Soil Water Extremes and Streambank Erosion: A Comprehensive Study of Erosion Dynamics and Predictive Modeling

Streambank erosion processes are influenced by complex interactions between soil moisture dynamics, erosion resistance, and hydraulic forces. This dissertation investigates these interrelationships through three interlinked studies that advance both the understanding and modeling of erosion under temporally and spatially variable hydrologic conditions. The first study explores the spatiotemporal variability of soil moisture extremes across geophysically distinct watersheds in Tennessee. Using projected soil moisture, it identifies how physiographic factors and hydroclimatic variability jointly influence the occurrence of hydrologic extremes. The second study quantifies the effect of soil moisture on key erosion resistance parameters—critical shear stress (τ_c) and erodibility coefficient (K_d), through extensive Jet Erosion Tests (JET) under varying seasonal moisture conditions. Quadratic relationships are established, revealing how low to high soil moisture transitions impact bank resistance and erosion susceptibility. The third study integrates these dynamic soil moisture relationships into the Bank Stability and Toe Erosion Model (BSTEM), creating a modified framework where τ_c and K_d evolve over time in response to observed moisture conditions. This dynamic BSTEM model is evaluated across three natural streambanks and compared with traditional static configurations. Statistical comparisons demonstrate that the dynamic model captures the timing and magnitude of observed lateral retreat accurately, particularly at sites where suitable Manning’s n values align with field conditions. While the dynamic model does not universally outperform static configurations, its ability to respond to temporal transitions in erosion behavior offers clear advantages, especially during rapid erosion transitions. Together, these studies demonstrate that integrating time-varying hydrologic drivers with erosion resistance improves both conceptual understanding and predictive accuracy in streambank erosion modeling.