River discharge is traditionally required with a lot of measurements and data. However there are areas where discharge data is not available. In this research a tool is developed based on hydraulic geometry relations to estimate the discharge of a river in bank full state near Narok, Kenya. With the help of the Global Navigation Satellite System (GNSS) coordinates are converted in to cross-sections of the river in a bank full state. Discharges are assumed and verified through the hydraulic geometry relations of the cross-sections. These discharges are simulated in Super-Fast INundation of CoastS (SFINCS) to simulate the flash floods of Narok and an accurate estimate of the river discharge in a bank full state was obtained.

In July 2021 severe flooding occurred in the South of Limburg, Belgium, and Germany due to heavy precipitation. Extreme precipitation events like this are expected to occur more often in our changing climate. Urbanization is thought to be another contributing factor to the July 2021 flood event. The Netherlands is expected to increase urbanization as a solution to its housing shortage. To make room for urbanization, while minimizing the effect of climate change, the government wants to make ”water and soil leading”, Water en Bodem Sturend, in the decision-making about the layout of the Netherlands.

Therefore, the goal of this research is to investigate the best suitable subcatchment for the construction of new residential houses within the Geul catchment, in terms of flooding. The July 2021 flood event is used as a reference. The first step was to investigate the hydrological response of the Geul catchment. Secondly, this hydrological response was modelled by the semi-distributed hydrological models HBV coupled to D-RR and by the distributed model Wflow_sbm. HBV and D-RR are set up in this research, while Wflow_sbm is adopted from Klein (2022) and Bouaziz (2022). The hydrological models are coupled to the Geul hydrodynamic model D-HYDRO of Hulsman, Weijers, Verstegen, and Goedbloed (2023). The building plans in the Geul catchment were investigated and scenarios were constructed. These scenarios were simulated in the hydrological models. This method resulted in a workflow that can be found in Idsinga (2024). The workflow can be applied on analyses of different land cover types.

The modelled hydrographs showed differences between the hydrological models. Each model better describes one part of the hydrological response compared to the other. HBV and D-RR better represent the subsurface flow and describe the hydrological response during consecutive precipitation events. Wflow_sbm represents the overland flow flux better and therefore describes the hydrological response during the July 2021 flood event. The modelled flood extents during the July 2021 flood event are also compared to the estimated extent by Slager, de Moel, and de Jong (2021). Wflow_sbm showed better similarity to the measured flood extent than HBV and D-RR. The Province of Limburg wants to build 18,730 new houses in the South of Limburg. This results in an increase of 6 km2 paved area. In this research, this increase is applied to different locations in the Geul catchment. Next, the impact of completely paved subcatchments was investigated. The relatively small 6 km² increase in paved area did not result in different discharge behaviour and the total area of the flood extent showed a small difference. However, it impacted the flooded paved area. Building far from the river on the hills resulted in no increase in the flooded paved area. New houses in the valleys, close to the river, are more exposed to flooding. In the Meerssen subcatchment, the added paved area was responsible for 95% of the total increase in the flooded paved area. This was also the case in the Gulp subcatchment, where about 90% of the increase in flooded paved area came from the added paved area.

The Meerssen subcatchment is the most vulnerable to flooding. This subcatchment contains the most paved area and more runoff will result in a more flooded paved area. A completely paved Gulp subcatchment results in a less flooded paved area than building 6 km2 close to the Geul in the Meerssen subcatchment. When the Belgians build new houses in the Sippenaeken subcatchment, the Netherlands will receive more water during an extreme event such as in July 2021.

The letter Water en Bodem Sturend states that new houses must be built in sensible locations. In this research, the location of new houses is found to be important for the hydrological response. Building close to the river results in a more flooded paved area than building far from the river. The Gulp subcatchment is the least vulnerable to flooding and can be considered the best building location for new houses among the three investigated subcatchments.

Fully distributed hydrological models take into account the spatial variability of a catchment, and allow for assessing its hydrological response at virtually any location. However, these models can be time-consuming when it comes to model runtime and calibration, especially for large-scale catchments. Meanwhile, deep learning models have shown great potential in the field of hydrological modelling, but a multivariable, fully distributed hydrological deep learning model is still lacking. To address the aforementioned challenges associated with fully distributed models and deep learning models, we explore the possibility of developing a fully distributed multivariable deep learning model by using Graph Neural Networks (GNN), an extension of deep learning methods to non-Euclidean topologies. We develop a surrogate model of wflow_sbm, a fully distributed, physics-based hydrological model, by exploiting the similarities between its underlying functioning and GNNs. The GNN model uses the same input as wflow_sbm: gridded static parameters based on physical characteristics of the catchment and gridded dynamic meteorological forcings. The GNN model is trained to approximate wflow_sbm outputs, consisting of multiple gridded hydrological variables such as streamflow, actual evapotranspiration, subsurface flow, saturated and unsaturated groundwater storage, snow storage, and runoff. Our results show that the GNN model accurately predicts multiple hydrological variables in unseen catchments (median KGE=0.76), and can serve as an emulator of wflow_sbm with a shorter runtime. We furthermore demonstrate how the GNN model can function up to a prediction horizon of a full year, using physical system states to account for system memory, as well as a curriculum learning strategy combined with a multi-step ahead loss function during training. Overall, this study contributes to the field of fully distributed modelling using a deep learning approach.

