Concentrações de cádmio (Cd), chumbo (Pb) e mercúrio (Hg) foram determinadas em amostras de água (teores dissolvidos e particulados), sedimento de fundo, peixes e bivalves, coletados em seis bacias hidrográficas do Estado de São Paulo: Alto Paranapanema, Peixe, Aguapeí, São José dos Dourados, Mogi-Guaçu e Piracicaba, as quais apresentam diferentes características quanto ao uso e ocupação da terra. Dentre estas bacias, as dos rios Piracicaba e Mogi-Guaçu apresentam maiores densidades demográficas e industrialização, o que gera, conseqüentemente, maior degradação da qualidade das águas, enfatizando a necessidade de estudos freqüentes nesses sistemas. As amostragens de água e sedimento foram feitas bimestralmente durante o ano de 2001. Os peixes e bivalves foram amostrados no mesmo ano em dois períodos: seca e cheia. As concentrações de Cd e Pb foram determinadas por espectrometria de absorção atômica com atomização eletrotérmica, e as de Hg por espectrometria de fluorescência atômica. Os teores de Cd e Pb na fração trocável do sedimento também foram determinados. Os resultados obtidos mostraram baixas concentrações dos metais Cd e Pb na coluna d’água, tanto nas formas dissolvidas quanto particuladas, no entanto, tais concentrações foram maiores nas bacias dos rios Piracicaba e Mogi-Guaçu. No sedimento de fundo e nos bivalves, as maiores concentrações de Pb foram observadas no rio Mogi-Guaçu, embora este elemento tenha ocorrido em pequenas quantidades na fração trocável do sedimento (<1%). O Cd ocorreu em maiores teores no sedimento e no material particulado do rio Piracicaba. Por outro lado, as maiores concentrações de Pb nos bivalves foram detectadas no rio Mogi-Guaçu. Na fração trocável do sedimento, o Cd ocorreu em altas proporções (até 76%), principalmente, nos rios Itapetininga (Bacia do Alto Paranapanema), Piracicaba e Mogi-Guaçu. Em algumas amostras de água e sedimento de fundo ocorreram quantidades relativamente elevadas de Hg. Por outro lado, nos bivalves os teores do elemento foram sempre baixos. No entanto, nenhuma tendência significativa de distribuição espacial ou temporal de Hg ocorreu nas amostras. Portanto, acredita-se que os altos teores registrados para algumas amostras foram resultantes de fontes de origem difusa não-pontual, como deposição atmosférica e fontes originárias da agricultura. Nos peixes, as concentrações de Cd e Pb foram sempre inferiores aos limites de detecção (0,005 e 0,08 µg/g peso seco, respectivamente). Por outro lado, peixes carnívoros coletados no reservatório de Jurumirim (bacia do Alto Paranapanema) apresentaram altas concentrações de Hg (média: 1,14 +/- 0,54 µg/g peso úmido) e metilmercúrio (média: 1,04 +/- 0,51 µg/g peso úmido). Este último, representou em média 92% do mercúrio total determinado no músculo dos peixes. Tais concentrações foram maiores que o máximo permitido pela legislação brasileira para peixes predadores (1 µg/g peso úmido) e, provavelmente, foram resultante de altos teores de metilmercúrio produzido no reservatório devido aos processos de produção e liberação de mercúrio a partir da vegetação e solos inundados. Portanto, concluiu-se que os peixes do reservatório de Jurumirim estavam contaminados com Hg, recomendando-se que novos estudos sejam realizados para se identificar as fontes e processos responsáveis por tais contaminações. As bacias dos rios do Peixe, Aguapeí, Alto Paranapanema e São José dos Dourados apresentaram baixas quantidades de todos os elementos estudados. Por outro lado, foram observadas concentrações relativamente elevadas de Cd no rio Piracicaba, e Pb e Cd no rio Mogi-Guaçu, os quais devem ser periodicamente avaliados. / Concentrations of cadmium (Cd), lead (Pb), and mercury (Hg) were determined in water (dissolved and particulate), bottom sediment, fish, and bivalves sampled in six watersheds located in the State of São Paulo (Southeast, Brazil): the Alto Paranapanema, the Peixe, the Aguapeí, the São José dos Dourados, the Mogi-Guaçu, and the Piracicaba. These watersheds show different levels of land-use. Among these basins, the Piracicaba and the Mogi-Guaçu River Basins show highest population and industrialization rates. Such features lead, consequently, to a higher degradation of water, thus decreasing its quality. These facts emphasize the need of new and frequent studies in those systems. Water and sediment samplings were carried out bimonthly during 2001. Fish and bivalves samples were collected twice a year: at the end of the rainy and dry season. Cd and Pb concentrations were determined through electrotermal atomic absorption spectrometry, whereas Hg through atomic fluorescence spectrometry. In addition, exchangeable