Date of Award

Spring 12-31-2013

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Geological Sciences

First Advisor

Gregory E. Tucker

Second Advisor

Robert S. Anderson

Third Advisor

Brian Hynek

Fourth Advisor

Cameron W. Wobus

Fifth Advisor

Eleanor R. Griffin

Abstract

Environments respond to large floods in complex and often spectacular ways. Connecting the topographic change caused by these rare events to principles of hydrodynamics and sediment transport is critical for quantifying their contribution to the long term evolution of landscapes.

Two sites in the southern Rocky Mountains serve as natural experiments on the effects of large floods. Vegetation along the Lower Rio Puerco, New Mexico, governs reach-scale patterns of erosion and deposition during overbank events. A multi-temporal lidar dataset bracketing a flood in 2006 captures the redistribution of sediment from a devegetated area of the floodplain onto a vegetated reach downstream. The observed depth of deposition is correlated with vegetation density, and the variability in the depth with vegetation type. A one-dimensional model of sediment transport suggests that material deposited within a few kilometers of its source, and that the uniform aggradation seen elsewhere reflects local erosion of the arroyo walls.

Numerical models that reproduce the interactions of flow, sediment, and vegetation can be used to explore the behavior of rivers in response to changing drivers. We can reproduce, to first order, the response to the 2006 flood, indicating that we have identified the minimum complexity required to quantitatively describe the behavior of the system. Historical reconstructions of the Rio Puerco suggest that the arroyo cycle could be driven by autogenic processes arising without anthropogenic forcings.

Simulations of the 2013 Front Range floods in Colorado reproduce the spatial and temporal variability seen in sediment fluxes. Small, low elevation catchments move large volumes of coarse sediment far from the range during intense, episodic flows, while the response of large basins is attenuated and prolonged. A probabilistic theory for sediment motion is needed to capture the distribution of particle travel distances and rest times during floods. I present a theoretical framework for this approach and show that it can predict the origin and fate of particles in the network.

This work contributes new methods for quantifying and predicting the behavior of landscapes in response to major floods, which will only become more frequent with a changing climate.

Share

COinS