Submitted to: United States Committee of Irrigation and Drainage Engineering Conference
Publication Type: Proceedings
Publication Acceptance Date: 3/1/2003
Publication Date: 5/15/2003
Citation: Strelkoff, T., Bautista, E., Clemmens, A.J. 2003. Errors inherent in the simplified infiltration parameter estimation. United States Committee of Irrigation and Drainage Engineering Conference. p. 735-745. Interpretive Summary: In the competition among the agricultural, municipal, and environmental sectors for water, agriculture, as a major user, must use irrigation water prudently. Furthermore, with the potential for runoff and deep seepage to carry agricultural chemicals off site into the environment, the issues of irrigation design and management take on a critical importance. About half of the irrigated acreage in the U.S. is watered by surface means. In many developing countries, over 90% of agricultural lands are supplied by surface irrigation systems. Often, these systems do not live up to their potential for efficient water application, which, under appropriate field conditions, can rival that of pressurized systems. Computer models of the surface-irrigation process and associated design aids have been constructed to facilitate the search for conditions that provide optimum performance. A key piece of information needed to run this software defines how water infiltrates into the soil of a given field. This paper examines the built-in inaccuracies of two popular methods for predicting soil infiltration from measurements of water advance over the field. The work forms part of a charge to the Environmental and Water Resources Institute (ASCE) Task Committee on Soil and Crop Hydraulic Properties to review existing field-parameter estimation methods and make recommendations. The study is intended for application by research, consulting, and extension personnel for the benefit of both growers and the population at large.
Technical Abstract: An adaptation of the two-point method of estimating Kostiakov infiltration parameters to one measured advance point, for soils characterized by the Merriam and Clemmens time-rated families, is reviewed. Errors in parameter estimation resulting from inappropriate assumptions of normal depth at the upstream end of a border or furrow, constancy of the surface-stream shape factor, and the particular selection of Manning roughness and surface shape factor are calculated for a number of sample data sets in a range of soil types and system designs. In order to ensure controlled accuracy of error calculation, the surface-irrigation simulation model, SRFR, is used to generate advance data in place of field measurements. Both one and two-point methods are found to be sensitive to both shape factor and Manning roughness, especially at low field slopes.