Loading...
Remote estimation of grain size distribution and simulating grain size effects on debris-flow rheology and post-fire variability of transport by dry ravel
Jacobson, Hayden Lee
Jacobson, Hayden Lee
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2024
Date Submitted
Keywords
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
The size distribution of granular material influences the length scale of sediment transport by dry ravel and debris flows. Variability in erosion rates and the grain size distribution of material delivered to channels has been observed due to an increase in the annual areal extent of wildfire in the Western United States. As these changes modify debris-flow hazard, it is essential to constrain the influence of grain size distribution on dry ravel and debris flows, especially in a post-fire setting.
To improve understanding of evolving transport regimes and propose grain-size-dependent constraints on process phenomena, the research presented in this thesis leverages increased access to remotely sensed morphological data and recent advancements in simulations of sediment transport. Here, remotely sensed point cloud data, statistical modeling, and numerical simulation were used to contribute towards the estimation of grain size and the modeling of grain size effects on dry ravel and debris flow.
First, a statistical description of the travel distance exceedance probability distribution of dry ravel was applied to a series of experiments simulating dry ravel with different grain sizes, aspects, and slope angles as a site in the Diablo Range, California recovered from wildfire. The forms of these distributions suggested that the largest naturally occurring grains exhibited a transition from more bounded to more runaway motion during the first year post-fire, after which seasonal changes in vegetation density controlled particle mobility.
Following these efforts, a workflow involving machine learning and a novel clustering approach (termed DBloops) was developed to extract the size distribution of angular cobble and boulder sized clasts from point clouds of nonplanar deposits. This workflow was successfully applied to point clouds of debris-flow and riverbank deposits, reproducing grain size distributions obtained from Wolman pebble counts and outperforming an alternative algorithmic approach.
To achieve the first formal connection between grain size distribution and controlling parameters in the numerical model D-Claw, a compositional mapping of grain size distribution to debris flow permeability and critical volume fraction was implemented. This mapping was found to perform adequately over a range of different compositions by simulating experiments conducted by others at small (2 m) and large (95 m) debris flow flumes.
By integrating knowledge of post-fire changes in the grain size distribution of ravel and the size distributions of the largest particles in debris-flow deposits, a better understanding of the influence of how debris-flow composition changes post-fire will be obtained. By formally implementing grain-size-based parameter selection in D-Claw, these findings will help assess and manage the risk debris-flow hazards present to communities and infrastructure as fire regimes change.
Associated Publications
Rights
Copyright of the original work is retained by the author.