Date of Award

1993

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Engineering Science (Interdepartmental Program)

First Advisor

Richard L. Bengtson

Abstract

The GLEAMS hydrology submodel was modified to account for the effects of shallow water table and subsurface drainage on runoff volume and soil loss. GLEAMS-WT (GLEAMS-Water Table) is a modified version of GLEAMS that accounts for shallow water table. GLEAMS-SWAT (GLEAMS with Subsurface drainage and WAter Table) is a modified version of GLEAMS that accounts for shallow water table and subsurface drainage. The simulation performances of GLEAMS, GLEAMS-WT, and GLEAMS-SWAT were evaluated by comparing their predictions with 7-years (1981-1987) of measured data from two plots of a runoff-erosion-drainage experiment at Baton Rouge, Louisiana. One plot was surface and subsurface drained, and the other was only surface drained. GLEAMS-WT predictions of surface runoff volume was very satisfactory. Total predicted surface runoff volume for 7-years was only 0.6 cm (0%) greater than the observed runoff volume from the non-subsurface drained plot, a significant improvement from the GLEAMS underprediction of surface runoff volume by 54%. GLEAMS-WT predictions of water table depth was satisfactory. GLEAMS-SWAT predictions of surface runoff volume, subsurface drainage volume, and water table depth were satisfactory. The total predicted surface runoff volume for 7-years was 6% less than the observed surface runoff volume from the subsurface drained plot, a significant improvement from the GLEAMS underprediction of surface runoff volume by 29%. The total predicted subsurface drainage volume was 1% less than the observed value. The erosion submodel of GLEAMS was linked with the two modified models. The GLEAMS, GLEAMS-WT, and GLEAMS-SWAT models seriously underestimated soil loss. An analysis was conducted to determine whether transport capacity or interrill + rill detachment was limiting soil loss prediction. It was found that soil loss prediction was limited by transport capacity. A calibration parameter was added to increase transport capacity. With calibration, the total soil loss predictions for both plots were satisfactory. However, for both models, the correlation coefficients relating calibrated predicted soil loss with the monthly observed soil loss were low. The regression slopes were significantly different and far from the ideal 1.0 slope. The transport capacity routine needs to be improved.

Pages

360

Share

COinS