Cd and Pb were determined in sediment. The results showed low concentrations of Cd and Pb in water (dissolved and particulate). However, these concentrations were higher in the Piracicaba and the Mogi-Guaçu River Basins. Pb concentrations in sediment and bivalves were the highest in the Mogi-Guaçu River, even so this element occurred in low levels in the exchangeable fraction of sediment (< 1%). The highest Cd concentrations in sediment and particulate suspended matter were related to the Piracicaba River, although the highest levels for bivalves occurred in the Mogi-Guaçu River. High Cd concentrations in the exchangeable fraction of sediment (up to 76%) were observed, mainly in the Itapetininga (the Alto Paranapanema Basin), the Piracicaba, and the Mogi-Guaçu River. High concentrations of Hg were detected in some water and sediment samples, whereas in bivalves Hg was always in low levels. However, no significant tendency of temporal and/or spatial distribution was observed for Hg, so it is belived that the high levels reported for some samples were a result from diffuse non-point source pollution, such as atmospheric deposition and residues from agricultural practices. Cd and Pb concentrations in fish were always lower than the detection limit (0.005 and 0.08 µg/g dry wt, respectively). On the other hand, high concentrations of Hg (1.14 +/- 0,54 µg/g wet wt) and methylmercury (1.04 +/- 0.51 µg/g wet wt), which represented 92% of the total mercury, were found in carnivorous fish samples from the Jurumirim Reservoir (the Alto Paranapanema Basin). These concentrations were higher than the maximum limit permitted by Brazillian legislation for predatory fish (1 µg/g wet wt) and, probably, were a result from high methylmercury levels produced and released in the reservoir supported by inundated soil and vegetation. Hence, we concluded that carnivorous fish from the Jurumirim Reservoir were contamined with Hg, therefore new research sould be carried out in that area in order to identify the sources and processes responsible for high levels of mercury. The Peixe, the Aguapeí, the Alto Paranapanema, and the São José dos Dourados River Basins showed relatively low concentrations of all elements studied. Conversely, high Cd concentrations were determined in the Piracicaba River, and Pb and Cd in the Mogi-Guaçu River, pointing out that these areas should be frequently monitored.

Event-based sediment budgets were developed in a small agricultural sub-watershed using radionuclide tracers in conjunction with traditional monitoring techniques. The result of these budgets quantified the flux of material from each sediment source in the sub-watershed. The first step in the study was to quantify the net flux of material through the watershed outlet for runoff events. Results indicated that a pronounced clockwise hysteresis effect occurred during all studied events. The cause of the hysteresis effect was attributed to exhaustion of the upland (i.e., hillslopes and floodplains) sediment source. The hysteresis effect was dampened during an extreme flash flood event that caused overbank flow. This dampening was attributed to an increase in upland sediment mobilization resulting from overbank flow. Results of the event-based monitoring were compared against a previously developed sediment rating curve that assumed a power-law relationship between suspended sediment transport and water flow rates. The results indicate that the power-law relationship grossly under predicted the sediment flux over each runoff event. A tracing technique was utilized to establish the relative contributions from the uplands and the stream channel (i.e., channel banks and bed). This technique used the relationship between the naturally occurring radionuclide tracers 7Be and 210Pbxs to differentiate eroded upland soils from channel-derived sediments in the suspended sediment loads. A simple two end-member unmixing model was used to determine the relative contribution from each source. Results indicate that the upland source was the dominant contributor to the suspended load early in the runoff events, but channel contributions were more prevalent at later stages, reinforcing the conclusion drawn from the hysteresis observation. Further analysis of the results indicated that the uplands contributed the majority of the material to the suspended load because the sediment transport rate during the early stages of the event was much larger than during the later stages. Therefore, watershed managers wishing to minimize non-point source pollution resulting from erosion should first focus their efforts on reducing erosion of upland soils.