Capillary barrier systems (CBSs) have been demonstrated to be effective in deviating infiltration in a landfill cover. However, their performance when combined with a cover soil optimised for methane oxidation had not been tested in the field yet. This study aimed to describe the water balance of a test field where such a configuration was built, located on the landfill in the Wieringermeer area, the Netherlands, over the period from 2009 to 2023.
During that time period breakthrough and deviated infiltration were measured. A one-dimensional finite difference model was built to model evapotranspiration and storage in the cover soil as well as the moisture retaining layer (capillary layer; CL). This model performed well in describing overall seasonal trends but generally overestimated evapotranspiration in spring and outflow in autumn, consequently underestimating the storage in these seasons.
The results show that the annual precipitation ranges between 770 and 990 mm. On average, 59% (494 mm) of this precipitation is evaporated, 33% (281 mm) is diverted by the CBS and 7% (63 mm) breaks through. Compared to other test fields and design standards, this breakthrough is high. A weakened functioning of the CBS by construction errors or the ingress of sand could be a reason for that. Another factor might be the coarse grain size distribution of cover soil, necessary for the efficient oxidation of methane. This type of soil has a relatively high hydraulic conductivity which does not control infiltration rates into the CBS as well as cover soils in other studies.
Furthermore, the results suggest that storage in the cover soil and CL is the main determinant for breakthrough to occur. Major breakthrough only occurred at a storage larger than the storage in the soil associated with field capacity, both on a seasonal and daily time scale. The available storage is mainly affected by the evapotranspiration which shows a seasonal cycle. Consequently, outflows for the CL and breakthrough occur mostly when evapotranspiration is low (autumn, winter). The relation of breakthrough with precipitation is less straightforward as a high rainfall does not necessarily lead to breakthrough on a seasonal or daily scale, depending on the available storage in the system. The distribution of precipitation can matter however. On a seasonal scale, precipitation in late summer can result in high storage in the beginning of autumn, potentially leading to breakthrough. Furthermore, a precipitation amount uniformly distributed over a day leads to less breakthrough than more concentrated precipitation.
Overall, with some adjustments regarding the design of the CBS and the cover soil, this landfill cover design is a promising alternative for the current standard.

The Bui Dam, the second-largest hydropower dam in Ghana, plays a significant role in the sustainable energy mix of the country. It is managed by the Bui Power Authority (BPA) and has a capacity of hydro-clean generation of 404MW, contributing to 17% of the country’s total electricity generation. However, decision-making at the dam lacks the use of predictive models and meteorological measurements. This can lead, in the case of perceived flooding risks and high dam water levels, to valuable water being spilled and endangering the downstream areas. Balancing the tradeoff between energy production and safety can be effectively achieved by implementing predictive models that anticipate peak flows in advance.

Since its commissioning in 2013, the Bui Dam has experienced two instances of emergency spillage, resulting in significant financial losses, property destruction, and displacement of downstream communities. Currently, the reservoir management decision-making process uses two discharge stations upstream, with one of them yielding some unreliable outcomes for high flows. Therefore, it is crucial to prioritize the analysis and updating of rating curves to ensure accurate forecasting.

This research aims to address these limitations by recalibrating the rating curve using the reservoir balance in a conservative manner, i.e. leaning on the safe side to avoid overestimation. Additionally, a conceptual, semi-distributed model was developed simulating high flows, specifically focusing on the years 2019 and 2022 when spillage events occurred. Five different hydrological conceptual models, with three different structures: single, serial, and parallel structures, were tested. The serial model yielded the best results. Then the Black Volta Basin was divided into five sub-catchments, and each sub-catchment was lumped. In the absence of discharge data for the upstream sub-catchments, remote sensing data from GRACE and satellite altimetry (3 virtual stations with data from 2016 to 2022) were used to impose restrictions on the feasible model parameter sets, thereby improving accuracy.

The final model output was calibrated using discharge data obtained from the recalibrated rating curve, along with satellite altimetry data. In the calibrated benchmark case, the model effectively reproduced daily river flows, demonstrating an optimum Nash-Sutcliffe efficiency (NSE) of 0.85 for the period of 2018 to 2022. Subsequently, the model underwent extensive testing under various conditions, including an independent time period without recalibration, different precipitation input sources, transitioning from actual evapotranspiration (AET) to potential evapotranspiration (PET) input, and a change in the testing discharge location. Throughout these testing phases, the model consistently produced favorable results, with NSE values ranging from 0.74 to 0.86.

Furthermore, the model was tested for its progressive predictive capability in simulating the unexpected peak inflows that led to the spillage event in 2019, utilizing iv only precipitation data from the TAHMO precipitation stations, which are openly accessible with near-live timing. The model successfully predicted the occurrence of the large peak inflow, on October 22nd, which ultimately caused the spillage. The model anticipated the occurrence of the ”unexpected” second peak, to some extent, as early as October 12th, providing an 11-day predicting window.

Overall, this research enhances the understanding of the Bui Dam system by implementing a recalibrated rating curve and developing a conceptual model that incorporates remote sensing data. The results demonstrate the model’s capability to simulate past events accurately and predict future inflow patterns, thereby providing valuable insights for effective dam management and spillage prevention.

One significant discovery regarding the character of the Black Volta River at the Bui Dam is the limitation of the prediction period to a strict maximum of two weeks. While the model proves effective within this time-frame, it is advisable for future research to consider incorporating weather predictions to extend this window further. Doing so would enhance the model’s forecasting capabilities and provide even more valuable information for dam operators and decision-makers.

