Adaptive flood risk management: A decision support system integrating deep learning, digital twins, and economic risk assessment

Flood risk is changing over time. Climate change, land-use change, human interventions and socio-economic developments have an influence on the evolution of flood risk. Thus, the future dynamics of drivers influencing hazard, exposure and vulnerability and consequently flood risk evolution is uncertain. Therefore, flood risk management is confronted with deep uncertainties and need to continuously adapt to future circumstances. New management strategies are required to ensure the safety level of humans and their assets and reduce losses from floods. Adaptive flood risk management is a way to cope with such uncertainties. However, the implementation of adaptive flood risk management requires a flood risk monitoring system that screens critical developments of hazard, exposure, or vulnerability and warns the user when a critical point in flood risk evolution is approached. In order to develop a conceptual framework for a flood risk monitoring system, we conducted a systematic review of flood risk evolution assessments. We analysed how flood risk is conceptualised, which factors are assessed to analyse evolutions in one or more risk component, which methods are used to assess flood risk evolution and which risk outcomes are identified. We discuss the main concepts of monitoring the spatiotemporal changes of the components of risks and how the changes of these components contribute to the evolution of risk. We furthermore discuss the data sources, issues of spatial and temporal scales, and how the components of risk coevolve.

This uniquely comprehensive guide for engineers and scientists demonstrates the feasibility of an integrated approach that combines resilient infrastructure with adaptive flood risk management. The book offers practical guidance on managing existing flood defences and designing and planning new ones. Key legislation is discussed together with insights into emerging flood risk management concepts and future developments in policy and practice. The authors address the core concepts of flood risk management and equip readers with the tools and techniques needed to better assess the reliability and adaptability of defences and - ultimately - make robust risk-based decisions. Flood Risk is Placing flood defence infrastructure within the context of a broader flood risk management approach, this key text is essential reading for engineers, scientists, and government and local authority stakeholders who must meet the increasingly complex challenges of flood risk, both now and in the future.

It is well recognized that adaptive and flexible flood risk strategies are required to account for future uncertainties. Development of such strategies is, however, a challenge. Climate change alone is a significant complication, but, in addition, complexities exist trying to identify the most appropriate set of mitigation measures, or interventions. There are a range of economic and environmental performance measures that require consideration, and the spatial and temporal aspects of evaluating the performance of these is complex. All these elements pose severe difficulties to decisionmakers. This article describes a decision support methodology that has the capability to assess the most appropriate set of interventions to make in a flood system and the opportune time to make these interventions, given the future uncertainties. The flood risk strategies have been explicitly designed to allow for flexible adaptive measures by capturing the concepts of real options and multiobjective optimization to evaluate potential flood risk management opportunities. A state-of-the-art flood risk analysis tool is employed to evaluate the risk associated to each strategy over future points in time and a multiobjective genetic algorithm is utilized to search for the optimal adaptive strategies. The modeling system has been applied to a reach on the Thames Estuary (London, England), and initial results show the inclusion of flexibility is advantageous, while the outputs provide decisionmakers with supplementary knowledge that previously has not been considered.

SummaryLand-use change, climate change, human interventions, and socio-economic developments influence the evolution of the risk components hazard, exposure, and vulnerability, and consequently of flood risk. Adaptive flood risk management is a way to cope with evolving risks, but it requires measuring the evolution of risks. To develop principles of flood risk monitoring, we systematically reviewed scientific literature on flood risk evolution analyses. The reviewed publications indicate a wide spread in increase or decrease of flood risk evolution over decades. Furthermore, the publications show a high diversity in factors and methods for flood risk evolution analysis and indicate the main challenges for developing flood risk monitoring. Flood risk monitoring needs the systematic detection of flood risk evolution by periodically (re)evaluate the factors that influence the risk components—hazard, exposure and vulnerability—modeling those risk components and combining them to quantify flood risk.

Climate change and rapid urbanization have increased the risk of urban flooding, making timely and accurate flood prediction crucial for disaster response. However, conventional physics-based models are limited in real-time applications due to their high computational costs. Recent advances in deep learning have enabled the development of efficient surrogate models that capture complex nonlinear relationships in hydrological processes. This study presents a deep learning-based surrogate model designed to efficiently reproduce the spatiotemporal evolution of urban pluvial flooding using data from physics-based models. For the Oncheon-cheon catchment in Busan, the spatiotemporal evolution of inundation at a 10 m spatial resolution was simulated using the physics-based model for various synthetic inundation scenarios to train the deep learning model based on a Convolutional Neural Network (CNN). The training dataset was constructed using synthetic rainfall scenarios based on probabilistic rainfall data, while the model was validated using both a synthetic flood event and a historical flood event from July 2020 with observed ground-based rainfall measurements. The model’s performance was evaluated using quantitative metrics, including the Hit Rate (HR), False Alarm Ratio (FAR), and Critical Success Index (CSI), by comparing results against both synthetic and real (historical) flood events. Validation results demonstrated high reproducibility, with a CSI of 0.79 and 0.73 for the synthetic and real experiments, respectively. In terms of computational efficiency, the deep learning model achieved a speedup 16.4 times the parallel version and 82.2 times the sequential version of the physics-based model, demonstrating its applicability for near real-time flood prediction. The findings of this study contribute to the advancement of urban flood prediction and early warning systems by offering a cost-effective, computationally efficient alternative to conventional physics-based flood modeling, enabling faster and more adaptive flood risk management.

