River Hydro- and Morphodynamics: Restoration, Modeling, and Uncertainty

Abstract

The study of fluvial geomorphology is one of the critical sciences in the 21st Century. The previous century witnessed a virtual disregard of the hydro and morphodynamic processes occurring in rivers when it came to design of transportation, flood control, and water resources infrastructure. This disregard, along with urbanization, industrialization, and other land uses has imperiled many waterways. New technologies including geospatially referenced data collection, laser-based measurement tools, and increasing computational powers by personal computers are significantly improving our ability to represent these complex and diverse systems. We can accomplish this through both the building of more sophisticated models and our ability to calibrate those models with more detailed data sets. The effort put forth in this dissertation is to first introduce the accomplishments and challenges in fluvial geomorphology and then to illustrate two specific efforts to add to the growing body of knowledge in this exciting field.First, we explore a dramatic phenomenon occurring in the Middle Rio Grande River. The San Marcial Reach of the Rio Grande River has experienced four events that completely filled the main channel with sediment over the past 20 years. This sediment plug has cost the nation millions of dollars in both costs to dredge and rebuild main channels and levees, along with detailed studies by engineering consultants. Previous efforts focused on empirical relations developed with historical data and very simple one dimensional representation of river hydrodynamics. This effort uses the state-of-the-art three-dimensional hydro and morphodynamic model Delft3D. We were able to use this model8to test those hypotheses put forth in previous empirical studies. We were also able to use this model to test theories associated with channel avulsion. Testing found that channel avulsions thresholds do exist and can be predicted based on channel bathymetric changes.The second effort included is a simple yet sophisticated model of river meander evolution. Prediction of river meandering planform evolution has proven to be one of the most difficult problems in all of geosciences. The limitations of using detailed three dimensional hydro and morphodynamic models is that the computational intensity precludes the modeling of large spatial or temporal scale phenomenon. Therefore, analytical solutions to the standard Navier-Stokes equations with simplifications made for hydrostatic pressure among others, along with sediment transport functions still have a place in our toolbox to understand and predict this phenomenon. One of the most widely used models of meander propagation is the Linear Bend Model that employs a bank erosion coefficient. Due to the various simplifications required to find analytical solutions to these sets of equations, efforts to build the stochasticity seen in nature into the models have proven useful and successful. This effort builds upon this commonly used meander propogation model by introducing stochasticity to the known variability in outer bank erodibility, resulting in a more realistic representation of model results.