Fluvial sediment transport is a process that has long been important in managing water resources. While we intuitively recognize that increased flow amounts to increased sediment discharge, there is still significant uncertainty in the details. Because sediment transport---and in the context of this dissertation, bed load transport---is a strongly nonlinear process that is usually modeled using empirical or semi-empirical equations, there exists a large amount of uncertainty around model parameters, predictions, and model suitability. The focus of this dissertation is to develop and demonstrate a series of physically- and statistically-based sediment transport models that build on the scientific knowledge of the physics of sediment transport while evaluating the phenomenon in an environment that leads us to robust estimates of parametric, predictive, and model selection uncertainty. The success of these models permits us to put theoretically and procedurally sound uncertainty estimates to a process that is widely acknowledged to be variable and uncertain but has, to date, not developed robust statistical tools to quantify this uncertainty. This dissertation comprises four individual papers that methodically develop and prove the concept of Bayesian statistical sediment transport models. A simple pedagogical model is developed using synthetic and laboratory flume data---this model is then compared to traditional statistical approaches that are more familiar to the discipline. A single-fraction sediment transport model is developed on the Snake River to develop a probabilistic sediment budget whose results are compared to a sediment budget developed through an ad hoc uncertainty analysis. Lastly, a multi-fraction sediment transport model is developed in which multiple fractions of laboratory flume experiments are modeled and the results are compared to the standard theory that has been already published. The results of these models demonstrate that a Bayesian approach to sediment transport has much to offer the discipline as it is able to 1) accurately provide estimates of model parameters, 2) quantify parametric uncertainty of the models, 3) provide a means to evaluate relative model fit between different deterministic equations, 4) provide predictive uncertainty of sediment transport, 5) propagate uncertainty from the root causes into secondary and tertiary dependent functions, and 6) provide a means by which testing of established theory can be performed.

No description available.

Semi-arid environments are mainly characterized by scarce water resources and are usually subject to risks of water stress. In these regions, water supply for drinking and irrigation purposes depends strongly on storage in surface reservoirs and sediment deposition in these reservoirs affects adversely the water storage. In order to reproduce the complex behaviour of sediment deposition in reservoirs located in semi-arid environments and the effects of using sediment management techniques, a reservoir sedimentation model is developed and coupled within the WASA-SED model, which simulates rainfall-runoff processes and sediment transport at the hillslope and river network. The reservoir sedimentation model consists of two modelling approaches, which may be applied according to reservoir size and data availability. For reservoirs with information about their geometric features (reservoir topography, stage-area and stage-volume curves) and physical properties of sediment deposits, such as deposition thickness, grain size distribution of sediment deposits and sediment densities, a detailed modelling approach of reservoir sedimentation may be applied. For reservoirs without those characteristics, a simplified modelling approach is used. The detailed modelling approach of reservoir sedimentation enables the assessment of sediment deposition pattern in reservoirs and the evaluation of sediment release efficiency of sediment management techniques. It simulates sediment transport along the longitudinal profile of a reservoir. The reservoir is divided into cross sections to elaborate the sediment budget. The sediment transport component is calculated using a non-uniform sediment transport approach based on the concept of sediment carrying capacity. Four different sediment-transport equations can be selected for the simulations. The simplified modelling approach of reservoir sedimentation is suitable to simulate water and sediment transfer in dense reservoirs network. Nevertheless, it allows simulating neither sediment management techniques, nor spatial distribution of sedimentation. In this approach, the reservoirs are classified into small and strategic reservoirs according to their location and size. Strategic reservoirs are medium and large-sized reservoirs located on main rivers at the sub-basin’s outlet or reservoirs of particular interest. The small reservoirs are located at tributary streams and represented in the model in an aggregate manner by grouping them into size classes according to their storage capacity. A cascade routing scheme is used to describe the upstream-downstream position of the reservoir classes. The water and sediment balances of small reservoirs are computed for one hypothetical representative reservoir of mean characteristics. Sediment trapping efficiency and effluent grain size distribution are estimated using the overflow rate concept. Three model applications are carried out within this research, as follows: • The detailed modelling approach of reservoir sedimentation is applied to the 92.2 Mm³ Barasona Reservoir, located in the foothills of the Central Pyrenees (Aragon, Spain). A two-stage calibration was performed to account for changes on the sediment