Under future warmer climates, drought events are projected to occur more frequently with increasing impacts in many regions and river basins. This study focuses on exploring the potential of the LSTM deep learning (DL) approach for operational streamflow drought forecasting for the Rhine River at Lobith with a lead time (LT) of up to 46 days.
The research investigates optimal spatial resolution, input and target variables, and loss functions. Four LSTM-based model architectures are developed and tested, incorporating both historical observation and forecast data to generate 46-step forecasts simultaneously. The robustness and stability of the models are assessed through cross-validation, and their performances are compared. Subsequently, the performance of the LSTM-based model is compared to the physically-based models, namely Wflow-Rhine and FEWS-Rhine, in forecasting streamflow drought.
The results suggest that utilizing a subbasin spatial resolution, including historical discharge as input, and training the model on time-differenced data enhance the forecast skill. Among the evaluated models, the model architecture with two LSTMs in cascade exhibits stable and robust performance across the forecast horizon and is considered for operational use in this study. Comparisons between the DL model and physically-based models indicate that: 1) When using observed meteorology forcing from ERA5, the DL model demonstrates a notable performance compared to Wflow-Rhine simulation using the same forcing data. 2) When utilizing SEAS5 for forecasting, the DL model demonstrates skill over Wflow-Rhine in predicting discharge levels during the dry season up to 10 days ahead, as well as for discharges between 950 and 2200 m3/s across the entire forecast horizon. However, for discharges between 700 and 950 m3/s with longer LTs beyond 20 days, Wflow-Rhine shows skill over the DL model. 3) While FEWS-Rhine successfully forecasts drought events in 2018 throughout the forecast horizon, it tends to produce more Type I errors (false positives). The DL model, forecasting with SEAS5, accurately predicts drought events in 2018 for LTs up to 30 days and generally has higher precision values. Despite using different forcing datasets, the DL model can predict the timing and trend of past drought events, indicating its potential in capturing streamflow patterns.
This study contributes to operational water management in the Netherlands by employing the LSTM deep learning approach in an operational framework for drought forecasting. By leveraging historical observation data and forecasted meteorology forcing data, these models achieve skillful performances for streamflow drought forecasts. Future research could focus on further enhancing model performance, exploring the applicability of the LSTM-based models in other river basins, and validating the results in real operational settings.

The integrated management of river catchments is a challenge to be addressed by many water authorities, where some of them even today do not yet incorporate the aquifers’ status in decision-making. This situation has led to the depletion of many aquifers, deeply affecting the drinking water supply, agriculture, and industries. This challenge is combined with the difficulty of decision-making under highly uncertain long-term weather forecasts.
This thesis research proposes a new water management methodology for the Elqui River basin in Chile by using an optimization model aligned with the water authorities’ main objectives and additionally incorporating the aquifer criteria. The optimization model is validated by comparing the results obtained over the 2010–2020 period with the water management practices employed during the same period.
Furthermore, an analysis of the performance of the model using different moving window lengths is executed by the implementation of a Receding Horizon Control (RHC) methodology, evaluating how well the solution is by comparing it with the historical simulation over the same period. The latter is done by looking at the performance of the key optimization goals and using a RMSE and R2 analysis.
Finally, a weather generator was used to randomly generate weather data, based on the 30-year period between 1990 and 2020. The random weather conditions are incorporated in a hydrological model to translate weather data into water volume into the reservoir. Making use of the optimization model, the RHC methodology, and the weather generator, the proposed methodology is tested, enabling the simulation of the decision-making processes. The results are again compared with the water management practices employed over the simulation period.
The research concludes that the proposed methodology brings significant benefits to the aquifers’ status, with neglectable impact on the Desmarque values. Receding horizon (RH) length plays a crucial role, with a balance between achieving optimal results and avoiding computational delays, recommending a RH length of 360 days for best results. The stochastic weather generator effectively replaces unpredictable forecast data, yielding comparable results to real future weather conditions, with temperature and accumulated snowpack playing important roles.

Floods are the most frequent natural disaster and due to climate change the frequency and intensity of these events are increasing. Therefore, it is becoming increasingly important to obtain accurate estimations of extreme discharges. Statistical modelling is widely used to estimate extreme discharges by fitting observed extreme discharges to an extreme value distribution. However, limited historical data makes it difficult to confidently model the tail behavior of extremes. Additionally, several modelling assumptions impact extreme discharge estimates including selection of the nonstationary method, extreme value distribution, parameter estimation method, and the impact of seasonality. In an effort to reduce uncertainties, a new method has been developed to derive design discharges for the Meuse in the Netherlands. This method, GRADE (Generator of Rainfall and Discharge Extremes) consists of three components: a stochastic weather generator, a hydrological model, and an extreme value analysis (EVA). However, the stochastic weather generator is not capable of producing daily rainfall that exceeds the range of historical data. Therefore, a physically based climate model, RACMO, is now being studied. RACMO is capable of generating 1,040 years of synthetic meteorological data that can be routed in a hydrological model to obtain 1,040 years of synthetic discharges. The physically based climate model makes it possible to capture the underlying physical processes of extreme events and the hydrological model can provide discharge information at locations where there are no observations. This thesis evaluates the impact various modelling assumptions have on estimated discharges using synthetic data generated by the RACMO through application of a case study in the Meuse.

