The transition towards a new risk approach for the Dutch national safety assessment of primary flood defences has been taken as an opportunity to improve the dealings with uncertainties. The probabilistic models for the new safety assessment (WBI2017) not only deal with the natural variability, but also with so-called epistemological uncertainties. One class of epistemological uncertainties is the model uncertainty in the hydrodynamic models used. The quantification of the water level uncertainty depends on the dominant hydraulic processes in each water system and is chosen to be independent of the return period. However, water level frequency lines derived including model uncertainty sometimes conflict with the physics. A comparative analysis is carried out to assess the performance of Hydra-NL w.r.t. observations and to analyse if physical processes that play an important role in each fresh water system are represented correctly after the inclusion of model uncertainty with the WBI2017 method. This study showed that the estimated exceedance probability of actual water levels in the tidal river area and lake area are often overestimated by the Hydra-model even without model uncertainty. When model uncertainties are included the overestimations become even larger. Furthermore, the inclusion of model uncertainty sometimes gives an incorrect representation of the underlying physical processes. The effect of the model uncertainty gets larger when the water level frequency line becomes flatter. This is e.g. the case at locations where the water levels are influenced by the closure of the Europoortkering and upstream of the flood channel near Veessen-Wapenveld. It can be argued that water level uncertainties are likely to decrease in these situations, because a reservoir is more predictable than a flowing river and the operation of the flood channel results in an enlarged conveyance that is less sensitive than an average river profile. The hypothesis that the inclusion of model uncertainty according to WBI2017 results in an incorrect representation of the underlying physical processes is further analysed for several case studies in the upper river area and the lake area. Varying river schematisations and two river interventions (flood channel and retention area) are modelled to compare the WBI2017 method with a more physics-based approach. In this physics-based approach the hydraulic roughness of the main channel and/or floodplains is incorporated as additional stochastic variable for the derivation of water level frequency lines. It is shown that the water level uncertainty becomes smaller for wide rivers where the water level frequency line becomes flatter. The cases where river interventions are present the water level uncertainty is bounded, because of the environment/geometry that influences the water levels. According to the physics-based method the water level uncertainties are location and discharge dependent, which are both not integrated in the WBI2017 method. For the lake area the focus is on locations where high water levels are predominantly determined by the wind that is causing a set-up on the lake. By considering an empirical parameter of the "capped Wu" formula as additional stochastic variable the uncertainty of the drag coefficient is modelled for the physics-based method. The physics-based method demonstrates that the water level uncertainty is larger for higher decimate heights and vice versa. It can be concluded that it is not valid to choose one uniform standard deviation of the water level for all wind-dominated locations, as it is done for WBI2017. It is recommended to include the model uncertainty by considering a model parameter in the hydrodynamic model for the upper river area and lake area as additional stochastic variable in the Hydra-models, to model uncertainty in the water level. In other water systems it becomes more complex, because several equally important sources of uncertainty are present. Still, the most important model uncertainty sources could considered stochastically, but computational effort would become the limiting factor

Conventional practices of rating curve computation fall short in many aspects. They are data-intensive and are notoriously inaccurate in high-flow regimes as a result of extrapolation of low-flow curve fitting. By making use of a simple commercial UAV, a physically-based rating curve is computed that is substantiated by a detailed representation of the section's geometry. Discharge data is still required for calibration of the roughness coefficient and determination of the stage at zero flow. The physically-based rating curve improves the accuracy of the rating curve in the high flows and allows for an easy separation between flow regimes.

This report introduces the use of Genetic Programming (GP) into hydrology by describing the results of GP using conceptual hydrological models as physical representation. First the possibilities of GP are tested on synthetic data, which results in a shortlist of good working objective functions and insight in the most important GP settings. The test on real data in the Belgium Ardennes showed that GP using the objective functions KG10, MM and Shafii performed better. Nevertheless all three models performed not well on simulating the low flows and high peaks. Furthermore GP using KG10 and MM both results in simple serial models which perform well overall, but bad on quick response runoff. Shafii resulted in parallel models which show quick response flow, however GP it is not able to capture the fast responses correctly (yet). GP has the potential to improve the understanding in the behaviour of catchments, however it still needs the human mind to observe, compare and analyse the modelling results. The main consideration with GP is to look for a balance between: model search space, objective function, randomness and (computational) time. The challenge is how to lead GP in an efficient way without removing the possibility of finding unknown patterns.

