

Urban stormwater infrastructure is at an increased risk of being overwhelmed by pluvial flood events due to climate change. Currently, there are no global standards or frameworks for approaching urban rainfall adaptation policy. Such standards or frameworks would allow cities that have limited time, finances or research capacities to make more confident adaptation policy decisions based on a globally agreed theoretical basis. Additionally, while adaptation via blue-green infrastructure is often weighed against traditional grey infrastructure approaches, its choice must be considered within the context of additional policy alternatives involved in stormwater management. Using six global and developed cities, we explore to what extent a standardized hierarchy of urban rainfall adaptation techniques can be established through a combined Analytic Hierarchy Process (AHP) Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) Multi-Criteria Decision Analysis. While regional and stakeholder differences emerge, our study demonstrates that green infrastructure undertaken by public bodies are the top policy alternative across the cities and stakeholder groups, and that there exists some consensus on best management practice techniques for urban stormwater adaptation.

Measuring emergency medical service (EMS) accessibility with the effect of city dynamics in a 100-year pluvial flood scenario

Emergency medical service (EMS) is important for rescuing victims suffering from life-threatening illnesses or accidents, and is highly time-sensitive by nature. Many uncertain contexts in the urban environment can prolong EMS response time and deteriorate its performance. Using the enhanced two-step floating catchment area (E2SFCA) method, this study measures EMS accessibility with the effect of a regular uncertain context (i.e., the city dynamics like time-varying population and traffic) and an irregular uncertain context (i.e., an extreme pluvial flood event which can cause extensive road closures). The results indicate that, in the central urban area of Shanghai, mid-west areas with denser populations have higher accessibility than eastern peripheral areas. Flooding can cause a remarkable decline of accessibility which falls to the lowest point slightly earlier than the time when the worst road connectivity emerges. The night time exhibits better accessibility than especially the peak hours during the daytime. The GWR results reveal that increasing facility richness and road density while decreasing flood-induced road closures have a positive effect on EMS accessibility. The study indicates that both regular and irregular uncertain contextual factors can influence EMS accessibility in a highly complex manner. Carefully taking these uncertainties into account would enable EMS planning in other contexts and regions to face the enormous challenges posed by the changing climate and increasingly complex urban environment.

Climate change impact assessment on pluvial flooding using a distribution-based bias correction of regional climate model simulations

The expected intensification of extreme precipitation events under climate change likely results in more frequent and intense pluvial floods worldwide. To study the climate change impact on urban pluvial flooding, fine-scale climate model simulations are needed. The number of such simulations is, however, still limited and not widely available, thus entailing the downscaling of the model outputs. One of the commonly used methods to meet this demand is statistical downscaling, where the statistical properties of large(r)‐scale climate simulations are used to derive local climate variables. This study focuses on applying a distribution-based bias correction method on regional climate model (RCM) simulations to explore how climate change affects extreme precipitation and urban flood events at the end of this century (2071–2100). A 1D-2D hydrodynamic model, implemented in Infoworks ICM, is used to simulate pluvial flood events for several return periods for a case study in the city of Antwerp in Belgium. The results show that the statistical downscaling approach can effectively decrease the bias in the model simulations and offer strong scaling relations to derive high-resolution extreme precipitation time series. The analyses also reveal that climate change may cause an increase of 16%, 31%, 47%, 63%, 73% and 84% in the flood volume for 2-, 5-, 10-, 20-, 30- and 50-year return periods, respectively. This projected increase in the flood volume enlarges the inundated area by 32%, 49%, 56%, 58%, 58% and 59% for the respective return periods. The flood frequency is also projected to almost double in the future, so that a 5-year flood event in the historical period will most likely be a 2-year event in the future period.

Impact‐Based Forecasting for Pluvial Floods

AbstractPluvial floods in urban areas are caused by local, fast storm events with very high rainfall rates, which lead to inundation of streets and buildings before the storm water reaches a watercourse. An increase in frequency and intensity of heavy rainfall events and an ongoing urbanization may further increase the risk of pluvial flooding in many urban areas. Currently, warnings for pluvial floods are mostly limited to information on rainfall intensities and durations over larger areas, which is often not detailed enough to effectively protect people and goods. We present a proof‐of‐concept for an impact‐based forecasting system for pluvial floods. Using a model chain consisting of a rainfall forecast, an inundation, a contaminant transport and a damage model, we are able to provide predictions for the expected rainfall, the inundated areas, spreading of potential contamination and the expected damage to residential buildings. We use a neural network‐based inundation model, which significantly reduces the computation time of the model chain. To demonstrate the feasibility, we perform a hindcast of a recent pluvial flood event in an urban area in Germany. The required spatio‐temporal accuracy of rainfall forecasts is still a major challenge, but our results show that reliable impact‐based warnings can be forecasts are available up to 5 min before the peak of an extreme rainfall event. Based on our results, we discuss how the outputs of the impact‐based forecast could be used to disseminate impact‐based early warnings.

Spatiotemporal analysis of heavy rain-induced flood occurrences in Germany using a novel event database approach

Flash floods are a worldwide threat to humans, which is why they are being intensively studied using historical event records. As measurements and event data increase, databases are becoming increasingly important for flash flood research. However, the recent literature on flood databases lacks technical details as well as discussions about a suitable database design for scientific investigations. In this paper, we thus show how an event database for the investigation of heavy rain-induced flood occurrences can be created. Based on the HiOS dataset (a German dataset with ~ 23,800 flash flood and pluvial flood events), we exemplify the database design and explore the spatiotemporal characteristics of floods caused by heavy rain in Germany. We outline all aspects relevant to database setup: from database requirements and system architecture through table and attribute design to a key and relationship definition. Furthermore, we clarify why a spatial database with interfaces for GIS softwares should be chosen, why a damage-based event definition is preferable to a hydrometeorological definition, and how table attributes support differentiated analyses. By means of the database, we investigated frequency, temporal evolution, spatial distribution and patterns, fatalities and injuries, as well as the seasonality of heavy rain-induced floods in Germany. The results indicate that floods caused by heavy rain occur throughout Germany but with a tendency toward fewer events in the northern direction. Across the country, we identified seven hot spots in urbanized and mountainous regions. Although heavy rain-induced floods in Germany take place mostly between noon and late afternoon, most people are injured and killed in events starting in the evening. Our investigation indicates an increased incidence of flash flood and pluvial flood-related injuries and fatalities in the identified hot spots. Overall, we observe a pronounced summer seasonality of the heavy rain-induced flood events. This study highlights the importance of event databases for flash flood research and advances our understanding of heavy rain-induced flood occurrences in Germany.