The two-phase flow approach has been the conventional method designed to study the sediment transport rate. Due to the complexity of sediment transport, the precisely numerical models computed from that approach require initial assumptions and, as a result, may not yield accurate output for all conditions. This research work proposes that Machine Learning algorithms can be an alternative way to predict the processes of sediment transport in two-dimensional directions under oscillating sheet flow conditions, by utilizing the available dataset of the SedFoam multidimensional two-phase model. The assessment utilized linear regression and gradient boosting algorithm to analyze the lowest average mean squared error in each case and search for the best partition method based on the domain height of the simulation setup.

Braided channels arise due to high sediment availability in conjunction with regular competent flows and readily erodible banks. Together, these boundary conditions lead to the deposition and reworking of a network of transient bars that characterize the braided planform. However, quantifying the geomorphic response of braided systems to alterations in these boundary conditions is not straightforward, as channels adjust over a wide range of timescales, rendering traditional field-based observation intractable. As such, the development of simple yet robust relationships between channel morphology and sediment transport has the potential to allow predictions of channel response to altered hydrologic or sediment regimes. In this research, I first use laboratory flume experiments to relate particle travel distance during floods (termed particle path length) and the spacing of channel bars in braided rivers (Chapter 2), finding that deposition sites for sediment in transport can be readily predicted by the characteristic confluence-diffluence spacing in a reach. I then use the relationship between path length and channel morphology to build a simple, open-source morphodynamic model for braided rivers that computes sediment transport using path-length distributions derived from bar spacing (Chapter 3). I explore the validity of this model, specifically noting that its modular framework allows exploration of process representations in morphodynamic modeling in ways existing models do not. Finally, I employ the model to determine the role of sediment supply in braided channel bar morphodynamics (Chapter 4). Specifically, I address the relative roles of sediment sourced from upstream versus sediment sourced from within a braided reach in terms of channel morphodynamics at decadal timescales. This research demonstrates that simple scaling relationships, while necessarily imperfect, nevertheless provide insight into morphodynamic processes in braided rivers, while also allowing predictions of channel response to sediment or hydrologic forcing at the timescales of channel adjustment.

Les grandes quantités de sédiments transportés par suspension dans les rivières alpines sont associées à d’importantes problématiques socio-économiques et environnementales telles que le transport de polluants, la dégradation des milieux aquatiques ou l’envasement des retenues hydroélectriques. Pour faire face à ces enjeux, il est nécessaire de mieux comprendre le rôle joué par le lit des rivières alpines sur la dynamique de ce transport. A partir de larges bases de données issues de la littérature et de nouvelles mesures de terrain, la première partie de cette thèse propose une étude à une échelle régionale de i) l’influence de la configuration des sources sédimentaires sur la variabilité du transport solide par suspension, ii) l’estimation de la quantité et de la disponibilité des particules fines dans le lit des rivières alpines et iii) l’analyse de la relation entre transport solide par suspension et mobilité du lit de ces rivières. Ces analyses montrent que la configuration du bassin versant en amont du point d’observation contrôle significativement la dynamique du transport solide par suspension observée en aval. De grandes quantités de sédiments fins sont en effet stockées dans les rivières alpines alluviales et il est ainsi possible de prédire une partie significative de leurs flux en suspension pour les forts débits à partir d’une modélisation de la mobilité de leurs lits. La seconde partie de la thèse teste ces résultats à une échelle locale. Pour cela, une campagne de mesures a été réalisée durant une saison complète de fonte sur un cours d’eau alpin typique, la Séveraisse. Un large panel de mesures directes et indirectes a été mis en œuvre pour mesurer la suspension, le charriage et les évolutions topographiques sur un tronçon de 3.5 km. Ces mesures confirment que les particules fines transportées par suspension interagissent fortement avec le lit dans ce type de tronçon morphodynamiquement actif. Ce dernier peut être perçu comme une zone tampon intermédiaire contrôlée par le forçage amont sédiments-débit liquide ainsi que par la mobilité et la morphologie de son lit. / The large quantities of sediments transported as suspension in Alpine rivers are associated with important socio-economic and environmental issues such as pollutant transfer, aquatic habitat degradation or dam siltation. To address these issues, it is required to better understand the role of Alpine river beds on the dynamics of this transport. In the first part of this thesis, we use large datasets from the literature and new field measurements to investigate at a regional scale i) the influence of sediment sources configuration on suspended load variability, ii) the quantity and availability of fine particles in Alpine river beds and iii) the relation between suspended load and river bed mobility. These analyses show that the catchment configuration upstream the observation point can significantly control the suspended load dynamics observed downstream. This first part also shows that large quantities of fine particles can be stored in alluvial Alpine rivers. For these rivers and for high flow rates, it was possible to predict a significant part of suspended load based on riverbed mobility modeling. The second part of the thesis tests these results at a local scale. To do so, an important field campaign was performed during the entire melting season of a typical Alpine river, the Séveraisse. A large panel of direct and indirect measurements was used to measure suspended load, bedload and topographic changes on a 3.5-km reach. These measurements confirm that suspended particles strongly interact with the river bed in that kind of morphodynamically active streams. The latter can be considered as an intermediate buffer controlled by the upstream hydro-sedimentary forcing and by the river bed mobility and morphology.