We live today on a human-dominated planet under unprecedented pressure on both land and water. The water cycle is intrinsically linked to vegetation and land use, and anticipating the consequences of simultaneous changes in land and water systems requires a thorough understanding of their interactions. This thesis aims to advance our knowledge of how land-use change influences the water cycle, i.e., focussing on the role of land-use in mediating water’s journey from land evaporation, to atmospheric moisture, and to precipitation on land. This thesis first presents the development (Chapter 2) and evaluation (Chapter 3) of the process-based water balance model STEAM (Simple Terrestrial Evaporation to AtmosphereModel). STEAM simulates five different evaporation fluxes, based on land-use representation with only a limited number of parameters. Comparison with independent data shows that STEAM produces realistic evaporative partitioning and hydrological fluxes over different locations, seasons and land-use types. Chapter 4 investigates the temporal characteristics of partitioned evaporation, and shows that terrestrial residence timescale of transpiration (days to months) is substantially longer than that of interception (hours). The vegetation’s ability to transpire by retaining and accessing soil moisture at great depth is critical for dry season evaporation, and the substantial differences in temporal characteristics between evaporation fluxes can create contrasting moisture recycling patterns. In response to the importance of root zone storage capacity for transpiration and moisture recycling simulation, Chapter 5 sets out to present an ’earth observation-based’ method for estimating this critical parameter in land surface modelling. By assuming that vegetation does not root deeper than necessary to bridge critical dry periods, satellitebased evaporation were used to derive root zone storage capacity. The new estimate improved evaporation simulation overall, and in particular during the least evaporating months in sub-humid to humid regions with moderate to high seasonality. The results suggest that several forest types are able to create a large storage to buffer for severe droughts, in contrast to e.g., grasslands and croplands. Based on the new insights, Chapter 6 analyses the effects of land-use change on river flows. In some of the world’s largest basins, precipitation was found to bemore influenced by extra-basin, than within-basin, land-use change. In fact, in several non-transboundary basins, river flows were considerably influenced by land-use changes in foreign countries, suggesting new transboundary water relationships in international politics. This thesis addressed different domains in the water cycle to improve our understanding of land-water interactions. Every water flux and stock requires our examination, whether they flow visibly in rivers, travel invisibly in the air, or hide deep in soil and roots. Because of the terrestrial water cycle’s interaction with land, and therefore human activities, we are in an extraordinary position to shape its path and pace.@en

Contemplating the Meuse or any other river of the world, one may wonder about the journey of rain in becoming river. This fascinates hydrologists, as they develop theories to understand movement, storage and release of water through the landscape across climates. These theories are translated to hydrological models, which describe the complex reality in a simpler way. Models are then used to predict the hydrological cycle for the nearby or long-term future. This thesis aims to assist the Dutch Ministry of Infrastructure and Water Management in improving the reliability of hydrological modeling of the Meuse basin for operational and policy applications. Using in-situ and remote-sensing data, the value of representing additional processes in models is explored, as well as the creative use of additional data to improve hydrological predictions. First, water balance data is used to identify the potential presence of intercatchment groundwater flows (Chapter 3). These underground flow paths cross topographic catchment boundaries and mainly play a role in headwater catchments (< 500 km2) of the Meuse basin, which are underlain by productive aquifers. Representing this flux as a preferential threshold-initiated process improves low and high flow model performance and increases the consistency between modeled and remote-sensing estimates of actual evaporation. Besides the importance of quantifying the long-term hydrological partitioning of precipitation into streamflow, evaporation and potentially intercatchment groundwater flows, another key element of the hydrological response is the amount of water available in the root-zone of vegetation. The temporal dynamics of root-zone soil moisture control how much more water can be stored in the soil and how much water is available for transpiration. In Chapter 4, meaningful estimates of root-zone soil moisture are inferred from satellite observations of near-surface soil moisture, by establishing a link between the catchment-scale root-zone storage capacity and the Soil Water Index. Interestingly, hydrological models with different internal process representations of root-zone soil moisture, evaporation, snow and total storage at the catchment scale may lead to a similar aggregated streamflow response (Chapter 5). This discrepancy implies that models are not