Flooding is the natural hazard most likely to affect individuals and can be driven by rainfall, river discharge, storm surge, tides, and waves. Compound floods result from their co-occurrence and can generate a larger flood hazard when compared to the synthetic flood hazard generated by the respective flood drivers occurring in isolation from one another. Current state-of-the-art stochastic compound flood risk assessments are based on statistical, hydrodynamic, and impact simulations. However, the stochastic nature of some key variables in the flooding process is often not accounted for as adding stochastic variables exponentially increases the computational costs (i.e., the curse of dimensionality). These simplifications (e.g., a constant flood driver duration or a constant time lag between flood drivers) may lead to a mis-quantification of the flood risk. This study develops a conceptual framework that allows for a better representation of compound flood risk while limiting the increase in the overall computational time. After generating synthetic events from a statistical model fitted to the selected flood drivers, the proposed framework applies a treed Gaussian process (TGP). A TGP uses active learning to explore the uncertainty associated with the response of damages to synthetic events. Thereby, it informs regarding the best choice of hydrodynamic and impact simulations to run to reduce uncertainty in the damages. Once the TGP predicts the damage of all synthetic events within a tolerated uncertainty range, the flood risk is calculated. As a proof of concept, the proposed framework was applied to the case study of Charleston County (South Carolina, USA) and compared with a state-of-the-art stochastic compound flood risk model, which used equidistant sampling with linear scatter interpolation. The proposed framework decreased the overall computational time by a factor of 4 and decreased the root mean square error in damages by a factor of 8. With a reduction in overall computational time and errors, additional stochastic variables such as the drivers' duration and time lag were included in the compound flood risk assessment. Not accounting for these resulted in an underestimation of 11.6 % (USD 25.47 million) in the expected annual damage (EAD). Thus, by accelerating compound flood risk assessments with active learning, the framework presented here allows for more comprehensive assessments as it loosens constraints imposed by the curse of dimensionality.

Worldwide, estuaries are increasingly constrained by human interventions, such as wetland reclamations. Intertidal area has an important influence on the extent of estuarine salt intrusion. Previous research has shown conflicting effects of intertidal area on the salt intrusion. Therefore, this study explores this interaction for three estuary classes: (a) salt wedge, (b) partially mixed, and (c) well-mixed. Our findings show that the effect of intertidal area on the salt intrusion depends on the estuary class: enlarging the intertidal area reduces the salt intrusion for salt wedge and partially mixed estuaries, but vice versa for well-mixed estuaries. These opposing responses are explained by the balance between salt fluxes driven by the estuarine circulation versus by the tidal oscillation. In general, enlarging intertidal area results in the suppression of the estuarine circulation. Such system understanding is especially relevant in an era of increasing coastal urbanization.

Worldwide, estuaries are centres of life. Their connection with the open seas while providing fertile lands and freshwater has resulted in these regions becoming major hubs of human settlement and activities, which is expected to continue growing in magnitude. However economically useful, the open connection to the seas also poses a challenge: freshwater availability. Via this open connection saline water can enter the estuary and contaminate the freshwater reserves - so-called (estuarine) salt intrusion. This may cause freshwater shortages and damage to all living beings that depend on this invaluable resource.

This research aims to develop mitigation measures to this estuarine salt intrusion following the Building with Nature-philosophy. Thus, the goal of this research is to develop nature-based solutions to mitigate salt intrusion. This is achieved by (1) exploring the potential of estuary-scale interventions to affect salt intrusion; (2) conceptualising various nature-based solutions that mitigate salt intrusion; (3) evaluating nature-based solutions in a multidisciplinary context; and, to conclude, (4) reflecting on the role of nature-based solutions.

The exploration of potential estuary-scale modifications is a computationally expensive endeavour for which a novel simulation strategy is developed. This strategy is presented in Chapter 2 and proposes the use of machine learning techniques to determine the input space - i.e., which model simulations to execute, and which to exclude. The aim of the strategy is to put more focus on exploring the output space instead of exploring the input space.

Subsequently, the model simulations are analysed in Chapter 3. Thus, Chapters 2 and 3 present respectively the method and the results of an extensive sensitivity analysis of estuarine salt intrusion to estuary-scale modifications. The end-result of Chapter 3 includes a shortlist of potential nature-based solutions to mitigate salt intrusion, including a ranking based on the sensitivity analysis.

The conceptualisation of nature-based solutions focuses mainly on two potential options: (1) a (temporary) sill, or submerged dam (Ch. 4); and (2) enhancement of intertidal area (Ch. 5). In addition to these novel nature-based solutions, Chapter 6 evaluates a third nature-based solution: shallowing of an estuary. From a physical perspective, this mitigation measure is well-known but mainly poses challenges in the socio-economic domain, which is why it is not extensively covered in the conceptualisation-phase of this dissertation but the evaluation-phase instead.

