Mechanics of River Erosion and its Effects on Floodplain Biogeochemistry

Citation

Douglas, Madison Marie (2023) Mechanics of River Erosion and its Effects on Floodplain Biogeochemistry. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/rwmn-hq80. https://resolver.caltech.edu/CaltechTHESIS:06012023-210614776

Abstract

Rivers transport water, sediment, and nutrients across Earth’s surface. They shape landscapes, eroding mountain ranges and building floodplains, simultaneously providing important resources and posing a hazard to nearby communities. Here, I present field work, flume experiments, numerical models, and laboratory analyses addressing three main themes: permafrost river and floodplain dynamics, river meandering without plants, and rates of bedrock incision. Arctic rivers migrate rapidly across their floodplains and their migration rates are predicted to increase as permafrost thaws due to climate change. However, no mechanistic model is capable of predicting permafrost riverbank annual erosion rates. To address this knowledge gap, I developed a calibrated numerical model for permafrost riverbank erosion. A previously published theory assumes that permafrost erosion rates are limited by pore-ice thaw, but underestimates thaw rates due to bank roughness increasing heat transfer from the river to its banks (Chapter 3). Results indicate that thaw-limited erosion is orders of magnitude higher than observed erosion rates, and permafrost riverbank erosion must instead be limited by sediment entrainment and the collapse of overhanging blocks to match observed rates (Chapter 2). Based on experimental results, I developed a 1D numerical model that includes roughness-dependent permafrost thaw and sediment entrainment and tracks how heat transfer within the riverbank can form a thawed layer (Chapter 4). Results indicate that permafrost riverbank erosion rates respond to changes in river discharge due to climate change, which affect both bank thaw and entrainment rates, and are only sensitive to changes in water temperature via thawed layer failure. As a case study, I conducted fieldwork along the Koyukuk River in Alaska, which is located in discontinuous permafrost. I found that changes in riverbank erosion rates may more rapidly erase permafrost from floodplains (Chapter 7) and change the spatial patterns of floodplain methane emissions (Chapter 5). While riverbank erosion releases eroded organic carbon to be oxidized as greenhouse gases or transported downstream, a portion of this carbon is re-deposited in the floodplain, modulating the effects of river migration on regional carbon cycling (Chapter 6). To understand the effects of vegetation on river migration rates and fluvial stratigraphy, I conducted long-term monitoring of the unvegetated, ephemeral Amargosa River in Death Valley, California (Chapter 8). This study found that the Amargosa is actively meandering at very slow rates and frequently avulses, producing muddy stratigraphy with isolated sand bodies that is thought to be unique to vegetated meandering rivers. Sediment transport has also been proposed as a primary control on bedrock river incision rates, where saltating grains gradually abrade the channel bed over geologic timescales. However, uncertainty about long-term sediment supply and the frequency of floods that cause significant bedrock incision has prevented using saltation-abrasion to model landscape evolution. Using a global data compilation, I calculated a best-fit sediment supply-normalized flood intermittency parameter so that the saltation-abrasion model can be broadly applied (Chapter 9). Together, these studies advance understanding of how riverine sedimentary transport governs permafrost riverbank erosion, Arctic floodplain biogeochemistry, stratigraphic deposits of unvegetated rivers, and bedrock incision rates.

Thesis Files