necessarily providing the right answers for the right reasons, as they cannot simultaneously be close to reality and different from each other. To circumvent the uncertainty of process representation, which is inherent to hydrological science, the use of multiple model structures is advocated for operational and policy applications. Nonetheless, testing the consistency between modeled hydrological behavior and independent remote-sensing data can foster model developments and lead to creating better models. Finally, we move beyond the use of historical in-situ and remote-sensing data to predict long-term hydrological behavior of the Meuse basin under projected global warming (Chapter 6). If environmental conditions change, it is likely to also assume ecosystem adaptation in response to climate change and a potential natural and/or anthropogenic shift in dominant species across the landscape. Non-stationarity in the representation of hydrological systems is introduced in a process-based model with three hydrological response units to account for the spatial variability of hydrological processes. More specifically, we adapt the root-zone storage capacity parameter using the information contained in the projected climate data. This is an important step forward in the great challenge of hydrological predictions under change. Despite data uncertainties and a lack of data at the required temporal and spatial resolutions, many possibilities are at hand with what is currently available to develop new theories, test and improve hydrological models. Requiring creativity, this is a beautiful challenge to further unravel the mysteries of the hydrological landscape.@en

Through precipitation retention and evaporation (by both interception and transpiration), woodlands play a significant role in the global moisture cycle. Evaporation is the largest, but at the same time, the most difficult flux to observe in a woodland. Accounting for woodland evaporation is important for hydrological modelling for the efficient development and management of water resources. Assessing evaporation is a challenging undertaking that involves the use of a wide range of equipment and requires skilled personnel. Much work has been conducted on assessing evaporation in agricultural crops. Even satellite data-based models are largely structured to assess evaporation in agricultural crops to the exclusion of understanding evaporation dynamics in natural woodlands, especially in African ecosystems. However, evaporation in woodland surfaces accounts for a significant portion of the water cycle over the terrestrial land mass. Understanding the characteristics of woodland ecosystem evaporation like interception and transpiration, is key to monitoring climate impact on woodland ecosystems, which is important for hydrological modelling and the management of water resources at various scales. One of the key aspects to enable this understanding is the knowledge of woodland phenological interaction with climate variables and the seasonal environmental regimes. “Vegetation phenology” refers to the periodic biological life cycle events of plants, such as leaf flushing and senescence, and corresponding temporal changes in vegetation canopy cover. Solar radiation, temperature and water availability (i.e., rainfall and soil moisture) are some of the key environmental variables that influence plant phenology. The attributes of woodland phenology, solar radiation, temperature and water availability differ across the diverse ecosystems globally, therefore, requires better understanding at a more local or regional level. Yet, evaporation of natural woodlands, especially in African ecosystems, with respect to phenological phases, are poorly characterised. This is largely because phenological studies have mainly focused on northern mid-latitude regions to the exclusion of other regions like the miombo of southern Africa. For increasing the predictive power of hydrological models, it is important to account for the interaction of woodland phenology with climate variables over the seasons and to characterise evaporation. This thesis aims at understanding the miombo woodland evaporation as a consequence of the vegetation phenological interaction with environmental and hydrological variables across seasons. Based on information in public domain, this study is the first independent field observation data-based characterisation of actual evaporation of the miombo woodland. The miombo is a heterogeneous woodland of the genus Brachystegia with the dominant species in the study location being Bauhinia petersenia, Brachystegia longifolia, Brachystegia boehmii, Brachystegia speciformis, Jubenerdia paninculata, Pericopsis angolensis, Uapaca kirkiana and Uapaca sansibarica. Unique phenological attributes are the simultaneous leaf fall, leaf flush and leaf colour changes that normally occur in the dry season between May and October. Most miombo woodland species are broad leaved and have developed dry season coping mechanisms such as deep rooting (capacity to access deep soil moisture and ground water) and vegetation water storage. The canopy closure is varied across the miombo woodland strata and is influenced by several factors including rainfall, soil type, soil moisture and nutrients, species diversity and temperature. These phenological attributes are species dependent, with varied response to phenological stimuli. This study sought to answer