Flood risk management (FRM) is moving towards more proactive and collaborative direction to enable adaptation to changing conditions. We present a case study on collaborative planning process, which contributed to the development of adaptive FRM in one of the largest river basins in Finland. The focus was on the possibility and acceptability of using large regulated lakes as storage for flood water in an extreme flood event to decrease flood damage at the downstream riverside towns. We defined an extreme flood event that would cause dramatic flood damage and developed tools for simulating the event with alternative regulation strategies using Watershed Simulation and Forecasting System (WSFS). We organized a stakeholder event to demonstrate the alternative lake regulation strategies, their socio-economic consequences, and to discuss their acceptability. We found that storing flood water in the lakes above the regulation limits and preparing for winter floods in advance by lowering the lakes in the autumn can minimize the total damage in the target area. The majority of stakeholders considered these actions acceptable in an extreme flood event, regardless of deliberately induced flooding of areas where no floods have occurred for over 50 years. However, lowering the lakes in the autumn on annual basis gained less support. We emphasize the importance of deliberations on the FRM procedures and responsibilities in extreme flood events with the stakeholders in advance to increase adaptive capacity and legitimacy of decisions.

Introduction. In the context of economic, political, and market instability, an effective business organization is based on modern project management approaches that combine and take into account various aspects, including economic analysis, risk assessment, and business ethics. Innovative methods, approaches, mechanisms, and technologies improve and facilitate project development and management. The purpose of the paper is to identify modern approaches of project management based on economic analysis, risk assessment and business ethics for effective business organization in a dynamic environment. Results. In the paper modern approaches of management projects, that include an economic analysis, evaluation of risks and business-ethics and allow to promote efficiency of organization of business, are considered. It was found out that the program facilities of project management are concentrated in directions: a management, administration, resources and also system management projects, that combine budgeting, management, administration of processes, resources and terms. It was pointed on the fact that for companies that will realize several projects in a complex, a project portfolio design is recommended. It was found out that for implementation of analysis of co-ordination of project and objects of investing it is expedient to apply hieratical approach. It was found out that programmatic items for management risks can come forward as applications to the systems of the projects calendar planning, or by the particular programs. It was found out that for management modern projects social analysis and audit, and also observance of principles of digital business-ethics was necessary. Conclusion. The project approach is now an integral part of the business organization of modern companies, ensuring giving continuous development according to the market conditions. An effective project management system ensures the achievement of the strategic goals of the enterprises’ activities, their financial stability, and market position. However, in order to build a modern project management system, modern science-based approaches should be implemented, including economic analysis, risk assessment and management, as well as compliance with business ethics.

Climate change adaptation creates significant challenges for decision makers in the flood risk‐management policy domain. Given the complex characteristics of climate change, adaptive approaches (which can be adjusted as circumstances evolve) are deemed necessary to deal with a range of uncertainties around flood hazard and its impacts and associated risks. The question whether implementing adaptive approaches is successful highly depends upon how the administrative tradition of a country enable or hinder applying a more adaptive approach. In this article, we discern how the administrative tradition in the Netherlands, England, and New Zealand impact upon the introduction of adaptive flood risk management approaches. Using the concept of administrative traditions, we aim to explain the similarities and/or differences in how adaptive strategies are shaped and implemented in the three different state flood management regimes and furthermore, which aspects related to administrative traditions are enablers or barriers to innovation in these processes.