Both the sill and the intertidal area show a dependency on estuary class in how effectively the mitigation measures are. For the sill it holds that the weaker the tide, the more effective the sill mitigates landward salt transport - i.e., salt intrusion. When the tide is limited, salt intrusion is largely (if not fully) driven by gravitational circulation, for which the sill functions as a wall beyond which the formed salt wedge can hardly penetrate. However, with tidal energy increasing, the tidal momentum to push the saline water over the sill also increases resulting in more salt intrusion - the sill functions more like a speed-bump than a wall. All in all, a sill is most effective for estuaries with little tidal influence.

Enhancement of intertidal area increases the mixing in the estuary. When the dominant salt transport mechanism is related to the estuarine circulation, this enhanced mixing by increasing the intertidal area reduces the salt intrusion. However, the opposite holds for estuaries in which the salt transport is dominated by the tidal oscillation. This means that in case of salt wedge and partially mixed estuaries, enhancement of the intertidal area reduces salt intrusion; and in case of well-mixed estuaries, the intertidal area increase promotes salt intrusion. Thus, the additional (vertical) mixing caused by the intertidal areas reduces salt intrusion as long as there is something to mix - i.e., as long as there is a (sufficient) vertical salinity gradient.

The evaluation of nature-based solutions is inherently multidisciplinary. In this dissertation,
two different perspectives are considered: (1) socio-economy (Ch. 6), and (2) socio-ecology (Ch. 7). In both cases, the Rhine-Meuse Delta is considered as case study.

The socio-economic evaluation addresses the effect of water depth on two major stakeholders in an estuary: a port, and water boards. These stakeholders have opposing interests regarding the water depth: a port benefits from enhanced water depth to facilitate larger vessels, but the resulting contamination of freshwater reserves via salt intrusion is negatively affecting water intakes, and everyone that depends on them. Chapter 6 presents a multidisciplinary evaluation method based on Pareto-fronts, which are to inform decision- and policy-makers. The Pareto-front in Chapter 6 shows that the port performance remains relatively unaffected for shallowing until two metres with the current bed levels, while the freshwater availability improves. However, beyond this level of shallowing, port performance drops substantially against limited gains in freshwater availability.

The socio-ecological evaluation focuses on the effects of reopening a closed-off estuary on two opposing interests as well: freshwater availability, and estuarine ecosystem functioning. Reopening is beneficial for the ecological diversity in the former estuary, but comes at the costs of returning saline influences into a freshwater lake - i.e., freshwater availability is hampered. The ecological implications are mapped by translating hydrodynamic (model) data to an ecotopes-potential map due to which the ecological impact of interventions can be quantified. As in Chapter 6, Pareto-fronts are used as method to inform decision- and policy-makers. The Pareto-front in Chapter 7 shows an increased ecological diversity in the former-estuary without impeding the freshwater availability when partially opening the gates. However, once the salt intrusion reaches the most western water intakes, the freshwater availability drops without major gains in diversity.

This dissertation has shown the complexity of developing future-proof nature-based solutions to mitigate salt intrusion. Building on the lessons learned, the reflection of this dissertation proposes a next step for nature-based solutions: DARE (diverse, adaptive, and robust engineering; Ch. 8). DARE is based on nature's own approach to dealing with uncertainties. The three axes of DARE show how to deal with uncertainties in the three dimensions of engineering: (1) a diverse set of solutions to deal with uncertainty in forcing conditions; (2) an adaptive approach with room to change course in response to the uncertainty that comes with time; and (3) robust - or even antifragile - solutions regarding the uncertainty in performance. With DARE, the focus shifts from the input to the output: Which part of the output is desirable, and what are the options to get there?

The nature-based solutions presented in this dissertation form a starting point for further explorations, with work in this dissertation already being used as a stepping stone for other studies. Besides further exploring the opportunities of nature-based solutions to mitigate salt intrusion, next steps also include bringing these findings into practice. This includes reassessing past estuarine modifications from a new perspective, one that is less susceptible to the challenges that the future may have in store.

At a global scale, deltas are vital economic hubs, in part due to the combination of their access to inland regions via river systems with their proximity to sea. However, with the sea in close vicinity also comes the threat of freshwater contamination by saline seawater, especially during droughts. This study explores the potential of a mitigation measure to estuarine salt intrusion, namely the construction of a (temporary) earthen sill—a measure implemented in the Lower Mississippi River near New Orleans (LA, USA). This study suggests design guidelines on how a sill can be used to mitigate estuarine salt intrusion: the design should focus on the longitudinal placement and the height of the sill, and the mitigating efficiency of the sill reduces with increasing tidal range. Overall, a (temporary) sill has great potential to reduce salt intrusion in salt wedge estuaries if there is sufficient water depth available.