Modeling the relationship between rainfall and runoff is a longstanding challenge in hydrology and is crucial for informed water management decisions. Recently, Deep Learning models, particularly Long short-term memory (LSTM), have shown promising results in simulating this relationship. The Transformer, a newly proposed deep learning architecture, has also demonstrated the ability to outperform LSTM in machine translation, text classification, etc. However, there has been limited research on applying Transformers for rainfall-runoff modeling.

The research examined the performance of using Transformer architecture, including its time series forecasting variants, to develop rainfall-runoff models using the CAMELS (US) data set. These models were compared to the LSTM regional rainfall-runoff models, with a particular focus on snow-driven basins as the attention mechanism in Transformer is believed to allow it to attend to the earlier precipitation events in the meteorological forcing. Additionally, the Transformer's potential as a global rainfall-runoff model was also tested using the global Caravan data to determine if it could learn and generalize a wide range of rainfall-runoff behaviors, allowing it to potentially be applied in ungauged basins.

The results suggest that while Transformer and its variants may not be able to fully replace LSTM for rainfall-runoff modeling, the variant called Reformer has shown promise for daily discharge forecasting in snow-driven basins, particularly in terms of peak flow and low flow prediction. However, using the global Caravan data for building a global rainfall-runoff model was not successful due to uncertainty in the forcing data, particularly precipitation. The code for Transformer-based rainfall-runoff modeling is available publicly at https://github.com/Numpy-Panda/neuralhydrology_Transformer.

Recent reports show that deforestation in Latin-America has been severe over the last decades. Especially Brazil is subject to an alarming rate of forest loss, which will remain a factor in the coming decades. In addition to deforestation, there is an increasing amount of hazards like floods and mudslides. These hazards result in major damage to human life and nature. This research analyzes the relation between deforestation and river discharge for observed data, a simple conceptual model and complex model WFLOW for 30 catchments subject to deforestation in Brazil. These three studies use satellite data provided by \cite{GEE} to determine annual cumulative deforestation per catchment. In the study with observed data the relation between deforestation and data from CAMELS-BR \citep{CAMELS2020} is researched by analyzing annual values of two discharge driven parameters: Runoff coefficient and recession coefficient. The second study consists of a simple self-made conceptual model. In this second study, the relation between recession coefficient $\alpha$ and annual deforestation is analyzed by calibrating the model per year for parameter $\alpha$. Lastly, a study is conducted on the performance of WFLOW in catchments subject to deforestation. During this study, several input parameters are changed to observe the response of hydrological processes in WFLOW.

The study with observed data shows both increasing and decreasing trends in coefficients for several degrees of deforestation. Literature shows that landcover type after deforestation is a major factor in the interpretation of these results. However, a lack of quality annual landcover data prevents better research in the non-masked impact of deforestation on discharge. The results of the simple model study show no significant relation between deforestation and recession coefficient $\alpha$. The simplicity of this self-made conceptual model is the weak and strong point of this sub-study. The simplicity makes the results less suitable for analyzing the exact impact of deforestation on discharge, but it is useful for observing general signals and is easily scalable to different catchments. The simulated discharge by WFLOW show a steep overestimation of discharge during peak flow in comparison to observed discharge by CAMELS-BR. Therefore it is not possible to analyze how WFLOW reacts to deforestation. Instead, an in-depth analysis on the cause of this poor performance is conducted by analyzing timeseries for hydrological factors like unsaturated zone depth. The results of this analysis indicate the overestimation of discharge is caused by a lack of outflow from soil layers. In addition, the difference between the Budyko framework of different data sets used in this research, show that uncertainty in quality of input data is a plausible factor on the output of WFLOW.

In conclusion it is observed that deforestation does not necessarily lead to higher runoff coefficients and recession coefficients for measured data in this study area. In addition, the conclusion of the simple model study is that no significant relation between deforestation and recession coefficient $\alpha$ is observed for this conceptual model. Finally, the performance of WFLOW is considered too poor to analyze the impact of deforestation on discharge. The root of this poor performance is considered to be a combination between lack of groundwater modelling and uncertainty on the quality of input-data.

Floods and droughts, also known as hydro-hazards, are phenomena that generally involve detrimental consequences to society and environment. Traditional practices for risk assessment consider flood and drought independently. However, they are two opposite extremes of the same hydrological cycle. Omitting their interaction might lead to an under- or overestimation of the current and future risks associated with such natural hazards. In history, a number of drought-flood interactions have been observed in various parts of the world. Research to these drought to flood interactions is still in its infancy. Therefore, this research explores the concept of consecutive dry and wet (CDW) events in the Netherlands. The aim of this research is two sided. First, the Consecutive Events Graph (CEG) is introduced. This is a radar chart type of graph used to quantify spatial and temporal changes in dry and wet indicators in consecutive seasons. These can be used to identify hot-spots prone to opposite extremes. Second, a fully probabilistic framework based on Non-parametric Bayesian Network is developed to model the dependence between dry and wet indicators. Such model can be used to infer expected wet conditions in a given region when dry conditions are known.
For the CEG and probabilistic model a number of settings were introduced to quantify meteorological dry and wet extremes and to couple them spatial-temporally. First, a number of indicators were selected to quantify both type of extremes. Second, the dry period is defined in summer (June, July and August) and consecutive wet period in fall (September, October and November). Third, the Netherlands was subdivided in five homogeneous regions such that both wet and dry indicators were characterized on a regional scale. Maximum values of the indicators in its corresponding period were calculated for each single region and for every single year between 1965 to 2020.
This resulted in a dataset consisting of quantities for dry extremes in summer and wet extremes in fall for 5 unique regions over 56 years. Application to the CEG shows potential to identify and quantify CDW extremes. Region-to-region and year-to-year comparison is possible to quantify changes between years or regions. Application to the NPBNs disclosed limited interdependencies across the dry and wet indicators. Using the NPBNs for precise forecasting of expected wet conditions is deemed unsuitable as of low precision. Making inference of wet indicators based on hypothetical mild to extremely dry indicators revealed multiple trends of those wet indicators. These trends are increasing for short term precipitation indicators (R1D, R3D, and R5D) and simple precipitation intensity index (SDII) and are mildly decreasing for the total precipitation (Ptot).
Extreme dry events, extreme wet events and consecutive occurrences of these events are inevitable. It is expected that these phenomena will occur more frequently and become more severe due to a changing climate. A number of recommendations for future research is proposed. Findings from this thesis will help to smooth the path towards better understanding of the identification, quantification and interaction of CDW events or multi-hazard events in general.