Chapter 32 - Economic Risk Assessment for Field Development

Introduction to the special issue on adaptive flood risk management

As we move into the digital age, the concept of a digital twin of the ocean - a virtual replica that integrates real-time data, environmental conditions and ecosystem dynamics - holds transformative potential for ocean science and resource management. Digital twins promise to provide unprecedented insight into ocean processes, supporting real-time scenario testing, predictive modelling and adaptive management. However, while their value for marine resource management is compelling, the integration of digital twins into fisheries management applications, such as Management Strategy Evaluation (MSE), presents a number of technical and operational challenges.MSE is a critical framework for evaluating and optimising fisheries management strategies, but the operating models (OMs) on which it is based are often simplified single-species population models with limited, if any, environmental feedback. Due to constraints on data availability, computational resources, and the need for transparency and stakeholder engagement, these OMs are typically built to prioritise practicality over complexity. Ecosystem models remain underused in MSE due to these constraints, and Models of Intermediate Complexity are somewhat more applicable as they can be fitted to data in a similar way to single-species models. Therefore, the full potential of sophisticated digital twins, even when they become technologically feasible, may be difficult to realise immediately in MSE and related fisheries management applications.This presentation will examine the technical and operational barriers to adopting Digital Twins of the Ocean for fisheries management applications such as MSE, highlighting the significant requirements for expertise, data integration, computational power and model transparency. We discuss key areas for development, including simplification and modularisation of Digital Twins to make them suitable for fisheries-focused applications, as well as the need for training, standardisation and stakeholder engagement. In conclusion, while Digital Twins of the Ocean hold great promise, their immediate adoption in fisheries management is limited by practical and technological barriers. Overcoming these challenges will require cross-disciplinary collaboration, investment in computational infrastructure, and a roadmap for bridging current OMs with emerging digital frameworks. The path to integrating digital twins into fisheries management must therefore balance ambition with a pragmatic approach, preparing the field for the eventual realisation of fully integrated, adaptive digital management tools.

Effective natural resource management in the face of ecological complexity and uncertainty requires a seamless integration of long-term ecological research (LTER) and adaptive management. Bayesian decision analysis (BDA) provides a robust theoretical foundation for addressing this need, offering a structured framework for decision-making under uncertainty. At its core, BDA integrates three critical components: a causal model that represents the underlying dynamics of the ecological system, a set of alternative management decisions, and a valuation of future states of nature to assess the outcomes of these decisions. a causal model that represents the underlying dynamics of the ecological system, a set of alternative management decisions, and a valuation of future states of nature to assess the outcomes of these decisions. Central to BDA is the concept of the value of information (VOI), which quantifies the benefits of reducing uncertainty in the causal model through additional research or monitoring. VOI provides a mechanism to prioritize data collection efforts that are expected to yield the greatest improvements in decision-making. Similarly, the value of control (VOC) evaluates the potential benefits of influencing system dynamics through management actions. VOC highlights trade-offs between current decisions and future flexibility, aiding in the identification of strategies that maximize long-term ecological and socio-economic objectives. These concepts underscore the dynamic interplay between learning (via LTER) and acting (via adaptive management). The causal model serves as a formalized representation of ecological knowledge, encapsulating the relationships between key system variables, external drivers, and management interventions. By accounting for uncertainties within these relationships, the model facilitates predictions of system responses to various management actions. Alternative decisions, representing possible management pathways, are evaluated based on their capacity to achieve desired outcomes. Valuation of future states of nature, incorporating ecological, economic, and cultural dimensions, quantifies the relative benefits and trade-offs of different management scenarios. Despite its theoretical strengths, implementing Bayesian decision analysis in practice presents significant challenges. One major hurdle lies in defining an accurate and comprehensive causal model of the ecological system. The complexity of ecological interactions, combined with data limitations, often impedes the development of models that are both realistic and computationally tractable. Furthermore, the ability to value future states of nature and apply VOI and VOC concepts remains a formidable challenge. Valuation requires integrating diverse perspectives, accounting for uncertain and delayed outcomes, and navigating trade-offs between competing objectives. To address these constraints, Management Strategy Evaluation (MSE) has emerged as a valuable approximation of full Bayesian decision analysis. MSE uses simulation-based approaches to test alternative management strategies against a range of plausible scenarios, accounting for uncertainty in system dynamics and stakeholder objectives. By generating insights into the robustness and trade-offs of different strategies, MSE provides a pragmatic framework for decision-making when full BDA is infeasible. Developing a real-time BDA model for decision analysis closely parallels the concept of creating a Digital Twin of the ecological system. A Digital Twin is a dynamic, real-time representation of a physical system, continuously updated with data and capable of simulating system behavior under various scenarios. In ecological management, a real-time BDA model serves as a Digital Twin by integrating live data from monitoring systems, updating causal models to reflect current conditions, and providing immediate insights for decision-making. This real-time capability enhances the relevance and accuracy of adaptive management, particularly in rapidly changing environments. Integrating LTER with adaptive management enhances both MSE and BDA. Long-term data refine system understanding, reduce uncertainty, and identify thresholds, improving VOI and VOC assessments. Adaptive management facilitates iterative learning, allowing for real-time updates to causal models, valuation frameworks, and control strategies as new information becomes available. Realizing this potential requires collaboration across disciplines, stakeholder engagement, and investments in computational tools that support decision analysis and Digital Twin implementations. In this presentation I explore the interplay between theory and practice in integrating LTER and adaptive management, emphasizing how Bayesian decision analysis, VOI, VOC, MSE, and Digital Twins provide a pathway to informed, flexible, and resilient decision-making in ecological systems. It also highlights practical limitations and identifies key areas for advancing methodologies. Bridging the gap between theory and practice is imperative for achieving sustainable and adaptive outcomes.

