A meso-scale gravel tracer model for large gravel-bed rivers

A simplified numerical gravel tracer model is presented that utilizes the flow field calculated by a three-dimensional hydrodynamic model. It operates on the meso-scale, predicting pathways and velocities of gravel tracers along several kilometres of river length for durations of weeks to months. The model simplifies tracer shape as an equipotential ellipsoid, assumes constant tracer movement and incorporates rest periods into a transport resistance coefficient. It considers local bed slope and performs a random walk based on the near-bed turbulent kinetic energy. The deterministic shear stress value of the hydrodynamic model is permuted by a random variable, for which both a Gaussian and a log-normal probability density function were tested. The model was calibrated on gravel tracer positions monitored at the Danube River. Validation was conducted for another discharge and an unsteady flood wave. Both paths and velocities of gravel tracers observed in the field could be reproduced by the model.