Sensitivity to grid and time resolution of hydrology components of DANSAT

A sensitivity analysis of the Dynamic Agricultural Nonpoint Source Assessment Tool (DANSAT) model to different grid sizes and time steps was conducted to investigate scale effects on hydrology and to provide users with a guideline for selecting an appropriate grid size and time step in order to enhance computational time efficiency. Response of the hydrology components to different grid sizes was analyzed by considering: (1) changes in input parameter values due to GIS manipulation processes, and (2) comprehensive response of the model through applications to a small agricultural subwatershed (QN2) in the Nomini Creek watershed, Virginia. In addition, model response to different time steps was analyzed in QN2 by changing the storm event time step (SET) of the model (1, 5, 10, 15, 30, 45, and 60 min) against a fixed grid size. A maximum acceptable grid size (MAG) of 90 m was selected for QN2, considering changes in the accuracy of spatial data for different grid sizes. Only the overall response of the hydrology components to down-scaled (60 m) grid size in QN2 was acceptable without any further calibration. Daily streamflow for storm events decreased with increases in time step from 1 to 60 min, while total runoff for the simulation period increased slightly by 8%. Use of MAG (90 m) with an acceptable larger time step (10 min) based on monthly runoff criteria exponentially reduced computational time compared to an application using the smallest grid size and time step. Site-specific sensitivity analysis is recommended due to the possible differences in response of the hydrology components to watersheds with different hydrologic characteristics.