Digital ocean twins represent a transformative tool for integrating marine ecosystem models, observational data, and direct stakeholder input to develop robust management strategies. Here, we introduce one of the first digital twins for the Black Sea—a pioneering model designed to deepen our understanding of this unique ecosystem, forecast its response to climate change and environmental stressors, and evaluate alternative socio-economic scenarios to support informed decision-making.The Black Sea digital twin includes a comprehensive ensemble of integrated simulations and resilience assessments, offering insights into ecosystem states and the risks to the valuable services they provide. Utilizing machine learning and Cumulative Effects Assessment (CEA) methodologies, it functions as a sophisticated decision-support system. This model tests a variety of socio-economic and blue economy scenarios, incorporating analyses of critical sectors and feedback from stakeholders through basin-wide living labs.Through this innovative digital twin, we aim to define a "safe operating space" for the Black Sea—where ecosystem services are preserved and understood, enabling resilient and sustainable coastal societies. The model not only enhances our capacity to predict future changes but also serves as a foundation for adaptive management in a region undergoing rapid environmental shifts.

The purpose of the article is a study of the basic principles, methods, models and tools of managing the ecological and economic security of economic structures. One of the challenges of the long-term future is to increase the growth constraints associated with environmental factors, with the increase of environmental and economic risks and environmental risks, including the occurrence of large-scale emergencies of varying nature. In this regard, further improvement of the ecological and economic safety management systems of production facilities, the creation of subsystems of preliminary ecological and economic risk assessment and the development of measures to reduce these risks are necessary. Accordingly, the principles, methods and tools of environmental and economic risk assessment must meet current requirements for the levels of danger of all aspects of the negative impact on the environment, personnel of the enterprise and the population of the region.Insufficient study of this problem, the need to scientifically substantiate the methodology and practice of improving the system of environmental and economic security management based on environmental risks have led to the choice of research topic.Methodology. The development of methodological aspects of environmental and economic security management involves defining the principles, approaches and methods that form the basis and ensure the effectiveness of enterprise safety research.Results. The article explores the basic principles of environmental management of an economic entity, among them five main principles of compliance that will lead to the effective and efficient management of environmental and economic security. The methods of management of ecological and economic safety are analyzed, namely: organizational-administrative, social-psychological and economic. It is proved that the most effective is the combination of organizational, administrative and economic management methods, because the first to a greater extent reflect the interests of citizens about the quality of the environment, and the second - the economic interests of the nature of users. The models of management of ecological and economic security, which are based on economic mechanisms, are investigated. The main tools of environmental and economic security management are analyzed, the key of which is the environmental management system in the form of an integrated management system. The environmental management system is considered on the example of the integrated management system of the Public Joint Stock Company Azot.Practical implications. The environmental management system in the form of an integrated management system functions effectively and is improved, ensuring the achievement of the goals and objectives defined by the ecological and economic policy of the enterprise.Originality. In order to improve the functioning of the system of management of ecological and economic safety and environmental protection, in addition to technical and technological issues, it is necessary to solve organizational and managerial issues, one of such solutions may be to create an integration of the management system.Creating a comprehensive management system for environmental and economic security provides the enterprise with the effect of synergy of all elements, consisting in the optimal use of material and organizational resources required for the enterprise, with the organization of this system based on the principles of complexity, economy and use of a process approach to all stages of the product life cycle. in conjunction with a functional organization of activities that will allow the company to improve the effectiveness of management and environmental activities ness and create conditions for economic growth.

Risk, including economic risk, is increasingly a concern for public policy and management. The possibility of dealing effectively with risk is hampered, however, by lack of a sound empirical basis for risk assessment and management. This article demonstrates the general point for cost and demand risks in urban rail projects. The article presents empirical evidence that allow valid economic risk assessment and management of urban rail projects, including benchmarking of individual or groups of projects. Benchmarking of the Copenhagen Metro is presented as a case in point. The approach developed is proposed as a model for other types of policies and projects in order to improve economic and financial risk assessment and management in policy and planning.