River channel dynamics mapped with dense lidar point cloud time series

Sinuous channel-like landforms are observed on planets and moons throughout the Solar System. On Earth’s terrestrial surface, a significant fraction of these landforms are created by lowland rivers that drain water, sediment, and organic matter from upland landscapes. Many of these are self-formed alluvial river channels that have a bed and banks made of mobile material. Interactions between the flow and the material lining the channel cause alluvial channel bends to migrate laterally, or meander, resulting in dynamic topographic changes observable over human time scales. The beauty and ubiquity of the patterns created by meander migration—as well as the consequences of rapid channel movement for river-adjacent communities—have led to a deep history of their study from both geologic and water engineering-driven perspectives. Observations and theory argue that many physical mechanisms driving alluvial channel migration scale with the weighted sum of upstream channel curvature, thus, many models rely on channel centerline curvature to predict river migration rates. Time series of satellite-derived images show that the migration rates of large alluvial river bends are indeed correlated with centerline curvature over multi-year time scales. This relationship forms the basis for many of the most widely-used models in geomorphology. However, migration does not actually happen at a constant rate and the channel does not actually maintain a constant width over time as is framed in these models. Channel bank erosion and bar deposition are driven by floods, which transport variable amounts of sediment that can quickly change the channel shape over time scales that range from hours to months. This raises a fundamental question about river behavior that is broadly unanswered—how will a channel respond to a single flood? My work aims to address this question by describing how two channel bends on the actively migrating Lower Trinity River in Texas respond to a range of flood sizes. To do this, I leverage high-resolution airplane- and drone-based lidar datasets that describe the detailed shape of the bank surface topography. Changes between the datasets show how the outer channel banks evolve over time scales ranging from two months to 14 years. A second motivation for my work is to advance how dense topographic time series data can be applied to understand the evolution of planetary surfaces. Lidar-derived topographic datasets have shifted how researchers study Earth surface processes—detailed, repeat measurements now present a rich opportunity to see precisely how landscapes change over time. To achieve this, I present a detailed collection and processing pipeline that I developed to map river channels using drone-based lidar. I also describe two workflows that I created in this context. One improves data coverage on steep surfaces in airborne- and drone-derived 3D lidar point clouds. The other reconstructs 3D surfaces between point cloud pairs to obtain measurements of volume change. Within this context, I examine channel migration rates and bank erosion volumes calculated with different approaches, highlighting the power of oblique data collection for mapping steep surfaces. By applying this approach at the Trinity, I show that bank surfaces erode after every flood, even floods that cause only a small change in river discharge. I then employ a graph theory-based analysis that pairs every sequential set of point clouds to examine differences between bank surface and channel edge movement over a wide range of time scales. When retreat magnitudes are small, bank surface erosion is larger than or comparable to channel edge movement, but this pattern reverses when retreat magnitudes are comparable to bank height. I find that erosion rates stabilize to within 25% of the equilibrium rate after ~2–3 years. Using only the drone datasets, I then measure channel profile evolution between floods and compare results to two controls on bank erosion: convolved channel centerline curvature and the fraction of the bank submerged during floods. I find that erosion magnitudes are well-correlated with predictions from the local bankfull channel geometry in both channel bends during the largest flood and the two second-largest floods, all of which are bankfull. Correlations decrease and become less consistent from flood to flood as well as from bend to bend during sub-bankfull floods. Results highlight the variability in bank erosion behavior at small erosion magnitudes and point to the influence of factors such as flood duration or hysteresis effects in controlling erosion patterns. These results provide a novel view into the variability of bank erosion behavior that aggregates to generate long-term channel migration. They also present a comprehensive example of how to treat analysis of a dense lidar time series. Future work includes comparing erosion patterns observed in my datasets to those in a river with different flood behavior to understand the fundamental processes controlling bank erosion.

Name thesis:degree_name

Doctor of Philosophy

Discipline thesis:degree_discipline

Geological Sciences

Grantor

The University of Texas at Austin

Year dc:date.issued

2025

Author dc:creator

• Nelson, Mariel D.

Advisors dc:contributor.advisor

• Goudge, Timothy A.
• Mohrig, David

Committee members dc:contributor.committeemember

• Joel Johnson
• Magruder, Lori A.
• Sylvester, Zoltán

• Surface reconstruction
• Lidar
• Sediment transport
• Bank erosion
• River evolution
• Topography
• Change detection

Language dc:language.iso

English

Identifier URI

https://doi.org/10.26153/tsw/63150

OAI identifier oai:identifier

oai:repositories.lib.utexas.edu:2152/135836