Submitted to: US Committee on Irrigation and Drainage/Environmental and Water Resources Institute Conference
Publication Type: Proceedings
Publication Acceptance Date: March 1, 2003
Publication Date: July 13, 2003
Citation: Bautista, E., Strelkoff, T., Clemmens, A.J. 2003. Sensitivity of surface irrigation to infiltration parameters: implications for management. US Committee on Irrigation and Drainage/Environmental and Water Resources Institute Conference. p. 475-485. Interpretive Summary: Nearly 50 percent of the irrigated farmland in the U.S. uses surface irrigation methods. Improving the design and management of such systems can foster water conservation and also help reduce water quality problems caused by agricultural runoff and deep percolation. Design and management alternatives can be analyzed with the aid of Irrigation computer models. Present use of modeling tools is limited, however, largely because key model inputs, in particular infiltration, are difficult to measure in the field and exhibit great variability. Because actual infiltration characteristics can depart significantly from the design specifications, modern surface irrigation system designs need to include a well-defined strategy for adjusting system management to compensate for the effect of these uncertain data inputs. This paper outlines such a strategy for a particular system. Such an analysis should be of value to irrigation specialists and farmers with knowledge of surface irrigation hydraulic models.
Technical Abstract: Infiltration characteristics are a major source of uncertainty in the design and management of surface irrigation systems. Understanding the sensitivity of the design to errors or variation in the design inputs is needed to develop management recommendations that account for this uncertainty. This paper further analyzes the sensitivity of the level basin design procedure proposed by Clemmens (1998). Results show that the recommended management approach, cutting off inflow when the water advances a fixed distance relative to the field length, works best when actual advance time is more than predicted. If actual advance time is the same or less than predicted, then cutoff based on time may be a better approach, independent from variations due to differences in infiltration, roughness, inflow, or all of these factors combined.