the question on the role of the phenology of the miombo woodland in the evaporation dynamics. The thesis also endeavoured to show how phenology, potentially, affects satellite-based evaporation estimates of the miombo woodland. The Luangwa Basin in southern Africa, a largely miombo woodland covered basin, was used as the study area. This basin was chosen because it is located in both the dry miombo woodland and wet miombo woodland in the Zambezian miombo woodland which is the largest strata of the miombo woodland. Furthermore, the Luangwa Basin is located in Zambia which is described as the country possibly with the highest diversity of trees and is said to be the centre of endemism for Brachystegia, with 17 species. To answer the questions on the importance of the phenology of the miombo woodland on the evaporation dynamics, the study used a coupled approach by applying both satellite data and field observations. Phenological changes of the miombo woodland across seasons were assessed using satellite-based data, the normalised difference vegetation index (NDVI) and leaf area index (LAI). Satellite-based data, land surface temperature (LST) and normalised difference infrared index (NDII), were used as proxies for climate variables canopy temperature and canopy vegetation water content. Point scale field estimates of evaporation across three different phenophases of the miombo woodland were obtained using the Bowen ratio distributed temperature sensing (BR-DTS) system. By measuring profiles of air temperature and wet bulb temperature, the evaporation could be estimated via the Bowen ratio method (BR-DTS). Six satellite-based evaporation estimates were compared across different phenophases of the miombo woodland. This was meant to observe the phenophases in which significant diferences in the trend and magnitude of satellite-based evaporation estimates occured. The general water balance approach was used to assess annual actual evaporation at basin scale. Consequently, satellite-based evaporation estimates were compared to the BR-DTS-based evaporation estimates at point scale and the water balance-based evaporation at basin scale. Results, based on satellite data, show that the phenology of the miombo woodland, i.e., changes in woodland canopy cover and photosynthetic activities, have a season-dependent correlation with climate variables. Woodland canopy cover, across phenophases and seasons, appear to be more influenced rather by water than temperature. This may explain the particular species-dependent buffering mechanisms during water limited conditions i.e., leaf shedding, deep rooting systems with access to ground water, and the vegetation water storage mechanisms. In agreement with available literature in public domain it appears there is little variation in canopy cover/closure (i.e., proxied by LAI) in wet miombo woodland in the dry season. At the wet miombo woodland site in Mpika, Zambia, the BR-DTS observations showed that, across the different phenophases, the actual evaporation trend and magnitude appeared to be more associated with the available energy than the changes in the woodland canopy cover. Further analysis showed that the net radiation has a greater influence on actual evaporation as it accounted for more variations in the actual evaporation compared to the changes in the woodland canopy cover (i.e., NDVI). The energy partitioning showed that available energy expenditure varied with phenological season. In the green down phenophase during the cool dry season the available energy was largely partitioned as sensible heat flux. As the temperature and net radiation begun to increase in the early dormant phenophase during the late cool dry season (July August) the available energy appeared to be equally partitioned between sensible and latent heat flux. In the late dormant phenophase during the early warm pre-rainy season (i.e., September) available energy was largely partitioned as latent heat flux. In the green-up phenophase during the late pre-warm rainy season (i.e., October) and early rainy season (i.e., November to December) the avialable energy was largely partitioned as latent heat flux. During the rain days the available energy appeard to be equally partition between latent and sensible heat flux. It appears that as the net radiation and canopy cover increased the available energy was largely partitioned as latent heat flux during the dry season. A remarkable observation was the continued rising trend of actual evaporation even during the lowest woodland canopy cover period in August and September. The rising trend in actual evaporation during the dry season may be due to the developed dry season water stress buffering mechanism such as deep rooting with access to moisture in deep soils and possibly access to ground water. The trend of the BR-DTS-based actual evaporation of the miombo woodland in the dry season points to the interaction between hydro-climate variables (i.e., precipitation linked soil moisture and net radiation) and the plant phenology. When compared to field observations, at point scale, all satellite-based evaporation estimates underestimated actual evaporation of a wet miombo woodland in the dry season and part of the early rainy season. Substantial underestimations were in the dormant and the green-up phenophases. Additionally, except for the WaPOR, the trends of all other satellite-based evaporation estimates differed from that of