|
" Evaluation of surface runoff water quality prediction under different water table management practices "
A. T. Mohammad
R. W. P. Skaggs, John E.
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1113273
|
| Doc. No
|
:
|
TLpq304378098
|
| Main Entry
|
:
|
A. T. Mohammad
|
|
|
:
|
R. W. P. Skaggs, John E.
|
| Title & Author
|
:
|
Evaluation of surface runoff water quality prediction under different water table management practices\ A. T. MohammadR. W. P. Skaggs, John E.
|
| College
|
:
|
North Carolina State University
|
| Date
|
:
|
1997
|
| student score
|
:
|
1997
|
| Degree
|
:
|
Ph.D.
|
| Page No
|
:
|
172
|
| Abstract
|
:
|
Data from a water table management research project near Plymouth, NC were collected and used to test the combined DRAINMOD-CREAMS model. The combined model was tested in predicting surface runoff as well as sediment and nutrient (nitrogen and phosphorus) carried by surface runoff under different water table management (WTM) practices, including conventional drainage (D), controlled drainage (CD) and subirrigation (SI). Surface runoff and subsurface volumes and water table depths were measured and compared to model simulated results. The comparisons showed that DRAINMOD performed well in predicting hydrologic responses under different water table management practices. In general, DRAINMOD overpredicted surface runoff by 5%. Runoff was overpredicted by 22% for SI and overpredicted by 49% for CD condition. Under D, runoff was underpredicted by 5%. Runoff was overpredicted by 1% for graded subplots and overpredicted by 9% for ungraded subplots. Analysis of storm events showed that water table depth prior to a storm affected runoff. The water table depth prior to the storm was affected by water table management practices and by previous rainfall and evapotranspiration conditions. From hydrologic evaluation of individual storm events, it is clear that the water table depth before each storm affected the amount of runoff and this varied considerably across WTM treatments. In predicting soil loss, the results showed that the CREAMS erosion submodel was very sensitive to changes in slopes in the experimental plots. The submodel overpredicted soil loss on graded plots and underpredicted soil loss on ungraded plots. Peak runoff rate, one of important storm characteristics affecting soil loss, was better estimated by SCS triangular hydrograph method than maximum hourly runoff and CREAMS regression equation. The DRAINMOD-CREAMS model overpredicted soil loss for the plots under SI and CD practices. This resulted from overprediction of surface runoff by the DRAINMOD model. In predicting total nitrogen (TN) and total phosphorus (TP), the model overpredicted cumulative TN and TP losses from graded plots and underpredicted TN and TP losses from ungraded subplots. Using measured runoff improved TN prediction only slightly. Using both measured runoff and measured soil loss, improved TN predictions some (r2 = 0.5 to 0.6 and underpredicted TN loss by 44% for CD and 16% for SI). Model performance in predicting total phosphorus (TP) was also poor with r2 values ranged from 0.16 to 0.66 and large overpredicted for CD and SI. In contrast to TN, however, model predictions improved significantly when measured values for surface runoff and soil loss were used in the CREAMS submodel. In this case, r2 values ranged from 0.55 to 0.87 and slightly overpredicted for CD (3%) and SI (14%) and underpredicted for D (9%). (Abstract shortened by UMI.)
|
| Subject
|
:
|
Applied sciences
|
|
|
:
|
Earth sciences
|
|
|
:
|
Health and environmental sciences
|
|
|
:
|
nitrogen loss
|
|
|
:
|
phosphorus loss
|
|
|
:
|
soil loss
|
| |