Riverbank characteristics determine the pattern of riverbank collapse: Aiding early disaster risk identification

Identifying riverbank collapse risk is critical for early warning systems and mitigation strategies. This study conducted field investigations in the lower Yellow River (China) and employed the Bank Stability and Toe Erosion Model to quantitatively assess fluvial-scour-induced instability through limit equilibrium analysis. Based on the field investigations and in-situ experiments, five idealized riverbank profiles (3–5 m high, 1.3–3.0 m water levels), were simulated to analyze failure mechanisms via boundary conditions and toe deposit interactions. Results demonstrate that the outer angle (∠BCA) of triangular basal deposits and water-level fluctuations primarily control the buttressing effect of toe deposits on riverbank stability. This influence follows an inverted U-shaped relationship with ∠BCA, while increased water levels significantly weaken buttressing performance, reducing extended stability duration by 7–43%. The submerged slope ratio (SSR) is the predominant factor governing riverbank reshaping by toe erosion, with erosion volume and erosion displacement exhibiting a significant negative correlation with slope angle yet a positive correlation with water level ( R ² >0.71). Furthermore, toe erosion constitutes the primary mechanism of bank destabilization, accounting for most sediment loss during floods, whereas gravity erosion becomes comparatively more important during non-flood periods. Rain-induced reduction in soil shear strength significantly destabilizes riverbanks, with a more than 7% decrease in safety factor observed after repeated erosion simulations. Although rising water levels can partially mitigate this effect, higher hydraulic conditions ultimately lead to greater variability in bank stability, highlighting a complex interaction between soil moisture, hydraulic regime, and geotechnical properties. Risk thresholds identified SSR > 0.4 and convex/concave–convex morphologies as high-risk configurations, particularly during flood-rainfall periods. Proactive risk identification enables targeted prevention, significantly enhancing regional flood resilience.