deposition pattern caused by sediment management. The reservoir sedimentation model is then validated for another simulation period which confirms that the processes related to reservoir sedimentation are well represented by the model. • An application is carried out to the 933-km² Benguê catchment, located in the semi-arid region of Northeast Brazil. The catchment is characterized by a dense reservoir network, covering almost 45% of the catchment area, with a significant lack of data. Water and sediment balances of those reservoirs are computed using the simplified modelling approach. Three spatial configurations describing the cascade routing scheme are tested. • The reservoir sedimentation model is applied again to the Barasona reservoir to evaluate the sediment release efficiency of sediment management strategies. Cost analysis is presented to help in the choice of the most promising sediment management technique for that situation. Thus, the model enables the assessment of technical features of the sediment management strategies. Overall, simulation results are characterized by large uncertainties, partly due to low data availability and also due to uncertainties of the model structure to adequately represent the processes related to reservoir sedimentation. / Semiaride Gebiete sind hauptsächlich durch geringe Wasserressourcen gekennzeichnet und unterliegen häufig dem Risiko der Wasserknappheit. In diesen Gebieten ist die Wasserbereitstellung für Bewässerung und Trinkwasserversorgung stark von der oberflächlichen Speicherung in Stauseen abhängig, deren Wasserverfügbarkeit nachteilig durch Sedimentablagerung beeinflusst wird. Zur Wiedergabe des komplexen Sedimentablagerungsverhaltens in Stauseen von semiariden Gebieten und die Auswirkungen von Sedimentmanagementmaßnahmen wird ein Sedimentationsmodell entwickelt und mit dem WASA-SED Modell gekoppelt, das für die Modellierung der Abflussbildung und des Sedimenttransportes in Einzugsgebieten geeignet ist. Das Sedimentationsmodell beinhaltet zwei Ansätze, die unter der Berücksichtigung verschiedener Stauseengrößenklassen und Datenverfügbarkeit eingesetzt werden können. Für die Stauseen mit verfügbaren Informationen über ihre geometrischen Eigenschaften (wie Stauseetopographie und Höhe-Fläche-Volumen-Beziehung) und weitere Kenngrößen wie Ablagerungsmächtigkeit, Korngrößenverteilung und Sedimentdichte, kann ein detaillierter Modellansatz für die Sedimentablagerung verwendet werden. Wo diese Informationen nicht verfügbar sind, wird auf einen vereinfachten Ansatz zurückgegriffen. Der detaillierte Modellansatz ermöglicht die Betrachtung von Ablagerungsmustern im Stausee und Einschätzungen über die Effektivität von Sedimentmanagementmaßnahmen hinsichtlich der Sedimententlastung. Dieser Ansatz beruht auf der Simulation des Sedimenttransportes entlang eines Stauseelängsprofils. Für die Berechnung des Sedimenttransfers wird der Stauseekörper in einer Folge von Querprofilen repräsentiert. Der Sedimenttransport wird dabei korngrößenspezifisch entsprechend der Transportkapazität berechnet. Dafür stehen vier verschiedenen Sedimenttransportgleichungen zur Verfügung. Der vereinfachte Modellansatz ist für die Simulation des Sedimenttransfers in Gebieten mit hoher Stauseedichte geeignet, jedoch können weder Sedimentmanagementmaßnahmen noch die räumliche Verteilung der Ablagerungen berücksichtigt werden. Dafür werden die Stauseen in Abhängigkeit von ihrer Größe und Position in kleine und strategische Stauseen unterteilt. Dabei sind strategische Stausseen solche mit mittlerem bis großem Volumen sowie einer Lage im Hauptgerinne oder solche mit sonstiger besonderer Bedeutung. Kleine Stauseen hingegen befinden sich an den Nebenflüssen und werden im Modell in aggregierter Form durch ihre Einteilung in Stauseegrößenklassen repräsentiert. Ein Kaskadenverfahren wird für den Wasser- und Sedimentlauf zwischen den Stauseeklassen verwendet. Dabei werden für jede Stauseeklasse der Wasser- sowie Sedimenthaushalt für einen hypothetischen repräsentativen Stausee mit mittleren Eigenschaften berechnet. Die Sedimentaufnahme und die Korngrößenverteilung des abgegebenen Sediments werden mit dem Überlaufanteil-Ansatz berechnet. In dieser Studie werden drei Modellanwendungen vorgestellt: • Für den 92,2 Mio.m³-großen Barasona-Stausee (Vorland der Zentralpyrenäen, Aragon, Spanien) wird die Modellierung der Sedimentablagerung mit dem detaillierten Modellansatz vorgenommen. Die Kalibrierung dafür wurde in zwei Schritten durchgeführt, um Änderungen im Stauseemanagement Rechnung zu tragen. Die ModellValidierung wird schließlich für eine andere Simulationsperiode vorgenommen. Dabei wird ersichtlich, dass die Prozesse der Sedimentablagerung gut durch das Modell wiedergegeben werden. • Das Modell wird auf das 933 km²-große Benguê-Einzugsgebiet, das sich im semiariden Nordosten Brasiliens befindet, angewendet. Dieses Einzugsgebiet ist durch eine hohe Dichte an kleinen Stauseen, charakterisiert, die fast 45% des Gebietes umfasst, wofür jedoch wenige Messdaten verfügbar sind. Deshalb werden der Wasser- und Sedimenttransport mit dem vereinfachten Modellansatz berechnet. Dabei werden drei Konfigurationen des Kaskadenverfahrens getestet. • Die Modellanwendung erfolgt erneut für den Barasona-Stausee bezüglich der Effektivität der Sedimentmanagementmaßnahmen. Eine Kostenanalyse ermöglicht die Auswahl geeigneter Maßnahmen für den Stausee. Dadurch wird eine Beurteilung der verschiedenen Sedimentmanagementstrategien ermöglicht. Im Allgemeinen unterliegen die Simulationsergebnisse großen Unsicherheiten, teilweise wegen der geringen Datenverfügbarkeit, andererseits durch die Unsicherheiten in der Modellstruktur zur korrekten Wiedergabe der Sedimentablagerungsprozesse.