field observations. Plausible explanations for the behaviour (trend and magnitude) of satellite-based evaporation estimates in the dry season include the non-integration of soil moisture directly into the modelling of transpiration and the optimisation of the rooting depth. For instance, the use of proxies such as the NDVI and LST for soil moisture in surface energy balance models, such as SSEBop, results in uncertainities as the proxies are unable to take into account other factors that influence the sensible heat flux. In MOD16 the use of relative humidity and vapour pressure difference as proxies for soil moisture may be a source of uncertainty in estimating transpiration. On the other hand it has been observed that direct integration of soil moisture in the MOD16 algorithm appeared to improve the accuracy of actual evaporation estimates. This may explain why the WaPOR which integrate soil moisture stress in the algorithm appeared to have a smilar trend to field observations and also had higher estimates of actual evaporation compared to the other satellite-based evaporation estimates. It has also been shown that optimising the rooting depth improves the accuracy of transpiration estimates in vegetation with a dry season. Most miombo woodland species are deep rooting with access to deep soil moisture and potentially groundwater. Therefore, direct integration of soil moisture into the algorithms for the satellite-based evaporation estimates and optimising the rooting depth is likely to improve the accuracy of actual evaporation estimates for the miombo woodland. The phenophase-based comparison at pixel scale in dry miombo woodland and wet miombo woodland and at the Luangwa Basin miombo woodland scale showed similar results. In all three scenarios substantially high coefficients of variation in actual evaporation estimates among satellite-based evaporation estimates were observed in the water limited, high temperature and low woodland canopy cover conditions in the dormant phenophase. The coefficients of variation in actual evaporation estimates were also substantially high in the green-up phenophase at the boundary between the dry season and the rainy season. The lowest coefficients of variation in actual evaporation estimates were observed in water abundant, high temperature, high leaf chlorophyll content and high woodland canopy cover during the maturity/peak phenophase. The high coefficients of variation in actual evaporation estimates, among satellite-based evaporation estimates, in the dormant and green-up phenophases, points to the challenge of estimating the actual evaporation of the miombo woodland in the dry season and early rainy season. The same scenario emerged as was observed at point scale, with reference to field observations, in which satellite-based evaporation estimates which directly integrate soil moisture in their algorithm appeared to have higher estimates of actual evaporation in the dormant phenophase in the dry season. For instance, the FLEX-Topo and WaPOR integrate soil moisture in their algorithms. Compared to each other the FLEX-Topo and WaPOR appeared to have no statistically significant (p-value > 0.5) differences in their trends and mean estimates of actual evaporation in the dormant phenophase in the dry season. Compared to the FLEX-Topo and WaPOR the other four satellite-based evaporation estimates, GLEAM, MOD16, SSEBop and TerraClimate showed statisticantly significant (p-value < 0.05) differences in the trend and mean estimates of actual evaporation in the dormant phenophase in the dry season. Considering the canopy phenology and the associated physiological adaptation of the miombo woodland plants in the dry season, it appears that the direct integration of the soil moisture in the algorithms and optimising the rooting depth is likely to improve the accuracy of the satellite-based evaporation estimates. In the maturity/peak phenophase(s) during the mid-rainy season, compared to other satellite-based evaporation estimates, the MOD16 appeared to have significantly (p-value < 0.05) higher estimates of actual evaporation. The plausible explanation for this observation could be that the interception module of MOD16 is more responsive to the miombo woodland phenology. The wet miombo woodland intercepts between 17-20 percent of rainfall annually. Compared to the general annual water balance-based actual evaporation all six satellite-based evaporation estimates underestimated actual evaporation of the Luangwa Basin. The implication of this observation is that satellite-based evaporation estimates likely underestimates evaporation even in non-miombo woodland such as the mopane woodland that are also part of the larger Luangwa Basin vegetation landscape. However, for a comprehensive overview of the performance of the satellite-based evaporation estimates there is need for vegetation type and land-cover type based assessments of actual evaporation for the Luangwa Basin. At both point and basin scale-based assessments, there was a negative linear relationship between the spatial resolution of satellite-based evaporation estimates and the estimated actual evaporation. Satellite-based evaporation estimates with fine spatial resolutions showed lower underestimates compared to those with coarser resolutions. The implication is that the finer the spatial resolution the lower the underestimation. However, at both assessment scales, the linear relationships