The Mississippi River is considered the ultimate single channel meandering river. Five hundred km upstream from its mouth, about 25% of the river’s discharge is diverted into the Atchafalaya River. This diversion is controlled by the Old River Control structure, built by the US Army Corps of Engineers in stages since 1963, to stop the avulsion of the Mississippi River into the Atchafalaya. The study area is a 119 km sandy bedded reach near Old River Control that is highly impacted by engineering measures. Channel dimensions average 1,000 m wide with average thalweg depths of 23 m. The mean annual discharge is 15,000 m3s-1 with a water surface slope of 0.06 m per river mile. In a sandy bedded river, the effective discharge is the discharge which cumulatively transports the most sand. This study evaluates how the Old River Control structure has influenced an adjusting effective discharge between 1978 and 2011. The bed load component of sand transport is included by employing a novel, automated, cross-correlation technique. It was found that the upper limit for discharges that cumulatively transport 85% of the total sand load has decreased from 34,000 m3s-1 to 28,000 m3s-1 between 1978 and 2011. Sand transport from 1982 to 2011 occurred during progressively greater ratios of water discharge to the Atchafalaya River and corresponded to an aggradational trend in the nearby Mississippi River at Red River Landing stream gage. The combination of this sand transport trend and nearby channel aggradation is some indication that the diversion may not be stable and that the avulsion of the lower Mississippi River is ongoing. However, sand was transported at similar discharge ratios in the 1978 to 1982 and 2002 to 2011 periods. Future trends will reveal definitively if these findings indicate that the lower Mississippi River avulsion is continuing. Two aims are pursued by placing the effective discharge approach into the geomorphologic context of the study area. Ten zones are categorized by four distinct geomorphologic classes: meander, no islands; meander, geologic control; meander, islands and divided flow; and straight zones. One, these classes have merit for future research and are shown to be geomorphologically distinct in several ways: natural levee height and channel planform adjustment relationships, sinuosity and width to depth ratios, and bedform depth to height scaling. Two, this approach allows the comparison of the effective discharge to the study area’s geomorphology. Process-form linkages can be made between sediment transporting events and the three levels in a fluvial hierarchy: fluvial bedforms and channels, bars, and levees. Median grain size and channel position of sediment samples from these three levels were plotted on combined LiDAR and bathymetric derived cross- sections from specific geomorphologic zones. This analysis indicates that the fluvial hierarchy coincides with the stages of effective discharge but seems to scale to the elevation of natural levees. This study interjects a geomorphologic approach into the lower Mississippi River discourse and raises a number of interesting questions for further research. / text

The purpose of this study is to analyze how estimates of an important geomorphic parameter, effective discharge, are impacted by the choice of bed material load equations and flow duration curves (FDCs). The Yang (1979), Brownlie (1981), and Pagosa equations developed by Rosgen (2006) were compared for predicting bed material load. To calculate the bed material load using the Pagosa equations, the bedload and suspended load are calculated separately and the results are added together. To compare the effectiveness of the equations, measured bed material load data from the USGS Open-File Report 89-67 were used. Following the calculations, the equation results were compared to the measured data. It was determined that the Pagosa equations performed the best overall, followed by Brownlie and then Yang. The superior performance of the Pagosa equations is likely due to the equations being calibrated. USGS regression equations for FDCs were compared to a method developed by Dr. David Rosgen in which a dimensionless FDC (DFDC) is developed. Weminuche Creek in southwestern Colorado was used as the study site. Rosgen's DFDC method requires the selection of a streamgage for a stream that exhibits the same hydro-physiographic characteristics as the site of interest. An FDC is developed for the gaged site and made dimensionless by dividing the discharges by the bankfull discharge of the gaged site. The DFDC is then made dimensional by multiplying by the bankfull discharge of the site of interest and the resulting dimensional FDC is taken as the FDC of the ungaged site. The USGS regression equations underpredicted the discharges while Rosgen's DFDC method overpredicted them. Rosgen's DFDC method produced more accurate results than the USGS regression equations for Weminuche Creek. To calculate the effective discharge, the FDC was used to develop a flow frequency curve which was then multiplied by the sediment rating curve. Effective discharge calculations were performed for Weminuche Creek using several combinations of bed material load prediction equations and FDCs. The USGS regression equations, Rosgen's DFDC method, and streamgage data were all used in conjunction with the Yang and Pagosa equations. The Brownlie equation predicted zero bed material load for Weminuche Creek, and was thus not used to calculate the effective discharge. When the USGS regression equations were used with the Yang and Pagosa equations, the calculated effective discharge was approximately 4.5 cms for both bed material load prediction equations. When Rosgen's DFDC method and streamgage data were used with the Yang and Pagosa equations, the effective discharge was approximately 13.5 cms. From these results, it was determined that the bed material load prediction equations had little impact on the effective discharge for Weminuche Creek while the FDCs did influence the results.

No description available.