Thesis (Ph. D.)--Joint Program in Oceanography (Massachusetts Institute of Technology, Dept. of Earth, Atmospheric, and Planetary Sciences; and the Woods Hole Oceanographic Institution), 2011. / Cataloged from PDF version of thesis. / Includes bibliographical references. / The sinking flux of particulate matter into the ocean interior is an oceanographic phenomenon that fuels much of the metabolic demand of the subsurface ocean and affects the distribution of carbon and other elements throughout the biosphere. In this thesis, I use a new suite of observations to study the dynamics of marine particulate matter at the contrasting sites of the subtropical Sargasso Sea near Bermuda and the waters above the continental shelf of the Western Antarctic Peninsula (WAP). An underwater digital camera system was employed to capture images of particles in the water column. The subsequent analysis of these images allowed for the determination of the particle concentration size distribution at high spatial, depth, and temporal resolutions. Drifting sediment traps were also deployed to assess both the bulk particle flux and determine the size distribution of the particle flux via image analysis of particles collected in polyacrylamide gel traps. The size distribution of the particle concentration and flux were then compared to calculate the average sinking velocity as a function of particle size. I found that the average sinking velocities of particles ranged from about 10-200 m d- and exhibited large variability with respect to location, depth, and date. Particles in the Sargasso Sea, which consisted primarily of small heterogeneous marine snow aggregates, sank more slowly than the rapidly sinking krill fecal pellets and diatom aggregates of the WAP. Moreover, the average sinking velocity did not follow a pattern of increasing velocities for the larger particles, a result contrary to what would be predicted from a simple formulation of Stokes' Law. At each location, I derived a best-fit fractal correlation between the flux size distribution and the total carbon flux. The use of this relationship and the computed average sinking velocities enabled the estimation of particle flux from measurements of the particle concentration size distribution. This approach offers greatly improved spatial and temporal resolution when compared to traditional sediment trap methods for measuring the downward flux of particulate matter. Finally, I deployed specialized in situ incubation chambers to assess the respiration rates of microbes attached to sinking particles. I found that at Bermuda, the carbon specific remineralization rate of sinking particulate matter ranged from 0.2 to 1.1 d', while along the WAP, these rates were very slow and below the detection limit of the instruments. The high microbial respiration rates and slow sinking velocities in the Sargasso Sea resulted in the strong attenuation of the flux with respect to depth, whereas the rapid sinking velocities and slow microbial degradation rates of the WAP resulted in nearly constant fluxes with respect to depth. / by Andrew M. P. McDonnell. / Ph.D.

Thesis (M. Eng. in Ocean Engineering)--Joint Program in Oceanography/Applied Ocean Science and Engineering (Massachusetts Institute of Technology, Dept. of Mechanical Engineering; and the Woods Hole Oceanographic Institution), 1999. / Includes bibliographical references (leaves 97-103). / by Rachel H. Levine. / M.Eng.in Ocean Engineering

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第17865号 / 工博第3774号 / 新制||工||1577(附属図書館) / 30685 / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 牛島 省, 教授 角 哲也, 准教授 米山 望 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Streambank erosion is a major problem and a major known source of sediment in impaired streams. Stream deterioration is mainly due to the excess sediment in the United States. Many models have been developed to predict streambank erosion and sediment transport in the streams. Determining the most sensitive soil-specific parameters of the CONCEPTS Model for Goodwin Creek, MS was the focus of the study. The Latin Hypercube Oneactor-At-a-Time (LH-OAT) method was used to complete the sensitivity analysis on soil-specific parameters in CONCEPTS. Overall results demonstrate that erodibility and critical shear stress parameters should be determined very carefully and realistic to determine streambank erosion and sediment transport rate more accurately. This sensitivity analysis also shows the minimum effects of suction angle and cohesion on results. In this case, making an assumption in a literal range, or safely ignoring them should not cause a big variation on CONCEPTS results.

