|Tile drainage simulation in SWAT2012: Parameterization and evaluation in an Indiana watershed
|Boles, C.M.W, J.R. Frankenberger and D.N. Moriasi
|Transactions of the ASABE
|URL (non-DOI journals):
|Broad Application Category:
|hydrologic and pollutant
|Primary Application Category:
|tile drainage effects and/or processes
|Secondary Application Category:
|nutrient cycling/loss and transport
|47 km^2 Matson Ditch, located in DeKalb County in northeast Indiana, U.S.
|Subsurface drainage is an important flow pathway in the poorly drained soils of the Midwestern U.S. and therefore needs to be included in modeling studies. The SWAT2012 tile drainage simulation method, based on the Hooghoudt and Kirkham equations with a drainage coefficient, shows promise for predicting drainage impacts in drained watersheds but has had limited application. In this study, SWAT2012 was implemented in a small agricultural watershed in Indiana using parameters based on knowledge of typical drainage systems and drainage design theory. Monthly Nash-Sutcliffe efficiency (NSE) values for streamflow exceeded 0.70 in all calibration and validation years but were lower for nutrients during some years, with values ranging between 0.14 and 0.88. Simulated tile flow compared well with measured values from field-scale studies in similar locations, ranging from 8.5% to 16.2% of annual precipitation when tile drains were implemented on 50.9% of the watershed. The depth to impermeable layer (DEP_IMP) parameter was found to be the most important calibration parameter, as it also controls seepage through the restrictive layer. Curve number values need to be substantially decreased for tile flow prediction, while two new tile drain parameters that control the ratio of lateral to vertical hydraulic conductivity (LATKSATF) and the static maximum surface roughness (SSTMAXD) likely need to be calibrated, as they could not be assigned values based on real-world properties. This study provides appropriate parameter values for the SWAT2012 tile drainage routine and demonstrates its ability to improve watershed-scale predictions in tile-drained watersheds.
|Hydrological modeling, Nitrogen, Phosphorus, SWAT, Tile drainage.