Cette thèse examine les impacts sur la morphologie des tributaires du fleuve Saint-Laurent des changements dans leur débit et leur niveau de base engendrés par les changements climatiques prévus pour la période 2010–2099. Les tributaires sélectionnés (rivières Batiscan, Richelieu, Saint-Maurice, Saint-François et Yamachiche) ont été choisis en raison de leurs différences de taille, de débit et de contexte morphologique. Non seulement ces tributaires subissent-ils un régime hydrologique modifié en raison des changements climatiques, mais leur niveau de base (niveau d’eau du fleuve Saint-Laurent) sera aussi affecté. Le modèle morphodynamique en une dimension (1D) SEDROUT, à l’origine développé pour des rivières graveleuses en mode d’aggradation, a été adapté pour le contexte spécifique des tributaires des basses-terres du Saint-Laurent afin de simuler des rivières sablonneuses avec un débit quotidien variable et des fluctuations du niveau d’eau à l’aval. Un module pour simuler le partage des sédiments autour d’îles a aussi été ajouté au modèle. Le modèle ainsi amélioré (SEDROUT4-M), qui a été testé à l’aide de simulations à petite échelle et avec les conditions actuelles d’écoulement et de transport de sédiments dans quatre tributaires du fleuve Saint-Laurent, peut maintenant simuler une gamme de problèmes morphodynamiques de rivières. Les changements d’élévation du lit et d’apport en sédiments au fleuve Saint-Laurent pour la période 2010–2099 ont été simulés avec SEDROUT4-M pour les rivières Batiscan, Richelieu et Saint-François pour toutes les combinaisons de sept régimes hydrologiques (conditions actuelles et celles prédites par trois modèles de climat globaux (MCG) et deux scénarios de gaz à effet de serre) et de trois scénarios de changements du niveau de base du fleuve Saint-Laurent (aucun changement, baisse graduelle, baisse abrupte). Les impacts sur l’apport de sédiments et l’élévation du lit diffèrent entre les MCG et semblent reliés au statut des cours d’eau (selon qu’ils soient en état d’aggradation, de dégradation ou d’équilibre), ce qui illustre l’importance d’examiner plusieurs rivières avec différents modèles climatiques afin d’établir des tendances dans les effets des changements climatiques. Malgré le fait que le débit journalier moyen et le débit annuel moyen demeurent près de leur valeur actuelle dans les trois scénarios de MCG, des changements importants dans les taux de transport de sédiments simulés pour chaque tributaire sont observés. Ceci est dû à l’impact important de fortes crues plus fréquentes dans un climat futur de même qu’à l’arrivée plus hâtive de la crue printanière, ce qui résulte en une variabilité accrue dans les taux de transport en charge de fond. Certaines complications avec l’approche de modélisation en 1D pour représenter la géométrie complexe des rivières Saint-Maurice et Saint-François suggèrent qu’une approche bi-dimensionnelle (2D) devrait être sérieusement considérée afin de simuler de façon plus exacte la répartition des débits aux bifurcations autour des îles. La rivière Saint-François est utilisée comme étude de cas pour le modèle 2D H2D2, qui performe bien d’un point de vue hydraulique, mais qui requiert des ajustements pour être en mesure de pleinement simuler les ajustements morphologiques des cours d’eau. / This thesis investigates the impacts of climate-induced changes in discharge and base level on the morphology of Saint-Lawrence River tributaries for the period 2010–2099. The selected tributaries (Batiscan, Richelieu, Saint-Maurice, Saint-François and Yamachiche rivers) were chosen because of their differences in size, flow regime and morphological setting. Not only will these tributaries experience an altered hydrological regime as a consequence of climate change, but their base level (Saint-Lawrence River water level) will also change. A one-dimensional (1D) morphodynamic model (SEDROUT), originally developed for aggrading gravel-bed rivers, was adapted for the specific context of the Saint-Lawrence lowland tributaries in order to simulate sand-bed rivers with variable daily discharge and downstream water level fluctuations. A module to deal with sediment routing in channels with islands was also added to the model. The enhanced model (SEDROUT4-M), which was tested with small-scale simulations and present-day conditions in four tributaries of the Saint-Lawrence River, can now simulate a very wide range of river morphodynamic problems. Changes in bed elevation and bed-material delivery to the Saint-Lawrence River over the 2010–2099 period were simulated with SEDROUT4-M for the Batiscan, Richelieu and Saint-François rivers for all combinations of seven tributary hydrological regimes (present-day and those predicted using three global climate models (GCM) and two greenhouse gas emission scenarios) and three scenarios of how the base level provided by the Saint-Lawrence River will alter (no change, gradual decrease, step decrease). The effects on mean annual sediment delivery and bed elevation differ between GCM and seem to be related to whether the river is currently aggrading, degrading or in equilibrium, which highlights the importance of investigating several rivers using several climate models in order to determine trends in climate change impacts. Despite the fact that mean daily discharge and mean annual maximum discharge remain close to their current values in the three GCM scenarios for daily discharge, marked changes occur in the mean annual sediment transport rates in each simulated tributary. This is due to the important effect of more frequent large individual flood events under future climate as well as a shift of peak annual discharge from the spring towards the winter, which results in increased variability of bed-material transport rates. Some complications with the 1D modelling approach to capture the complex geometry of the Saint-Maurice and Saint-François rivers suggest that the use of a two-dimensional (2D) approach should be seriously considered to accurately simulate the discharge distribution at bifurcations around islands. The Saint-François River is used as a test case for the 2D model H2D2, which performs well from a hydraulics point of view but which needs to be adapted to fully simulate morphological adjustments in the channel.

No description available.