between the spatial resolutions and the evaporation estimates were statistically insignificant (i.e., p-value > 0.05). The reason for this outcome is exhibited in that some satellite-based evaporation estimates with relatively coarser spatial resolutions, i.e., SSEBop at both point and basin scale and TerraClimate at basin scale, underestimated less compared to MOD16 which had a finer spatial resolution. Furthermore, at basin scale a coarser spatial resolution estimate FLEX-Topo and a finer spatial resolution estimate WaPOR showed similar magnitude of actual evaporation in the dormant phenophase in the dry season. The implication of this observation is that other factors (i.e., heterogeneity in the landscape, model structure, processes and inputs) influence more the estimated actual evaporation rather than the spatial resolutions of the satellite-based evaporation estimates. Consequently, it appears that satellite-based estimates at finer spatial resolution with the structure, processes and inputs that couple canopy transpiration with the root zone storage, taking into account the vertical upward (beyond 2.5 m) and horizontal moisture flux as well as the canopy phenological changes, are likely to provide actual evaporation estimates that reflect actual conditions of the miombo woodland. This is demonstrated by the WaPOR estimates which appears to include these aspects in simulating actual evaporation. The field-based actual evaporation assessments were conducted in the wet miombo woodland. It is possible that the phenological response to changes in hydrological and climate regimes in the drier miombo woodland are different from the observations at the Mpika site. Therefore, there is need for similar observations to be performed in the drier miombo woodland and to compare the results. However, this thesis has demonstrated the importance of understanding and incorporating the canopy phenology and dry season physiological adaptation (i.e., deep rooting) of the miombo woodland in modelling actual evaporation. Additionally, for basins with heterogenous woodland types like the Luangwa, it is important to conduct actual evaporation assessments in the different vegetation types. This is likely to give a more representative understanding of basin scale evaporation dynamics. Nevertheless, this study has provided a foundation on which other studies can build towards a more comprehensive understanding of the actual evaporation dynamics in this unique woodland. @en

The unavailability of consistent accurate river flow data is a significant impediment to understanding water resources availability, and hydrological extremes. This is particularly true for remote, difficult to access, morphologically active and therefore rapidly changing rivers. The state of global river discharge monitoring with respect to water infrastructure and frequency of data collection has been on the decline over the past few decades. This is despite the significant importance of these data for river flow predictions. Fortunately, rapid advancements in technologies open up possibilities for water resource authorities to increase their ability to accurately, safely and efficiently establish river flow observation through remote and non-intrusive observation methods. Low-cost Unmanned Aerial Vehicles (UAVs) in combination with Global Navigation Satellite Systems (GNSS) can be used to collect geometrical information of the riverbed and floodplain. Such information, in combination with hydraulic modelling tools, can be used to establish physically based relationships between river flows and permanent proxies. This study attempts to monitor flow in volatile, dangerous and difficult to access rivers using only affordable and easy to maintain new technologies. This thesis consists of three main components: i) generating a workable framework for monitoring rivers using low-cost technologies; ii) establishment of river geometry using a combination of airborne photogrammetry and low-cost GNSS equipment iii) and physically based rating curve development through hydraulic modelling of surveyed river sections. The first three chapters of this thesis provide an introduction in the form of a literature review, justification for the study and a description of the study area. In chapter 4, a framework is developed through an intensive review of traditional river monitoring processes. Uniquely effective and low-cost individual components are selected and placed within a framework. The ideal outcome is an interconnected framework which clearly presents the steps which are necessary for river monitoring in remote locations. The manner in which each critical step is related to the other is explained. Furthermore, the method by which modern technologies are assimilated into the method is described. Within the framework, critical thresholds are set up in order to signal the to the water manager whether the proposed model in its current state continues to perform as required. Chapter 5 investigates how low-cost technologies such as UAVs in combination with low-cost GNSS devices can be used to generate river geometry for the purposes of application in a hydraulic model. Furthermore, performance of the open-source photogrammetry software substantiated the claim that, free and open-source available packages are capable of producing results which are as good as proprietary alternatives as shown by the RMSE analyses. A novel approach to