Search

Drainage is an essential land-management practice for growing crops in the Lower Mississippi Valley and Gulf coastal areas. Current crop production and water-management methods pose significant pot...

Simulation performances of GLEAMS, GLEAMS-WT, and GLEAMS-SWAT were evaluated by comparing their soil loss predictions with measured data from two runoff-erosion-drainage experimental plots at Baton Rouge, Louisiana, One of the experimental plots was surface drained only, and the other was both surface and subsurface drained. Although the hydrology components of GLEAMS-WT and GLEAMS-SWAT predicted surface runoff more accurately than the original GLEAMS, all three models seriously underpredicted total soil losses over a seven-year period (1981 to 1987). Transport capacity limited soil loss prediction values in the models. Hence, we recommend that any changes or modifications in the erosion submodel be focused on improving transport capacity simulation; changes in the detachment simulation routine may not be needed. A calibration parameter was added to the erosion subroutine to adjust transport capacity. However, even when the models were calibrated for a specific site, there were still substantial annual and monthly differences between predicted and observed soil losses. Keywords. GLEAMS, Models, Runoff, Erosion.

Atrazine and metolachlor were applied preemergent on 22 April 1987, to silage corn on high watertable fields which were either subsurface drained (104 mm drain tubes 1 m deep) or surface drained only. Surface runoff and subsurface outflow were sampled and analyzed for atrazine and metolachlor from 22 April to 31 August 1987. Subsurface drainage reduced atrazine and metolachlor losses by 55 and 51%, respectively. Two-thirds of the losses occurred within 30 days after the application of the herbicides, and most of the losses were in the surface runoff. The edge-of-field average atrazine concentrations, 12.5 and 9.0 ug/1, for the non-drained and drained fields, respectively, were four and three times greater than the U.S. Environmental Protection Agency advisory of 3 ug/1 for drinking water. The edge-of-field average metolachlor concentrations were near the U.S. EPA advisory of 10 ug/1.

A sub surface drainage experiment on clay loam soil near Baton Rouge, Louisiana, was conducted from 1980-1983. The expeiment consisted of four plots, of which two were both surface and subsurface drained. Corn silage was grown each year. Average yields from the subsurface drained plots were 84% higher than the yields from plots with surface drainage only. The growing season high water table stress as quantified by SEW30 had a highly significant affect on corn silage yields.

Fluctuations in depth of the water table midway between subsurface-drainage/subirrigation conduits in a sump-controlled water table management system were predicted within an average deviation of 8 cm in DRAINMOD simulations conducted for a Commerce silt loam soil in the Mississippi Delta. The average daily sump water level from a field experiment was used as a model input to establish changes in the drainage outlet water level boundary conditions.

SIX years of data from erosion plots at Holly Springs, MS, and 10 yr of data from plots at Morris, MN, were used to study variation in soil erodibility through the year. Monthly values of soil erodibility were related to time with a cosine curve. Erodibility varied from a high of 169% of annual average K (K of the Universal Soil Loss Equation) on February 4 to a low of 31% of annual average on August 5 for the Mississippi data. Minnesota erodibility data were described similarly except the maximum and minimum of the erodibility function lagged the Mississippi curve by about 2 months. It was suggested that erodibility could be predicted using a variability function based on normal air temperature data described by a cosine function and average annual K factor values.