Precipitation-induced overland and groundwater flow and mixing processes are quantified to analyze the temporal (event and pre-event water) and spatial (groundwater discharge and overland runoff) origins of water entering a stream. Using a distributed-parameter control volume finite-element simulator that can simultaneously solve the fully coupled partial differential equations describing 2-D Manning and 3-D Darcian flow and advective-dispersive transport, mechanical flow (driven by hydraulic potential) and tracer-based hydrograph separation (driven by dispersive mixing as well as mechanical flow) are simulated in response to precipitation events in two cross sections oriented parallel and perpendicular to a stream. The results indicate that as precipitation becomes more intense, the subsurface mechanical flow contributions tend to become less significant relative to the total pre-event stream discharge. Hydrodynamic mixing can play an important role in enhancing pre-event tracer signals in the stream. This implies that temporally tagged chemical signals introduced into surface-subsurface flow systems from precipitation may not be strong enough to detect the changes in the subsurface flow system. It is concluded that diffusive/dispersive mixing, capillary fringe groundwater ridging, and macropore flow can influence the temporal sources of water in the stream, but any sole mechanism may not fully explain the strong pre-event water discharge. Further investigations of the influence of heterogeneity, residence time, geomorphology, and root zone processes are required to confirm the conclusions of this study. Copyright 2011 by the American Geophysical Union.