generate a seamless bathymetry through merging and volumization was successfully tested. Results presented in this chapter encourage future studies to investigate the impact of variations in the number of Ground Control Points (GCPs) on discharge estimations in a hydraulic model with different hydrodynamic boundary conditions. This follow up was instituted in Chapter 6. In this sixth chapter we accept that uncertainties in the data acquisition may propagate into uncertainties in the relationships found between discharge and state variables. This uncertainty prompts the need to understand the impact of varying geometries on hydraulic models. Specific attention is placed on variations caused by differing GCP numbers since the task of GCP placement is time consuming, potential dangerous and resource intensive in certain location and instances. We are successfully able to determine the minimum number of control points required to reproduce geometry. Overall, we successfully develop and test a workable method for water resources authorities to estimate river flows accurately through the application of advanced, low-cost technologies with minimal contact with measured variables. The development and application of low-cost technologies for river flow monitoring has led to the following important conclusions: • For the purpose of flow estimation, there is no need to use more than seven GCPs to establish accurate UAV-based geometry. Rather, it is more crucial to distribute the available markers to be maximally representative of the terrain elevations. Furthermore, it may be necessary to place more markers in close proximity to locations where one may expect the largest challenge for photogrammetry software (e.g.: water, thick forest/vegetation) • In order to limit the impact of the “doming” effect on terrain geometry measurements, one of the most effective, yet easily implementable mechanisms is to measure a river line using Real Time Kinematic (RTK) Global Navigation Satellite Systems (GNSS) equipment. This data can then be used to correct the terrain post photogrammetry processing. @en

Forest evaporation (Ei) is considered the main source of water vapor at a continental scale. Its quantification has been carried out in many ecosystems worldwide, applying the classical partitioning method to differentiate among sources of water vapor. This partitioning differentiates between transpiration (Et), soil evaporation (Es), water intercepted by plant and ground surfaces (Ei), and open water evaporation (Es) in flooded forests and mangroves. The partitioning of evaporation has been carried out by applying different methodologies such as eddy-covariance, conventional micro-meteorological measurements, stable water isotopes, and the combination of some of these methodologies. However, the classical partitioning approach can have large uncertainties in specific forest ecosystems as a consequence of the canopy structure. Instead, including canopy structure into the evaporation partitioning allowed us to better understand this flux. Forest canopy structure is difficult to assess and is determined by latitude, altitude, water availability, and growing stage of the forest. However, using the canopy layering (overstory, understory, ground layer, and forest floor layer) we can assess the contribution from the structural point of view. Forest succession is one factor affecting the classical partitioning in Tropical Deciduous Broadleaf Forest. Using cumulative daily collectors in three different stages of Tropical Dry Forest in Costa Rica, we were able to depict how the increment in forest complexity affects the interception of precipitation. Also, the Plant Area Index was the only structural parameter significantly correlated with the estimates of both, interception and effective precipitation. The capacity of the other parameters (e.g., tree densities, tree heights, number of species) was not enough to describe the effect of a growing forest on the interception of precipitation. Tropical forests with less water stress during the dry season allocate more biomass to their canopies. This increases the forest complexity in terms of the number of species, canopy height, and plant types. Tropical Evergreen Broadleaf forests have a more complex canopy structure than the Deciduous ones. The tropical wet forest in Costa Rica has a canopy of 45 m height and a large number of plant species including trees, lianas, palms, and bushes that provide a completely different canopy structure than mono-specific forests. Here, we were able to define three canopy layers according to canopy height (overstory, lower and upper understory) and monitor the evaporation process during one dry season. Applying conventional micro-meteorological measurements we were able to determine that the lower and upper understory layers contributed 9 % and 15 % of the evaporation, respectively. Meanwhile, the use of water stable isotopes did not allow us to determine the contribution of transpiration using the keeling plot method. However, the signatures of the stable water isotopes allowed us to determine that the source of water used by the plants depends on its type (liana, tree, palm, or bush). Also, we quantified the evaporation during precipitation events as one-third of the amount measured during dry sunny days. The proportion did not change during rain events per canopy layer. This water vapor was produced by the "splash droplet