In this thesis, an easily reproducible modeling approach was developed for assessing the climate change impact on streamflow. This approach was tested by using it to assess the impact of climate change on streamflow in 5 different contrasting catchments across the United States. Many studies show that climate change is expected to influence streamflow regimes all over the world. However, these studies are often difficult to reproduce because the modeling approaches used are usually only locally applicable. In the approach used in this study, hydrological model calibration and validation were done using open-accessible ERA5 forcing together with observed streamflow data provided by the GRDC. The model performed best in a mountainous catchment, while the worst performance was found in a dry catchment and a catchment containing several lakes. The low performances here are mainly caused by imperfect forcing data used for calibration and the neglection of lake processes. The climate change impact analysis used forcing from two CMIP6 models with the SSP245 and SSP585 scenarios. The projections showed significant changes in streamflow in colder regions, which are most likely related to changing snow melt processes. The main finding in warmer regions is that streamflow is generally expected to decrease in the drier periods. Changes of streamflow in these regions are most likely related to changes in precipitation and evaporation processes. However, results remain very uncertain due to disagreements between climate models and sometimes doubtful performance of the hydrological, caused by oversimplification of the model and imperfect ERA5 calibration data. The designed modeling approach facilitates reproducibility of climate change impact analyses in a wide range of catchments using different climate models and scenarios. Its use makes it easier to expand similar analyses to a large ensemble of these aspects.

The Greenland ice sheet (GrIS) is an important component of the climate system and is a key contributor to future sea level rise, as it is storing frozen water that would raise sea levels by 7.4 m should it all melt (Bamber et al., 2018). Of particular concern is the amount of global warming we are facing now and in the future, as it is becoming more likely that even if our emissions are significantly reduced, global warming will reach at least 2∘𝐶 (Arias et al., 2021). Much research concerns the future contribution of the GrIS to sea level rise for high emissions scenarios and low emission scenarios, but there are few studies giving the main focus to what is becoming a more likely future, the moderate emissions scenarios. This research aims to quantify the mass loss of the Greenland ice sheet and subsequent contribution to future sea level rise under a moderate CO2 concentration scenario over a multimillennia timescale. An idealised simulation of 3000 years, where CO2 concentrations are increased by 1% annually until reaching two times pre-industrial values and then kept constant, is run with the high-resolution Community Earth System Model version 2.1 (CESM2.1) and Community Ice Sheet Model version 2.1 (CISM2.1). The climate, run with CESM2.1, is simulated for 1000 years. After 500 years, it is assumed that the climate is close to equilibrium, and thus one climate year is used for five years of forcing the ice sheet in CISM2.1, resulting in 3000 years of ice sheet simulation. At the end of the simulation, the global mean annual temperature has increased by 5∘𝐶 and the temperature over Greenland is 9∘𝐶 warmer than pre-industrial. The rate of sea level contribution in the first centuries is lower than the observed contemporary mass loss of 0.7 mm/yr (Shepherd et al., 2020) but increases after year 710 to a rate of 1 mm/yr. Another increase in mass loss is happening from the year 1380 until the end of the simulation where the rate is 2 mm/yr and the total contribution to sea level rise is 4.1 m. The limited mass loss in the period between years 71-400 and its increase thereafter is found to relate to temporal strong weakening and posterior recovery of the NAMOC. This study projects that the GrIS is a major contributor to future sea level rise, even in a moderate warming scenario, and that the changing NAMOC has a noteworthy effect on the GrIS mass budget.

