Research papersPredicting the temporal variation of flow contributing areas using SWAT

- Temporal variation of flow contributing areas during each rainstorm was predicted.
- Continuous changes in soil water contents predicted by a current precipitation index.
- SWAT model over-predicts the flow contributing area for events with lower rainfall.
- SWAT closely simulates flow contributing area for higher rainfall events.

This study assessed the capability of soil and water assessment tool (SWAT) to identify areas contributing to flow in the Gully Creek Watershed in Ontario. The SWAT model predicted the streamflow at the outlet of the watershed, with monthly and daily Nash-Sutcliffe efficiencies of 0.75 and 0.60 during the validation period. In addition to the daily streamflow data, the flow was also observed at 16 monitoring stations during 6 different events. The validated model was then used to simulate flow at the monitoring stations. The effect of watershed delineation on streamflow and events at 16 monitoring station were then examined by SWAT. The delineation of 99 subbasins, with highest efficiency was selected for the purpose of predicting the potential flow contributing areas with the model. Overall, the flow events were overestimated by SWAT. Temporal variations in the potential flow contributing areas during each event were then analyzed. Flow contributing areas during each event was predicted by the model first and the results showed a good agreement with available information. A current precipitation index was used to simulate the continuous change of soil water content during each rainstorm, and the modeling results of the individual events were used to explore the capability of the model to predict the temporal variation of flow contributing areas during each event. The results revealed that the SWAT model over-predicted the areas contributing flow for events with lower rainfall; while for the events with higher rainfall amount the model closely simulated the time-varying contributing area. The results of this study provide some insight into the possible capability of SWAT model to predict the temporal variations in potential contributing area, and therefore provide an important contribution to the modeling of runoff generation in watersheds, a vital aspect in the evaluation and planning of best management practices.