Sediment has profound effects on water quality. Correspondingly, water quality modeling often needs sediment transport modeling. However, simplified descriptive sediment transport was originally employed for water quality modeling, and the linkage between sediment transport models and water quality models is less developed. Therefore, the main purposes of this study were to develop general methods of coupling sediment transport and water quality models and to improve sediment transport modeling for water quality modeling. Linkage of sediment transport and water quality was discussed and a comprehensive sediment transport literature review was conducted. SEDDEER (Sediment Deposition and Erosion), a stand-alone sediment and contaminant fate and transport model, which simulates one water box and the underlying multiple sediment bed layers, was developed. SEDDEER for Visual Basic for Application (SEDDEER_VBA) was written in VBA. SEDDEER for FORTRAN (SEDDEER_FOR) is the corresponding FORTRAN model. To improve WASP in terms of sediment transport, SEDDEER_FOR was incorporated into the WASP TOXI7 module as the starting point to generate the coupled WASP model (WASP_SEDDEER). Verification and validation of SEDDEER_VBA were conducted prior to model application and incorporation. A comprehensive model test was performed to show that SEDDEER_FOR is computationally identical to SEDDEER_VBA. Simple tests were carried out to verify the fluxes across the sediment-water interface and ensure that the coupling of the WASP water column and SEDDEER bed models is correct. The testing results indicated that these models were verified and/or validated. SEDDEER was used to evaluate the effects of sediment on contaminant transport. WASP_SEDDEER, WASP7.4, and EFDC were applied to Mobile Bay to demonstrate the capabilities of WASP_SEDDEER, and WASP_SEDDEER produced a reasonable and consistent modeling result. The results of the study indicated that SEDDEER can be used for one-box sediment and contaminant fate and transport modeling, and also incorporated into water quality models. In addition, WASP_SEDDEER coupling was implemented correctly and can be applied to the real world. Finally, study results show that sediment affects contaminant fate and transport mostly by external forcing and flow conditions, and contaminant fate and transport varies with different sediment and contaminant characteristics and sediment transport processes.

The Town Creek Watershed (TCW) is a representative area of the Tombigbee River Basin and the Southeastern Plains Ecoregion in Mississippi. The principal channel and four main tributaries have been included for several years within the MS Section 303(d) list of waterbodies biologically impaired due to sediment. The TMDL developed for TCW recommended that streams located near cultivated lands, road crossings and construction activities are a priority for streambank and riparian buffer zone restoration and sediment loads reduction. Development of remedial measures and future BMPs within TCW for reducing water quality impairment and downstream dredging costs requires identification of sediment sources and loads currently transported within TCW. Streambank erosion processes were hypothesized to be an important mechanism driving sediment supply from TCW. The overall goal of this research was to identify mechanisms and the potential effects of streambank erosion processes and to quantify and model the magnitude and rates of these processes within TCW. Research goal and specific aims were addressed in four substudies combining field reconnaissance and detailed data collection, laboratory analysis and computational modeling techniques. The first substudy involved a temporal and spatial analysis of observed suspended sediment transport rates, determined the stage of channel evolution and identified streambank erosion as an important source of sediment supply for reaches in TCW. Streambank erosion contributions of up to 28.5 Mg per m of streambank were quantified in a second substudy monitoring and determining streambank erosion processes and factors within TCW. Results from a third substudy assessed predictions of the computational model CONCEPTS for time of occurrence and magnitude of streambank failures and top width retreat along a 270-m modeling reach. Empirical and analytical approaches to estimate rates and depths of fluvial erosion were developed in a final substudy. The rate and depth of fluvial erosion were estimated as a function of hydraulic and hydrologic properties of flow events, vegetation on streambanks, flow induced forces and streambank geometry and soil properties. Reduction of suspended sediment loads should focus on attenuation of geomorphic processes and stabilization of reaches and agricultural lands adjacent to streambanks along incised headwaters within TCW.

No description available.