Michael H. Gardner

Predicting thresholds of sediment motion is critical for a range of applications involving sediment transport. However, thresholds for sediment motion can vary over an order of magnitude for a single characteristic flow and bed configuration. Lacking simple ways to incorporate this variability, many assume thresholds are constant for rough, turbulent flow. Here, we quantify variability of incipient‐motion thresholds based on a commonly used grain‐scale force‐balance model, with model parameter distributions determined from published experiments. We show that variability in the threshold of motion within the 2D force‐balance model occurs predominantly due to variability in the lift coefficient and grain protrusion, and secondarily due to drag coefficient variability. For a known grain size, the mean threshold of motion, and variability about the mean, can be predicted from a family of power laws. These power laws can be altered with site‐specific parameter distributions, allowing for site‐specific application to well‐studied reaches and other planets. Using compiled flume and field data we show that constraining force‐balance parameter distributions with independent data results in narrower distributions of the predicted threshold of motion, consistent with constrained flume experiments. This analysis highlights that while the threshold of sediment motion is variable, the magnitude of variability is predictable within the force‐balance model based on site‐specific physical constraints of local flow and bed conditions.