Water shortage is a serious threat for many societies worldwide. In drylands, water management measures like the construction of reservoirs are affected by eroded sediments transported in the rivers. Thus, the capability of assessing water and sediment fluxes at the river basin scale is of vital importance to support management decisions and policy making. This subject was addressed by the DFG-funded SESAM-project (Sediment Export from large Semi-Arid catchments: Measurements and Modelling). As a part of this project, this thesis focuses on (1) the development and implementation of an erosion module for a meso-scale catchment model, (2) the development of upscaling and generalization methods for the parameterization of such model, (3) the execution of measurements to obtain data required for the modelling and (4) the application of the model to different study areas and its evaluation. The research was carried out in two meso-scale dryland catchments in NE-Spain: Ribera Salada (200 km²) and Isábena (450 km²). Adressing objective 1, WASA-SED, a spatially semi-distributed model for water and sediment transport at the meso-scale was developed. The model simulates runoff and erosion processes at the hillslope scale, transport processes of suspended and bedload fluxes in the river reaches, and retention and remobilisation processes of sediments in reservoirs. This thesis introduces the model concept, presents current model applications and discusses its capabilities and limitations. Modelling at larger scales faces the dilemma of describing relevant processes while maintaining a manageable demand for input data and computation time. WASA-SED addresses this challenge by employing an innovative catena-based upscaling approach: the landscape is represented by characteristic toposequences. For deriving these toposequences with regard to multiple attributes (eg. topography, soils, vegetation) the LUMP-algorithm (Landscape Unit Mapping Program) was developed and related to objective 2. It incorporates an algorithm to retrieve representative catenas and their attributes, based on a Digital Elevation Model and supplemental spatial data. These catenas are classified to provide the discretization for the WASA-SED model. For objective 3, water and sediment fluxes were monitored at the catchment outlet of the Isábena and some of its sub-catchments. For sediment yield estimation, the intermittent measurements of suspended sediment concentration (SSC) had to be interpolated. This thesis presents a comparison of traditional sediment rating curves (SRCs), generalized linear models (GLMs) and non-parametric regression using Random Forests (RF) and Quantile Regression Forests (QRF). The observed SSCs are highly variable and range over six orders of magnitude. For these data, traditional SRCs performed poorly, as did GLMs, despite including other relevant process variables (e.g. rainfall intensities, discharge characteristics). RF and QRF proved to be very robust and performed favourably for reproducing sediment dynamics. QRF additionally excels in providing estimates on the accuracy of the predictions. Subsequent analysis showed that most of the sediment was exported during intense storms of late summer. Later floods yielded successively less sediment. Comparing sediment generation to yield at the outlet suggested considerable storage effects within the river channel. Addressing objective 4, the WASA-SED model was parameterized for the two study areas in NE Spain and applied with different foci. For Ribera Salada, the uncalibrated model yielded reasonable results for runoff and sediment. It provided quantitative measures of the change in runoff and sediment yield for different land-uses. Additional land management scenarios were presented and compared to impacts caused by climate change projections. In contrast, the application for the Isábena focussed on exploring the full potential of the model's predictive capabilities. The calibrated model achieved an acceptable performance for the validation period in terms of water and sediment fluxes. The inadequate representation of the lower sub-catchments inflicted considerable reductions on model performance, while results for the headwater catchments showed good agreement despite stark contrasts in sediment yield. In summary, the application of WASA-SED to three catchments proved the model framework to be a practicable multi-scale approach. It successfully links the hillslope to the catchment scale and integrates the three components hillslope, river and reservoir in one model. Thus, it provides a feasible approach for tackling issues of water and sediment yield at the meso-scale. The crucial role of processes like transmission losses and sediment storage in the river has been identified. Further advances can be expected when the representation of connectivity of water and sediment fluxes (intra-hillslope, hillslope-river, intra-river) is refined and input data improves. / In vielen Regionen der Erde stellt Wassermangel ein Problem für die menschliche Gesellschaft dar. Insbesondere in Trockengebieten werden jedoch Maßnahmen des Wassermanagements, wie die Wasserspeicherung in Stauseen, durch die im Fluss transportierten Sedimentfrachten negativ beeinflusst. Somit stellen eine adäquate Beurteilung von