Dissertation - Open Access
Doctor of Philosophy (PhD)
William F. Lytle
The utilization of distributed input parameters enables hydrologists to more adequately cope with watershed management problems. A distributed system aids in evaluating alternative solutions designed to alleviate a problem situation. As a result, the objectives of this study were (1) to develop a hydrological framework delineating the surf ace configuration of a watershed to consider parameter distribution; (2) to develop a model of hydrologic performance of a watershed using existing techniques and amenable to the physical framework; and (3) to demonstrate the versatility of the model using hypothetical data. The model versatility was demonstrated by showing the effects of different time and areal rainfall intensity distributions on infiltration and runoff at different locations on the watershed surface and in the channel. The watershed framework was based on flow net theory. Contour lines on a watershed were considered to be equipotential lines. Streamlines were constructed normal to the contour lines, resulting in a collection of incremental units representing the natural drainage network of the watershed. Each incremental unit is assumed to be homogeneous and its size can be easily adjusted by selecting arbitrary streamline or contour line intervals. The size flexibility permits a more refined flow net for the more physiographically complex watersheds. The model computations for each unit starting with rainfall were accomplished using the Green and Ampt infiltration model for layered soils and the kinematic flow equations for routing overland and channel flow. The computations proceed from the watershed divide to the channel between adjacent streamlines utilizing input parameters for each incremental unit encountered. In this manner, hydrographs were calculated at any desired location on the watershed or in the channel. The parameters required for each incremental unit in addition to its geometry were the hydraulic roughness at the surface and the thickness, the bulk density, the hydraulic conductivity, the initial moisture condition, and the wetting front pressure head characteristic for each soil layer. Channel and reach geometry together with hydraulic roughness were the input required for channel routing. Methods were described whereby all the soil parameters can be measured. To demonstrate the model versatility, four different time distributions of rainfall were considered. They were constant intensity and three triangular patterns with the peak intensity occurring at the beginning, middle, and end of the storm. All the patterns had the same duration and produced the same volume of water. In general, the effects of these distributions on peak discharge were attenuated with distance from the watershed divide. However, hydrograph shapes were considerably different. The time to peak discharge and lag times for the hydrographs also varied with the type of distribution. Three different areal rainfall distributions were investigated. They were uniform coverage and high rainfall concentrations at the upper and lower ends of the watershed. Hydrograph discharges and shapes on the land were different largely because of differences in quantities of water. However, channel flow at the watershed outlet before and after peak discharge varied considerably indicating hydrograph shape differences.
Library of Congress Subject Headings
Number of Pages
South Dakota State University
Onstad, Charles Arnold, "An Incrementally Distributed Mathematical Model of a Watershed" (1972). Electronic Theses and Dissertations. 5450.