The water balance model has been an indispensable tool for quantifying water supply and demand and regulating water resources for decades. By convention, hydrologists use in-situ measurements of river discharge for model calibration. Whereas hardly is it possible for areas suffering from data storage. Researchers have been working on methods dealing with this problem for many years. Satellite datasets are potential substitutes for the in-situ observation of hydrological variables. This research proposes a monthly strategy using satellite time series as model inputs. This research aims to apply this strategy in the absence of gauged data for discharge simulations and predictions. Instead of streamflow, the strategy uses variables like actual evapotranspiration (ETa) and/or water storage for calibration. The research challenges lie in the situation that there can be significant errors in data derived from satellite products. Also, hydrological models designed for discharge simulation cannot necessarily function well when calibrated on other terms. This boils down to the research questions as follow:
1. How large are water balance data errors and to what extent can they be reduced?
2. How large are water balance model errors and to what extent can they be reduced?
3. To what extent does quantifying and reducing data and model errors eliminate trade-offs in fitting multiple datasets?
The strategy is constructed based on an error estimation and water balance data fusion method and the original and the advanced version of the Water Partition and Balance model (Wapaba), then tested in the Smoky Hill River catchment. When using unprocessed data, the water balance is not closed for the basin. The discharge simulation has the fitting precision index the Box-Cox transformed root mean squared error (TRMSE) in the range of 0.70 - 1.43 for different datasets and in calibration and validation period. Indexes of discharge fitting σma$ exceed 1.66. After closing the water balance with the mean time series of all fluxes, TRMSE decreases to 0.55. Considering data uncertainties, TRMSE is further declined to 0.29 and σ drops to 0.46. After that, the model structure is also improved. When using the modified model to calibrate on only ETa and TWSA for calibration (TRMSE for discharge = 0.87), the performance is similar to that of using the original Wapaba on all fluxes (TRMSE for discharge = 0.86). The fitting precision index σ for TWSA also decreases.

The research demonstrates the effectiveness of the data fusion method in correcting satellite time series and sheds light on the potential of application of this strategy in the ungauged area through the comparison of different calibration cases. After modification, the strategy is able to reproduce the flow regime, without using in-situ data, to the same degree as all three hydrological components (discharge, actual evapotranspiration and water storage) are used for calibration.

Urban climate adaptation is generally attained --- in a hydrological perspective -- by implementing stormwater management techniques, such as grey, blue and green adaptation measures. Implementing such multi-functional adaptation measures touches on the interests of many stakeholders who need to work together to find resilient and suiting solutions in urban climate adaptation. To facilitate this cooperation, Deltares developed the Adaptation Support Tool. The AST defines eight different performance indicators to compare the effectiveness of urban runoff reduction measures, among which the runoff volume reduction factor. This factor is defined as the rate at which a specific adaptation measure increases the return time of the extreme runoff event. This factor is determined by the underlying model of the AST, the Urban Water Balance Model (UWBM).
The UWBM has the benefit of using long times series instead of single precipitation events (design storms) as input to determine the effectiveness of measures; antecedent conditions for every extreme event are therefore known. Currently, the UWBM is only determining the runoff volume reduction factor. At the same time, the literature shows the importance of analysing the effectiveness of measures using both flow peak reduction and storage peak reduction. Exploration of the possibility to implement these two reduction factors into the UWBM forms the challenge of this study.
This research aims to define a flow peak- and storage peak reduction factor based on modelling systems in the UWBM using long time series. Because the model describes these factors over the measure area, the second goal of this study is to investigate the possibility of converting this measure factor to a whole project area factor and combining this factor for a combination of measures. Main conclusions of this research are as follows: A flow peak reduction factor can be empirically determined based on the proposed method in this study. A storage peak reduction factor can not be empirically determined based on the proposed method in this study. It is recommended to do more research into the analysis of the storage peaks using time series, since it can give an extra insight into the effectiveness of measures. Converting the found flow peak reduction factor for the measure area to the project area with multiple interventions is substantiated with a proposed equation. However, it is recommended to do more research into the conversion equation to be used in the Adaptation Support Tool. By determining the flow peak reduction factor together with the runoff volume reduction factor, the effectiveness of urban runoff reduction measures in a project area can be quantified and used to compare alternative solutions for their effectiveness in flood risk reduction. The added flow peak reduction factor gives a more thorough insight into this effectiveness of measures.

Accurate estimates of terrestrial water storage variations (TWSV) are critical for a variety of applications, e.g., model calibration and climate studies. This study aims to find the added value of river run-off data for regularizing GRACE mascon solution, from which TWSV can be estimated. Most subbasins of the Mississippi Basin show an exponential relationship between run-off and TWSV with moderate to good predictive value. The mean value of the explained variance (R^2) for the models is 0.4, when excluding the Lower Mississippi and Middle Missouri subbasins the mean value increases to 0.5. The Lower Mississippi and Middle Missouri subbasins are the only subbasins with no clear correlation between run-off and storage, most likely due to aquifer depletion.

Since ready-to-used GRACE mascon solutions cannot be modified, the mascon method developed at the Geoscience and Remote Sensing Department of the Delft University of Technology is used in this study. This study proposes modifications to tailor this method to the Mississippi Basin and mass anomalies of hydrological origin. After applying standard Tikhonov regularization, the exponential relationship is used to determine the most probable model outcome for each subbasin, adding a new term to the estimator. The resulting GRACE mascon solutions (Tikhonov and run-off regularized solutions) have a root-mean-square deviation (RMSD) of around 3.5 cm when compared to ready-to-use mascon solutions from the Jet Propulsion Laboratory and Goddard Space Flight Center. The run-off regularized solutions show smaller deviation (on average 7 %) than the Tikhonov regularized solutions with respect to an independent validation data set. The largest difference between the Tikhonov and run-off regularized solutions occur in the dryer subbasins (Platte, Middle, and Upper Missouri, and Upper Mississippi subbasin). These areas show a stronger inter-annual trend in TWSV, these trends are captured better by the run-off regularized solutions. Additional research, co-estimating the regularization bias, shows that the run-off regularized solutions induce less bias into the solutions. This study shows that using run-off data when processing GRACE data could be of added value, especially in semi-humid to arid areas, since the TWSV in these areas is more susceptible to inter-annual variations.