Wasser- und Sedimentflüssen eine wichtige Voraussetzung für Entscheidungen in Wassermanagement und -planung dar. Dieser Problematik widmete sich das SESAM-Projekt (Sediment Export from large Semi-Arid catchments: Measurements and Modelling). Im Rahmen dieses Projektes befasste sich diese Dissertation mit (1) der Entwicklung und Umsetzung eines Erosions-Moduls für ein Einzugsgebietsmodell auf der Meso-Skala, (2) der Entwicklung von Skalierungs- und Generalisierungsmethoden für die Parametrisierung eines solchen Modells, (3) der Durchführung von Messungen, um die notwendigen Daten für das Modell zu gewinnen und (4) die Anwendung des Modells für verschiedene Einzugsgebiete und seiner Bewertung. Die Studie umfasste zwei mesoskalige Trockeneinzugsgebiete in NO-Spanien: Ribera Salada (200 km²) und Isábena (450 km²). Im Hinblick auf Zielstellung 1 wurde WASA-SED, ein räumlich semi-distribuiertes Modell für Wasserflüsse und Sedimenttransport, entwickelt. Das Modell simuliert Abfluss- und Erosionsprozesse auf der Hangskala, den Transport von suspendierten und Geschiebesedimenten auf der Skala von Flussabschnitten sowie Rückhalt- und Remobilisierungsprozesse von Sedimenten in Stauseen. Die vorliegende Arbeit stellt das Modellkonzept und Modellanwendungen vor und beschreibt Fähigkeiten und Grenzen des Modells. Die Modellierung auf größeren Skalen beinhaltet das Dilemma, dass relevante Prozesse beschrieben werden müssen, gleichzeitig aber die Anforderungen an Eingabedaten und Rechenzeit realisierbar bleiben. In WASA-SED wird diesem durch die Anwendung eines innovativen Hangprofil-basierten Skalierungsansatzes Rechnung getragen, indem die Landschaft durch charakteristische Toposequenzen repräsentiert wird. Um derartige Toposequenzen hinsichtlich verschiedener Landschaftseigenschaften (z.B. Relief, Böden, Vegetation) abzuleiten, wurde in Bezug zur Zielstellung 2 der LUMP-Algorithmus (Landscape Unit Mapping Program) entwickelt. LUMP beinhaltet ein Verfahren zur Berechnung repräsentativer Hangprofile und ihrer Attribute aus einem digitalen Geländemodell und optionalen Zusatzdaten. Durch die Klassifikation dieser Hangprofile wird die Grundlage der räumlichen Diskretisierung des WASA-SED Modells bereitgestellt. Im Zusammenhang mit Zielstellung 3 wurden Abfluss und Sedimentkonzentration (SSC) am Auslass und in einigen Teileinzugsgebieten des Isábena-Einzugsgebietes gemessen. Um den Sedimentaustrag zu bestimmen, mussten die Einzelmessung der Sedimentkonzentration interpoliert werden. Diese Arbeit vergleicht die Eignung traditioneller Eichkurvenansätze (SRCs), Generalized Linear Models (GLMs) und der nichtparametrischen Regressionstechniken Random Forests (RF) und Quantile Regression Forests (QRF). Da die beobachteten SSC-Werte stark über sechs Größenordnungen variieren, erwiesen sich die traditionellen SRCs als unzureichend. Gleichfalls versagten GLMs trotz der Einbeziehung weiterer relevanter Prozessgrößen wie Niederschlagsintensitäten und Abflusscharakteristika. RF und QRF stellten sich hingegen als sehr robust und für die Rekonstruktion der Sedimentdynamik geeignet dar. QRF liefert darüber hinaus auch Informationen zur Genauigkeit dieser Schätzungen. Die darauf aufbauende Analyse ergab, dass der Großteil der Sedimentfracht während der Starkregenereignisse des Spätsommers transportiert wurde. Spätere Niederschlagsereignisse erzeugten deutlich geringeren Austrag. Durch den Vergleich von Sedimentfrachten im Oberlauf mit Austragsmengen am Gebietsauslass konnte die Bedeutung der Sedimentspeicherung im Flussbett identifiziert werden. Zielstellung 4 wurde bearbeitet, indem das WASA-SED-Modell für zwei Untersuchungsgebiete in NO-Spanien unter unterschiedlichen Gesichtspunkten angewendet wurde. Für das Ribera-Salada-Einzugsgebiet lieferte das unkalibrierte Modell plausible Ergebnisse hinsichtlich der Wasser- und Sedimentflüsse. Damit war es möglich, die potentiellen Änderungen dieser Größen durch verschiedene Landnutzungsszenarien zu quantifizieren. Diese wurden den prognostizierten Veränderungen, die durch Klimaänderungen hervorgerufen würden, gegenübergestellt. Im Gegensatz dazu konzentrierte sich die Anwendung im Isábena-Einzugsgebiet auf die Untersuchung der bestmöglichen Modellanpassung. Im Validierungszeitraum ergab sich eine befriedigende Modellgüte für Wasser- und Sedimentflüsse. Diese Gesamtgüte wurde maßgeblich durch die unzureichende Abbildung der Unterliegergebiete beeinflusst, wohingegen die Gebiete des Oberlaufs, trotz ihrer stark kontrastierenden Sedimentausträge, gut dargestellt wurden. Die Anwendung des WASA-SED-Modells auf drei verschiedene Untersuchungsgebiete bestätigt die generelle Eignung des Modellkonzepts als einen sinnvollen multiskaligen Ansatz, der in einem Modell effektiv die Hangskala mit der Einzugsgebietsskala sowie den Einfluss von Flüssen und Stauseen vereint. Er stellt somit eine mögliche Grundlage für die Bearbeitung von wasser- und sedimentbezogenen Fragestellungen auf der Meso-Skala dar. Die besondere Bedeutung der Prozesse der Sickerverluste und Sedimentspeicherung im Gerinne konnten herausgearbeitet werden. Mögliche Verbesserungen betreffen die Berücksichtigung der Konnektivität von Wasser- und Sedimentflüssen (auf dem Hang, zwischen Hang und Fluss, innerhalb des Flusses) und die Qualität der Eingangsdaten für das Modell.

