Journal of Irrigation and Drainage Engineering

This report addresses the problem practicing engineers face as they try to evaluate the usefulness of watershed models for solving engineering problems. The report addresses the need for more complete parameter descriptions, unrealistic data needs, documentation, testing, and the lack of uniform criteria for evaluating a model's performance. The report makes recommendations for using some basic statistical measures to describe the performance of the models.

A new method for describing infiltration during filling and draining of cavities such as, for example, irrigation furrows is presented. The model development decomposes in a first step the two‐dimensional (2D) infiltration phenomenon into a series of one‐dimensional (1D) processes, which are described by an extended analytical solution of the 1D Richards equation (which incorporates the varying effect of gravity). A special integration over the wetted perimeter of the cavity includes rigorous description of the filling and draining mechanism of the furrow, thus allowing for accurate modeling of infiltration opportunity times. Then, two higher levels of approximation are still considered. Model version 2 describes with a constant geometry parameter (derived from the curvature of the cavity) the impact of the furrow shape on infiltration; and version 3 finally takes into account that the increasing wetted area around the furrow also affects infiltration. All three model versions employ only physically based parameters. The method decomposes the transient infiltration opportunity times; and the varying effect of gravity along the wetted perimeter are taken into account. The semianalytical character of the new approach avoids expensive and complex numerical solution procedures such as the finite‐difference or finite‐element methods, which recommends its use within a furrow‐irrigation simulation.

A physically based infiltration model describing two‐dimensional infiltration from cavities, for example irrigation furrows, is discussed. According to the different stages of model development, three versions with varying levels of sophistication are analyzed. These analyses include the investigation of short and long time behavior and the impact of fundamentally different soil types on the outcome of the models. The results show clearly that the curvature of the cavity has to be taken into account, which excludes the simplest model version from this kind of application. The other two model versions performed very well. Comparisons with the results of a laboratory experiment (together with simulations of all phases of furrow irrigation in different soils) show that up to considerable infiltration times the time‐dependent component of the shape parameter can be neglected. The new principle is recommended as an alternative to the empirical infiltration formulas presently used in numerical modeling of furrow irrigation, enhancing its computational efficiency.

The events leading to selenium contamination at the Kesterson Reservoir, a man‐made marsh in the Central Valley of California, began in the early 1900s, with the diversion of water to irrigate saline soils containing naturally high concentrations of selenium. With the advent of a systematic plan to collect and dispose of subsurface drainage water from these lands, problems associated with concentration and biological cycling of selenium became apparent at Kesterson, a series of shallow‐drainage‐water disposal ponds that was also managed as a wetland habitat. When chicks of hundreds of shorebirds failed to survive and deformed embryos were discovered, the California State Water Resources Control Board initiated a series of regulatory proceedings. This led to the closure of the Kesterson Reservoir and a series of remedial measures which went through numerous iterations as a team of scientists provided a more thorough understanding of processes leading to selenium transport and biologic exposure in this environment. In particular, field experiments and intensive monitoring demonstrated three factors leading to a cost‐effective resolution of the Kesterson cleanup. First, ground water under Kesterson was protected from selenium contamination by naturally occurring biogeochemical immobilization. Second, extensive monitoring of pore water in the vadose zone demonstrated that the favored soil‐excavation plan would be ineffective in reducing the elevated selenium concentration in ephemeral pools that occurs each winter at Kesterson. Finally, extensive biological monitoring demonstrated that selenium concentrations in the dominant species of upland vegetation at Kesterson was near or equal to “safe” levels. The combination of these factors led to a risk‐based remedial strategy that would leave the selenium inventory in place, and filling the low‐lying areas of Kesterson to prevent formation of ephemeral pools that attract wildlife. Observations that may be applicable to other areas experiencing similar environmental problems include the following: (1) Naturally occurring biogeochemical processes can be the cornerstone of effective remedial strategies; (2) although “safe levels” of toxic constituents based on total elemental analysis may be misleading and species‐specific targets may be more appropriate, seasonal cycles in speciation and analytical uncertainties make species‐specific goals difficult to define; and (3) scientific investigations and fast‐track remedial activities must go hand in hand to achieve cost‐effective, risk‐based remedial strategies.

Soil‐water distributions in homogeneous soil profiles of Yolo clay loam and Yolo sand irrigated from a circular source of water were measured at several times after initiation of irrigation. The effect of trickle discharge rates and soil type on the locations of the wetting front and soil‐water content distributions were determined. Soil‐water potential and hydraulic conductivity, as a function of soil‐water content, were also measured. A finite element solution of the two‐dimensional transient soil‐water equation, the theory of time‐dependent, linearized infiltration from a circular source, the effective hemisphere model, and a generalized solution for axially symmetric flow were compared with the experimental results. In general, the computed vertical advances of the wetting front were closely related to those observed for both soils. All of the solutions provided a better prediction of the wetting front positions for the clay‐loam soil than for the sandy soil. The calculated and measured horizontal wetting fronts did not agree for large time, especially for the generalized solution. A more pronounced underprediction of the wetting front positions by the effective hemisphere theory occurred compared to those predicted by the linearized and finite element theories. Soil‐water content distributions computed by linearized and numerical solutions, agreed reasonably with measured values in both soils.

