Calibrated Granular‐Flow PIV From DEM Simulations: Flow‐Substrate Dynamics in Flows With Fixed and Erodible Substrates, Modeling, and Field Implications

Granular flows are central to geophysical and industrial processes, yet their internal properties remain difficult to quantify. Understanding how energy and momentum are exchanged at the flow–substrate boundary is key to predicting their erosion and mobility. Here, we assess the accuracy of particle image velocimetry (PIV) in resolving velocity and granular temperature () in analog granular flows using like-for-like comparisons with discrete element method simulations. Synthetic image sequences from simulations reveal that PIV systematically underestimates by ∼34% owing to Eulerian spatial averaging. Applying this correction factor enables accurate quantification of within stated uncertainties. This calibrated approach integrates laboratory, numerical, and field perspectives, offering new constraints on granular rheology and flow–bed coupling. Applying this calibrated workflow to analog flows over fixed and erodible beds reveals fundamental behavioral contrasts. Fixed rough beds exhibit basal spikes caused by intense shear and grain interlocking, whereas erodible substrates transmit momentum and into the substrate, producing velocity profiles that rise exponentially within the substrate. The flow–substrate interface is dynamic, oscillating and alternating erosion and deposition. These results identify as a key link between local and non-local rheology: agitation within the flow can propagate, weaken the substrate, and control entrainment. Frame-by-frame velocity-based boundary tracking reveals that standard erosion–deposition laws in depth-averaged models fail to reproduce even simple experiments, highlighting the need for revision, especially for steep, mobile beds. Findings allow the observation and tracking of the fluctuating flow-substrate interface observed in the deposits of many geophysical flows.