Coastal low oxygen areas are expanding globally and are predicted to increase in size and duration due to climatic changes associated with a warming ocean. The Gulf of Mexico Hypoxic Zone (GoMHZ) is the second largest regularly occurring hypoxic habitat in the world and has increased in size since it was first mapped in the 1980s. The Mississippi Atchafalaya River System (MARS) floods the Louisiana continental shelf with fresh water high in nitrogenous compounds enhancing primary production which sinks to the sea floor. Stratification that occurs as a result of density differences and coastal currents creates a strong pycnocline that prevents bottom waters from being aerated causing seasonally hypoxic bottom waters (< 2.0 mg L^-1). The Mechanisms Controlling Hypoxia (MCH) project (hypoxia.tamu.edu) made regular cruises during 2004-2005 and 2007-2009 to the GoMHZ performing shelf wide hydrographic surveys and occupying central mooring sites within theoretical zones of differing hypoxic potential. Sediment cores were collected for incubation experiments using Batch Microincubation Chambers (BMICs) to measure rates of sediment community oxygen consumption and nutrient regeneration. Results of incubation experiments characterized sediments as net sources of dissolved inorganic nitrogen, mostly ammonium, and silicate and a net sink of phosphate. Modeling simulations of benthic-pelagic coupling focused in the western study zones related field measurements of benthic nutrient regeneration and primary production to important processes that maintain summertime hypoxia when surface waters are nitrogen limited. After incubations were completed macrofaunal individuals were removed from the cores enumerated and identified to the lowest possible taxon. Macrofauna communities in 2004-2005 were dominated by a hypoxia tolerant community dominated by polychaetes. Hurricanes Katrina and Rita in August and September of 2005 drastically reorganized macrobenthic communities decreasing abundances and negatively impacting diversity. These new communities collapsed under hypoxic stresses potentially impacting the ability of demersal foragers to utilize an important food resource. Large variations in biogeochemical fluxes and patchy distribution of fauna impeded the delineation of significant zones in benthic function and structure.

Fluvial sediment transport is a process that has long been important in managing water resources. While we intuitively recognize that increased flow amounts to increased sediment discharge, there is still significant uncertainty in the details. Because sediment transport---and in the context of this dissertation, bed load transport---is a strongly nonlinear process that is usually modeled using empirical or semi-empirical equations, there exists a large amount of uncertainty around model parameters, predictions, and model suitability. The focus of this dissertation is to develop and demonstrate a series of physically- and statistically-based sediment transport models that build on the scientific knowledge of the physics of sediment transport while evaluating the phenomenon in an environment that leads us to robust estimates of parametric, predictive, and model selection uncertainty. The success of these models permits us to put theoretically and procedurally sound uncertainty estimates to a process that is widely acknowledged to be variable and uncertain but has, to date, not developed robust statistical tools to quantify this uncertainty. This dissertation comprises four individual papers that methodically develop and prove the concept of Bayesian statistical sediment transport models. A simple pedagogical model is developed using synthetic and laboratory flume data---this model is then compared to traditional statistical approaches that are more familiar to the discipline. A single-fraction sediment transport model is developed on the Snake River to develop a probabilistic sediment budget whose results are compared to a sediment budget developed through an ad hoc uncertainty analysis. Lastly, a multi-fraction sediment transport model is developed in which multiple fractions of laboratory flume experiments are modeled and the results are compared to the standard theory that has been already published. The results of these models demonstrate that a Bayesian approach to sediment transport has much to offer the discipline as it is able to 1) accurately provide estimates of model parameters, 2) quantify parametric uncertainty of the models, 3) provide a means to evaluate relative model fit between different deterministic equations, 4) provide predictive uncertainty of sediment transport, 5) propagate uncertainty from the root causes into secondary and tertiary dependent functions, and 6) provide a means by which testing of established theory can be performed.