evaporation" process, that together with the energy convection and low air temperature produced the visible vapor plumes. We were able to identify the conditions during which the visible vapor plumes can be spotted. These conditions are the presence of precipitation, air convection, and a lifting condensation level at the top of the canopy with values lower than 100 m. Plants growing in arid environments developed strategies that help them to cope with the scarcity of water. Usually, these plants grow lumped in patches and the introduction of tree species to fight desertification changed the landscape introducing a forest-like land cover. In a Temperate Shurbland in China, we evaluated the effect of Willow trees (Salix matsudana) and Willow bushes (Salix psammophila) on the soil water after summer. Using stable water isotopes we identified the redistribution of groundwater beneath the plants through the hydraulic lift process. Mono-specific forest ecosystems such as the Temperate Evergreen Needleleaf Forest may modify the micro-meteorological conditions beneath their canopies. In Speulderbos, we monitor the evaporation process through eddy-covariance and stable water isotope techniques in a Douglas-Fir (Pseudotsuga menziesii) stand. Also, the evaporation process in the forest floor layer was analyzed in detail under laboratory conditions. Different forest floor layers evaporates up to 1.5 mm d-1, differing from field conditions, where the evaporation from these layers do not exceed the 0.2 mm d-1. This evaporation, represents only the 5.5 % of the total measured during the monitoring period. However, there is no evidence that the forest floor evaporation move upwards to contribute to the total evaporation measured above the overstory. This was confirmed by the eddy-covariance footprint and stable water isotopes signatures of the air measured continuously on the forest. Finally, the partitioning of evaporation based on canopy structure is suitable for complex ecosystems with a large number of species and a multilayered canopy. This leaves the classical partitioning for more homogeneous ecosystems where it can be carried out with a smaller monitoring investment.@en

Forests cover a large part of the globe, and are responsible for a large amount of evaporation and the fixation of carbon. To be able to better understand this atmospheric exchange of forests, and how the forests will behave under future climate change, both accurate measurements as well as models are required. However, due to their height and heterogeneity they are difficult to model and measure. Standard theories do not apply well to forests, and as such more effort is required to understand the exchange between the forests and the atmosphere. However, precise measurements are made difficult due to a number of issues. The most prominent are the non-closure of the energy balance, and so-called ‘decoupling’ of the canopy. Non-closure of the energy balance is where all the measured inflows and outflows of energy do not add up to the measured change in energy storage in the forest system. The size and heterogeneity of forests makes this difficult to assess. Second is ‘decoupling’, where the vertical mixing of air within the canopy is hampered, and measurements performed above the canopy are not representative of what happens in the entire canopy down to the forest floor.@en

Throughout the world, many people have been affected by water related issues in the past, some more extreme than others. In this context, hydrological models have often been used to gain more insight into the situation and to limit negative impacts as much as possible. There are many different types of hydrological models with each their strengths and weaknesses, but all models need a certain amount of reliable data. However, many river basins throughout the globe are poorly gauged which means there are only limited reliable ground observations available. That is why satellite observations provide many interesting opportunities to fill this gap of which many are not yet explored. Therefore the goal of this research was to answer the following main research question: What is the added value of satellite-based observations for hydrological modelling in a semi-arid, data-scarce river basin? This research focused on the Luangwa River in Zambia which is a large tributary of the Zambezi River. This semi-arid river basin is poorly gauged, mostly unregulated and sparsely populated. A process-based distributed hydrological model with sub-grid heterogeneity was developed in this research and modified step-wise when exploring the added value of different satellite observations for different aspects within hydrological modelling. This included analysing the information content of satellite-based river water level, i.e. altimetry, for model calibration, as well as evaporation and total water storage for step-wise model structure improvement and spatial-temporal model calibration. Overall, satellite-based observations have been used successfully to improve our understanding of the hydrological processes in the data-scarce Luangwa river basin, to improve the hydrological model structure and to allow for more reliable parameter identifications in the absence of reliable discharge data. This research focused on a selection of satellite-based observations and hydrological model applications. In other words, there remain many opportunities yet to be explored!@en