Could DHI elaborate on how reverse particle tracking is implemented for transient flow simulations.  For example how does it incorporate the different velocity fields as they change during a simulation.  Are particle positions reset at each time step?