On ocean beds around the world sinking particles of residual biomass from animals and vegetation have, over time, accumulated into thick layers of sediment. For researchers these layers of sediment contain valuable information about the earth’s past climate, information that can be used to predict future climate change and to estimate natural vs. human causes of today’s global warming.The Institute of Geosciences at Stockholm University studies sediment cores that have been acquired with an apparatus called piston corer. For future coring expeditions additional piston corers are needed. However, the dimensions of the current piston corer only allow use of specially ordered polycarbonate pipes which are used as liners. This pipe is relatively expensive compared to pipes of other plastic materials and of standard dimensions. In order to lower the cost of upcoming coring expeditions a new piston corer design, that allows use of standard dimension plastic pipes as liners, is required.The process of redesigning the piston corer offers an opportunity to identify structural disadvantages and generate ideas on improvements that can be accomplished in a new construction. It also offers an opportunity to satisfy a need to increase the dimensions of the liners in order to acquire a larger volume of sediment, thereby increasing the reliability of the information gathered when examining the sediment. This thesis paper presents the process of redesigning and improving the piston corer currently used by Stockholm University. The resulting design allows use of standard dimension PVC pipes as liners. It has also been improved regarding the process of extruding liners after recovering sediment cores as well as performance and handling.

High rates of soil loss and high sediment loads in rivers necessitate efficient monitoring and quantification methodologies so that effective land management strategies can be designed. Constructing a sediment budget is a useful approach to address these issues. Quantifying a sediment budget using classical field-based techniques, however, is labour intensive and expensive for poorly gauged, large drainage basins. The availability of global environmental datasets in combination with GIS techniques provides an opportunity for studying large basins. Following this approach, a framework is presented for constructing sediment budgets for large, data-sparse drainage basins, which is applied to the mountainous upper Indus River basin in northern Pakistan. The methodological framework consists of five steps: (1) analyzing hydro-climatological data for dividing the drainage basin into characteristic regions, and calculating sediment yields; (2) investigation of major controls on sediment yields; (3) identification and mapping of sediment source areas by spatially distributed modelling of erosional processes; (4) spatially distributed modelling of sediment yields; and (5) carrying out the sediment budget balance calculation at the basin outlet. Further analysis carried out on the Indus data has enabled a better understanding of sediment dynamics in the basin.<p> Analysis of the available hydro-climatological data indicates that the basin can be subdivided into three characteristic regions based on whether runoff production and subsequent sediment generation is controlled by temperature (Region 1, upper, glacierized sub-basins), precipitation caused by the monsoon and western disturbances (Region 3, lower sub-basins), or a combination of the two (Region 2, middle reach sub-basins). It is also demonstrated that contrary to the conventional model, the specific sediment yield increases markedly with drainage area along the Indus River. An investigation of major controls on specific sediment yield in the basin indicates that percent snow/ice cover is a major land cover control for specific sediment yield. Spatially distributed erosion modelling predictions indicate that 87% of the annual gross erosion takes place in the three summer months with greatest erosion potential concentrated in sub-basins with high relief and a substantial proportion of glacierized area. Lower erosion rates can be explained by the arid climate and low relief on the Tibetan Plateau, and by the dense vegetation and lower relief in the lower monsoon sub-region. The model predicts an average annual erosion rate of 3.2 mm/a or 868 Mt/a. Spatially distributed sediment yield predictions made with coupled models of erosion and sediment delivery indicate that the Indus sub-basins generally show an increase of sediment delivery ratio with basin area. The predicted annual basin sediment yield is 244 Mt/a and the overall sediment delivery ratio in the basin is calculated as 0.28. The long-term mean annual sediment budget, based on mass balance, is characterized by a gross erosion of 762.9, 96.7 and 8.4 Mt, and a gross storage of 551.4, 66.1, and 6.5 Mt in the upper, middle, and lower regions of the basin, respectively. The sediment budget indicates that the major sources of eroded sediment are located in the Karakoram, in particular in the Hunza basin. Substantial sediment storage occurs on the relatively flat Tibetan Plateau and the Indus River valley reach between Partab Bridge and Shatial. The presented framework for sediment budget construction requires relatively few data, mostly derived from global datasets. It therefore can be utilized for other ungauged or poorly gauged drainage basins of the world.