Novel Numerical Methods and Future Projections in Debris-Flow Hazard Assessments, Overland Flow Routing, and Global Suspended Sediment Fluxes
Author
Prescott, Alexander B.Issue Date
2024Keywords
BayesianClimate change
Diffusive wave approximation
Iterative numerical methods
Markov Chain Monte Carlo
Suspended sediment
Advisor
Pelletier, Jon D.
Metadata
Show full item recordPublisher
The University of Arizona.Rights
Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.Abstract
Natural processes continuously shape the earth’s surface as bedrock weathers to grains of sediment that are systematically transported through the continental interiors to coastal river deltas. From hillslopes to low-order stream valleys and into the largest terrestrial fluvial networks, flowing water slowly but surely transports the weathered remnants of the continents into the global oceans. While these phenomena are generally well appreciated and have been studied by scientists for more than a century, nonlinear relations, threshold responses, and feedback loops characterize hillslope erosion and fluvial sediment transport as well as the moving water that drives them, prohibiting the use of closed-form solutions in all but the most idealized of cases. Properly chosen numerical methods that approximate solutions of differential equations enable the investigation of research questions involving the complex interactions of overland flow and sediment transport processes. In the chapters contained in this dissertation, we employ novel numerical methods to solve problems at the intersection of hydrology and geomorphology. We rely heavily on iterative methods to simulate the flow of water or water-laden sediment mixtures over topography, and Bayesian calibration methods are combined with Markov Chain Monte Carlo sampling methods to incorporate underlying uncertainties into output simulated quantities. In Appendix A, a methodological framework for the probabilistic prediction of postfire debris-flow inundation is introduced that ties together existing models of debris-flow likelihood, volume, and runout with an ensemble of high-resolution atmospheric models. We demonstrate that even with considerable uncertainties in each of the model components, the modeling framework is capable of capturing observed debris-flow extreme events in the range of simulated inundation outcomes. In Appendix B, we revisit flow routing algorithms for the calculation of contributing area that have been in use for several decades across earth science disciplines, demonstrating that one of the most popular contains inherent biases that cause grid-orientation dependent results. We then introduce a novel flow routing algorithm that iteratively and efficiently constructs a steady-state water surface profile that compares well with solutions of the shallow water equations. In Appendix C, future climatological, hydrological, and biological variables from global climate model simulations are used to drive a fluvial suspended sediment flux model, producing spatially distributed projections of suspended sediment yield around the globe by the year 2100. Although each future variable results in significant latitudinal changes in suspended sediment yield, temperature and vegetation changes drive the largest magnitude changes at high-latitudes, while vegetation dominates the sediment response at lower latitudes. The results of this study indicate that globally averaged suspended sediment delivery to the coastal margins will be amplified under future climate conditions, although regional uncertainty remains, largely a consequence of divergent projections of vegetation cover by different earth system models.Type
Electronic Dissertationtext
Degree Name
Ph.D.Degree Level
doctoralDegree Program
Graduate CollegeGeosciences