Tuesday, November 14, 2006
200-5

Evaluation of Distributed Hydrology-Soil-Vegetation Model (DHSVM) For Simulating Tile Drained Landscapes.

Bibi Naz, Laura C. Bowling, Eileen J. Kladivko, and Keith Cherkauer. Purdue Univ, 915 W. State St, West Lafayette, IN 47907-2054

Although subsurface drainage systems improve field conditions by removing excess water from poorly drained soils, there are also concerns about the potential impacts of these systems on watershed hydrology and water quality. Unfortunately, the current generation of hydrologic simulation models either do not explicitly represent artificial subsurface drainage or are poorly suited to watershed scale applications.  The purpose of the study was to examine and improve upon the potential use of the Distributed Hydrology Soil-Vegetation Model (DHSVM) for predicting the hydrological response of tile drained landscapes. The model was applied to the subsurface drainage facility at the Southeast Purdue Agricultural Center (SEPAC), Butlerville, Indiana using observed drain flow and water table depth data from tile drains installed at 5, 10 and 20 m spacing. Following initial evaluation, DHSVM was modified to accurately describe the water table elevations, subsurface tile lines, deep seepage through a confining layer and seasonal crops. The model was calibrated from October 1986 to June 1990 and validated from January 1991 to December 1995. Most of the input parameters for soil physical properties were measured at the site, while hydraulic conductivities, exponential decrease in hydraulic conductivity and maximum infiltration rate were determined by calibration. The daily volume errors (dV) ranged from 5.13 to 0.46 and the coefficient of determination (D!) ranged from 0.6 to 0.9 for the calibration period. The shape and timings of observed and predicted hydrographs were well predicted, however, the model overestimated the drain flow volumes for most years. In general the model response appears reasonable and the results of this study document the potential of DHSVM for predicting streamflow in tile-drained watersheds.