The control of infiltration and seepage of water is one ofthe most challenging tasks in water management and civil­-engineering and, inan attempt to control this, methods for forming a water­impermeable layer inthe soil have been widely practised in soil engineering (Laumann et al., 2018;Proto et al., 2016). The use of natural processes to modify the engineeringproperties of the subsurface could help to develop cost-­effective, robust andsustainable engineering technologies and is attracting increasing attentionfrom the industry (Zhou, 2020). This research aims to reduce the permeabilityby using aluminium (Al) and organic matter (OM) precipitates mixed ex­-situwith porous media to create a horizontal barrier. The Al­-OM precipitates wereexpected to clog the pore space with a reduction in permeability and hydraulicconductivity as result. To find out if it is feasible to use Al­-OMprecipitates mixed ex­-situ with porous media for a permeability reductivelayer, the Al­-OM precipitates and the permeability of the medium wereresearched.  To characterise the flocculation reaction, experiments onthe flocculation were performed. The yield of the reaction was obtained byadding certain amounts of Al and OM solutions to form particular amounts of drymass of flocs. This experiment confirmed the hypothesis that 85% of the addedmass of OM will result in dry mass of flocs. The concept of a critical metal tocarbon ratio (M/C­-ratio), indicating flocculation regardless of the inputconcentrations, was tested by measuring the pH over an increasing M/C-­ratio.From this titration curve, the found critical M/C ratio is between 0.023­0.031,and the pH stabilises at a level lower than pH 4. This result proves theconcept behind the numerical scenarios describing the titration of OM solutionswith Al3+, with a critical molar M/C ratio independent of the inputconcentrations (Veerkamp, 2018; Zhou, 2020). After determining the yield of thereaction and concluding that the concentration of Al and OM was not ofinfluence, the by-products of the Al­-OM reaction were quantified. The ionicstrength of the supernatant of an increasing density of flocs in solution wasdetermined by measuring the electrical conductivity (EC). The results showedthat the ionic strength increased linearly with an increasing density of flocs.The relationship between the ionic strength and the density of flocs was coupledto the linear relationship between concentrations potassium chloride (KCl) andits EC. From the results, the measured EC can be used as a tracer since the K+and Cl­ are non­reactive.  The hydraulic conductivity measurements were conducted by afalling head test to be able to make an indication of the change inpermeability when adding the Al­-OM precipitates to the sand. To find theoptimal method to mix the Al, OM and porous media, the influence of differentmethods of producing, adding and mixing the materials on the permeabilityreduction is explored. In the first mixing method, the Al and OM were added insolution, the solution containing Al-­OM precipitates was centrifuged until thereduced ionic strength was at an EC value less than 700 µm/cm. The hydraulicconductivity measurements were used to obtain the relationship between thehydraulic conductivity reduction over an increasing concentration of flocsretained by one kilogram of sand. The hydraulic conductivity was reducedexponentially over an increasing concentration of flocs up to a magnitude of 3.The results imply a large variability in the achieved reduction dominated bythe amount of retained flocs. The second method is using Al and OM in powderformat and adding them to the dry sand and adding 500 ml of water to thismixture. This mixing method resulted in a completely different floc structure.The flocs produced by mixing in solution have a shear dependency feature, whiledry mixing created particles that have a constant size. For this method, theincrease of the reduction is linear over an increase of concentration of flocsretained by the soil. The highest reduction for this method was found to be ofa magnitude of two, measured at 50 grams of flocs retained by one kilogram ofsand. Finally, this research gave proof of principle of using Al­-OMprecipitates mixed directly with sand could reduce the permeability up to amagnitude of 3. These results present a new road to research on this Al-­OM­-sandmixture’s strength parameters and compaction over time over an increasing flocdensity, since these parameters are critical for using the layer in practice.  

Droughts are considered to be one of the most damaging, yet least understood, natural hazards of all. Despite their prevalence, a thorough understanding of them lacks because they are such complex phenomena, and their manifestation can differ depending on the region they occur in. Monitoring hydrological variables and processes is imperative for a good understanding of how droughts develop and persist. Backscatter from ASCAT and previous scatterometers has long been used for soil moisture retrieval. The first and second order derivative, slope and curvature respectively, of the backscatter - incidence angle relation in the TU Wien Soil Moisture Retrieval algorithm are used to correct for vegetation effects. Recently, new developments to this algorithm have allowed to account for interannual variations in the slope and curvature. This has given rise to the potential of monitoring vegetation directly with slope and curvature, rather than only using it to correct for vegetation effects in soil moisture retrieval. The long data record of ASCAT and previous scatterometers combined has the potential to provide valuable information for drought monitoring. This study investigates if ASCAT could be used as a self-contained dataset in drought monitoring. The spatial variability, the seasonal cycle, and the drought response of backscatter, slope and curvature across different vegetation types in Australia is assessed. Simulated surface- and root zone soil moisture, LAI and GPP from the land surface model ISBA are used to aid in the interpretation of the ASCAT signal. The results from this study show that backscatter, slope and curvature can adequately capture vegetation dynamics in times of drought across dry semi-arid grasslands and croplands. Over these regions the soil moisture and vegetation anomalies observed with ASCAT and simulated in ISBA correspond well. Considerable information into the vegetatin dynamics can be gained from analyzing the backscatter - incidence angle relationship. Especially the ability to monitor drought in crops with a coarse spatial resolution is promising for future applications. It proved more difficult to accurately capture the propagation from a soil moisture anomaly into vegetation anomaly across forests and mixed vegetation with grasses and trees. The first reason for this is the increased attenuation of the signal by vegetation, which hampers accurate measurements of soil moisture content. The second reason is that it is more difficult to separate the soil moisture and vegetation effects due to the fact that less is known about the scattering mechanisms induced by vegetation structure and moisture distribution. Overall the results support earlier findings the slope can be used as a measure of vegetation wet biomass and confirm that curvature is also a valuable source of information that gives insight into the relative contribution from surface or volumetric scattering to total backscatter. These relations have been shown to also adequately describe vegetation dynamics in times of drought.