Estuaries worldwide are of substantial ecological value due to the presence of various gradients, such as salinity. Preserving the natural value of estuaries is vital for meeting the climate stabilization goals of the Paris Agreement. Recognizing nature as a stakeholder is imperative, given the surpassing value of ecosystem services over global gross domestic product. Quantifying the current ecological state and future ecological shifts faces challenges, including variable dependencies, spatial-temporal disparities, and the limitations in available information. This study introduces EMMA (Ecotope-Map Maker for Abiotics), a method for quantifying the effects of human interventions or climate change scenarios on estuarine ecosystems by linking abiotic characteristics derived from a hydrodynamic model to ecotopes. The Western Scheldt, an estuary connecting the Scheldt river to the North Sea in the Netherlands, serves as a case study. The method successfully reproduced an existing ecotope-map, which is dependent on real-time data such as aerial photographs. The developed method not only proves applicable in assessing the current ecological state and future ecological shifts for hypothetical scenarios but also demonstrates utility in predicting future situations, providing valuable insights for decision-makers in estuarine ecosystem management and contributing to climate and environmental preservation goals.

In recent years, coastal management has been facing new challenges: socio-economic growth and consequent climate change impose new boundary conditions pushing coastal systems towards unseen states. For adaptation and mitigation strategies as well as risk management, the resilience of systems to these projected changes must be tested and quantified using predictive tools, given the scarcity of observations. Process-based models, which limit the number of assumptions, are the preferred tools. However, these models are computationally expensive and therefore unattractive for global sensitivity and uncertainty analyses. Input and model reduction techniques, as well as behavioural empirical models, have been widely used to overcome these computational difficulties. In this paper, we propose a process-based hybrid workflow—that combines statistical and machine learning with a process-based numerical model—to provide sensitivity analyses on complex systems. As an example we explore salt intrusion in estuaries. The novelty of the method presented is the implementation of an adaptive sampling technique of numerical experiments with a process-based hydrodynamic model, and the training of a neural network to augment the set of numerical runs executed. The first uses predictive uncertainty to automatically explore the response of the complex system to varying environmental boundary conditions and geomorphological configurations. The second is trained to provide system responses around the sampled points. This exploration is closed by simulating the extremes in the output space as found by a genetic algorithm. This scheme is shown to be highly efficient in non-linear, heteroscedastic, and highly non-stationary systems.

All around the world, deltas are among the most densely populated and heavily utilised regions, where crucial functions, such as freshwater availability and safety against flooding, strongly relate to the natural dynamics of the system. Therefore, a thorough understanding of the estuarine system is crucial, especially when developing nature-based solutions for safeguarding these essential functions for today’s society as well as future generations. To better understand the effect of different estuarine parameters on salt intrusion, an extensive sensitivity analysis has been executed based on an idealised estuary layout. The idealised estuary is parametrically designed using thirteen parameters that represent both boundary conditions and geometric features, such as river discharge and water depth. Subsequently, the Delft3D Flexible Mesh (DFM) model has been used to determine the salt intrusion, allowing the exploration of a wide range of estuary layouts.

Estuaries are among the most densely populated and heavily utilised regions in the world, where crucial functions – e.g., freshwater availability and water safety – strongly relate to the natural dynamics of the system. When developing nature-based solutions to safeguard these essential functions, a thorough understanding of estuarine dynamics is required. This study describes an elaborate sensitivity analysis on the salt intrusion length using an idealised estuary, which is parametrically designed using key estuary-scale parameters – e.g., river discharge and tidal flats – to cover a wide range of estuary classes. We were able to systematically investigate such a wide range of estuary classes due to the combination of (1) state-of-the-art hydrodynamic modelling software, (2) high performance computing, and (3) reduction and analysis techniques using machine learning. The results show that the extent of the estuarine salt intrusion length is largely determined by four estuarine features: (1) river discharge; (2) cross-sectional area (especially water depth); (3) tidal damping/amplification; and (4) tidal asymmetry. In general, the salt intrusion length shows clear correlations with (a combination of) estuary-scale parameters, which all put an upper limit on the salt intrusion length. These relations provide crucial insights for successful development of nature-based solutions to mitigate salt intrusion in estuarine environments.

The increasing pressure on Earth's ecosystems due to climate change is becoming more and more evident and the impacts of climate change are especially visible on coral reefs. Understanding how climate change interacts with the physical environment of reefs to impact coral growth and reef development is critically important to predicting the persistence of reefs into the future. In this study, a biophysical model was developed including four environmental factors in a feedback loop with the coral's biology: (1) light; (2) hydrodynamics; (3) temperature; and (4) pH. The submodels are online coupled, i.e. regularly exchanging information and feedbacks while the model runs. This ensures computational efficiency despite the widely-ranged timescales. The composed biophysical model provides a significant step forward in understanding the processes that modulate the evolution of coral reefs, as it is the first construction of a model in which the hydrodynamics are included in the feedback loop.