Predicting passive dispersal using particle hitting distance estimates in the near field in rivers

This is the authors' abstract. We don't add key points for this paper.

Particles, including biologically relevant ones, released in rivers may interact with the bed multiple times throughout their dispersal downstream. Passive dispersal is influenced by local variations in bed morphometry and hydrodynamics, especially in the near field—the area downstream of a point source where complete mixing across the river's width has not yet occurred. We examined predictors (e.g., depth [], shear velocity [], velocity at 50% ) of the distance at which particles first contact the bed (hitting distance) using a theoretical model of turbulent transport. We identified as the strongest predictor of hitting distance and tested how different estimates of affected model predictions. We aimed to determine whether increased sampling effort (e.g., multiple estimates in a reach) improves the accuracy of the model. Specifically, we compared a reach‐based derived from the depth–slope product (DSP) vs. local estimates based on velocity measurements at 50% H , calculated using the turbulent kinetic energy method. We evaluated predictions against empirical data from three river reaches (Conestogo, Grand, and Speed rivers) with different hydrodynamic conditions (e.g., H < 50 cm; 2 < < 5 cm s −1 ). Model predictions based on reach‐averaged or single values (RMSE local = 2–3 m) were more accurate than DSP‐derived estimates (RMSE DSP = 3–7 m) and an implementation that accounts for longitudinal variations in local (RMSE modelled = 1–8 m). Our results suggest that accounting for variations in bed morphometry and hydrodynamics improves predictions of particle transport and dispersal at reach scales.