Rainfall-runoff modelling is essential for short- and long-term decision-making in the water management sector. The accuracy of streamflow predictions of hydrologic models increases with the availability of and the access to streamflow observations. Therefore, one of the key challenges in the field of hydrology is to produce Predictions in Ungauged Basins (PUB), where observations are lacking. Recent research has shown the potential of deep learning neural networks as an alternative approach to conceptual and process-based hydrologic models for this purpose. 
In this study, the existing Multi-Timescale LSTM (MTS-LSTM) architecture is used to investigate if such a deep learning network is able to learn universal hydrologic behaviour. Therefore, a MTS-LSTM is trained on a large variety of >500 US catchments and subsequently tested outside the US, in the European Meuse river basin. The model is not retrained or finetuned to simulate an ungauged situation and to asses whether the streamflow predictions can compete with those from the uncalibrated distributed model wflow_sbm. 
Results indicate that the MTS-LSTM trained on US data cannot compete with wflow_sbm in the Meuse catchments. The simulated streamflow time series can be unrealistically shifted and scaled compared to the time series of observed streamflow due to sensitivity regarding static model input. This means, the values for catchment characteristics cannot be extremer in the testing data than in the training data. Therefore, it is recommended to select catchments for the model training such that the most extreme conditions are covered. In the case that the MTS-LSTM is trained for a specific region, results clearly compete with or outperform the distributed model. For the Meuse test catchments, the neural network achieves Nash-Sutcliffe Efficiency (NSE) values>0.46 where the application of wflow_sbm is problematic and yields negative NSE values. To exploit the potential of an LSTM, the model should be trained on all available data of the entire Meuse basin instead of on the subset of catchments used here. For some water management applications it is important to accurately predict high flow events. Training the MTS-LSTM with the Mean Quadrupled Error (M4SE) loss function showed that the peak flow representation can improve for catchments where the use of a NSE loss already leads to good predicting performance. Thus – for and ungauged catchments – an implementation of a combined loss function appears a valuable follow-up research. 
For ungauged catchments, these results imply that the global neural network model as tested in this study should be supplemented with finetuning. Thereby, the global model could not yet be applied everywhere, however, in regions where only few years of streamflow records are available. Alternatively, a regional neural network model trained on nearby catchments could be applied for PUB, if streamflow observations are accessible in the surrounding area. Finally, it is of high importance to maintain and extend the network of streamflow gauging stations globally, and to ensure easy access to the data.

Human interventions in the White Volta River in Ghana increase the flood risk in living spaces of the local communities, causing great challenges in the area. Burkina Faso has built a hydropower dam near the border with Ghana which has a large impact on the discharge extremes of the White Volta River. As a result of poor water management of this dam, water is spilled during the wet season, causing more floods in northern Ghana. Several communities, living near the White Volta River, depend on the river water for their drinking water demand and use the sand in the river bed as construction material. This puts them in a vulnerable position when changes in the natural flow regime of the river occur.

This thesis studies the impact of constructing multiple small dams in the White Volta river in order to decrease the flood risk in villages located close to the river and create an opportunity for controlled sand mining as a result of sedimentation.

Tamale is, with 672.000 inhabitants, the largest city in this area and struggles with the consequences of the floods as well. The drinking water company has trouble meeting the drinking water demand during both the wet and the dry season. When the area is flooded, the water intake point shuts down, and during low water levels, the pumps can get clogged. This clogging occurs even more often as a result of increased illegal sand mining from the river banks by the local communities. This has broadened the river, resulting in lower water levels and a higher turbidity.

A possible solution to ensure the water intake in Tamale and reduce the flood risk in northern Ghana is the construction of multiple small dams in the river bed to flatten the discharge peaks and slow down the water in the White Volta River. At the same time, these dams can create an additional advantage by causing upstream sedimentation that provides a possibility for controlled sand mining in the river bed. To research the effectiveness of this solution, the impact of the dams on flood risk and sedimentation, near seven villages in northern Ghana, was modelled.

A base case of the flood of 2003 was compared to a scenario in which seven dams were implemented at locations close to villages that are often exposed to floods. This was done by building a hydromorphological model of the White Volta River, using the D-HYDRO 1D2D software.
The results of the morphological simulation show that in total, the sedimentation upstream of the dams can fill 81 trucks with sand per day. However, extracting the sediment at the most upstream dams decreases the possibilities for sedimentation near the downstream dams. The results of the 1D2D hydraulic model show that the proposed dams are not able to reduce the flood extent in villages and can even increase inundation depths, resulting in more flood damage.

It is recommended to increase the 2D grid of the hydraulic model in order to receive more realistic backwater curves.