The treatment, reuse, and disposal of agricultural drainage water present formidable technical problems in the management of salts and associated trace elements. Although many promising treatment technologies and disposal methods have been identified and tested in minipilot plants, they need to be further developed before technical and economic feasibility can be established for full‐scale applications. Reuse of drainage is possible with salt‐tolerant crops, but has yet to be proven in large‐scale operations. Over time, salt buildup in soil system can be a problem. Brine recovery in solar gradient ponds for the generation of electrical power and salt harvesting for reclamation show promise both in theory and in pilot demonstration, but the economics of these integrated processes have yet to be proven for large‐scale operations. The technologies and economics of treatment, reuse, and disposal methods will require further research and development before they can be efficiently applied to solving drainage problems.

The concept of treating soil matrix as a bundle of capillary tubes was used extensively in the analytical description of hydraulic conductivity of unsaturated soils. Such a concept has seldom been used in the development of infiltration models. The advantage of considering soil matrix as a collection of capillary tubes is that the water flow in large tubes can be separated from the flow in small tubes. Such a separation is desirable in the study of soil macropore flow where the flow in large pores is primarily responsible for ground‐water contamination and recharge. The purpose of this paper is to derive an infiltration model based on the concept of a bundle of capillary tubes for the study of soil macropore flow. The model presented a spatially varied wetting front and was an improvement over the Green‐Ampt infiltration model, which predicted a constant depth front. The soil macropore flow was represented graphically as a finger on a depth‐water content plot.

Concentrations, distribution, and sources of selenium from irrigated lands were studied between 1986 and 1990 at 20 reconnaissance project areas in 17 western states under the Department of Interior's National Irrigation Water Quality Program. Samples of water, bottom sediment, whole‐body fish, and bird livers for analysis of selenium concentrations were collected before, during, and after irrigation season from streams, canals, lakes, and ground water in each project area. Selenium concentrations in water ranged from less than the detection limit of 1 μg/L in 42% of the 586 samples collected to 4,800 μg/L from a well in the Pine River Area, in southern Colorado. Selenium concentrations in 223 samples of bottom sediment ranged from less than the detection limit of 0.1 μg/g to 85 μg/g in a sample from the Middle Green River Basin, in Utah. Selenium concentrations in whole‐body fish (all species) ranged from 0.1 μg to 50 μg per gram dry weight, with the maximum concentration observed in a carp from the Gunnison River Basin, in western Colorado. Selenium concentrations in bird livers (all species) ranged from less than 0.32 μg to 170 μg per gram dry weight with the maximum concentration observed in the liver of an avocet from the Kendrick Reclamation Project, in Wyoming. Cretaceous‐age marine shales probably are the original sources for selenium in 16 of the 20 areas studied. In several areas, evaporative concentration of applied irrigation water and drainage of this water to wetlands, canals, streams, and lakes appears to be responsible for elevated concentrations of selenium.

Subsurface drainage systems in regions with poor‐quality, shallow ground water collect water that potentially may have adverse affects on the environment. In semiarid regions, the poor‐quality ground water may result from evaporative salinization, leaching of the soil root zone from excess irrigation, or some combination of these two processes. Traditional subsurface drainage system design procedures do not consider the water‐quality aspects associated with the depth and spacing of drains in a particular hydrogeologic setting. Typically, deep drains at relatively wide spacings tend to collect a larger fraction of deep ground water as compared to shallow ground water. In many cases, the deep ground water is of poorer quality. The purpose of this study is to illustrate and qualitatively describe the results of numerical simulations that examine the influence of drain spacing and depth on drain water quality. This paper considers steady and transient recharge to the water table in a layered anisotropic soil. Results of this and subsequent analyses can be used to propose conceptual modifications of traditional drainage system design for water quality control in irrigated regions.

Nitrate leaching was measured under no‐till and conventional tillage irrigated corn on a sandy loam soil. Ground water was sampled monthly for two years. Soil samples were taken three or four times a year to a depth of 150 cm. Nitrate concentrations were above 10 mg/L for most sampling dates. Most of the nitrogen was leached during the fall and winter months. The mass of nitrate leached from the no‐till plots ranged 55.0 to 78.8 kg/ha/yr. The mass of nitrate leached from the conventional plots ranged from 57.1 to 94.0 kg/ha/yr. The mass of nitrate leached was slightly higher under conventional tillage than no‐till. Nitrate leaching would be difficult to control, because most of the nitrogen is leached during the late fall and winter. Proper nitrogen applications for realistic yield goals is probably the best method for controlling nitrate leaching.

Analytical equations that characterize hydraulics of center‐pivot laterals with and without end‐gun sprinklers were developed. The equations include computation of pressure‐head distribution; pressure‐head loss; and maximum, minimum, and average pressure heads and their corresponding locations of occurrence along a lateral. The proposed equations are important in design and evaluation of center‐pivot irrigation systems since critical reference pressures can be easily determined and their corresponding locations can be readily identified in the field. The velocity‐head term was included in the complete set of developed equations. A simplified solution that neglects the velocity‐head term and uses the Darcy‐Weisbach formula for computing friction loss in a typical center‐pivot lateral with an end‐gun sprinkler is also included. The results are in good agreement to those computed by a stepwise solution in a variety of test conditions. Discrepancies between results from these methods are within the accuracies of pressure gages normally used in irrigation evaluations. The proposed equations simplify important practical applications, such as system design and evaluation. Besides design, other potential applications of the proposed equations include analysis of energy consumption and uniformity of water application due to hydraulic behavior under different design and operating conditions.

Principle considerations in examining water‐management options for the control of toxic trace elements in agricultural drainage waters are addressed, including a number of improvements in off‐farm water delivery systems, as well as efficiency of on‐farm irrigation systems and water reuse. Potentials exist to improve off‐farm water‐delivery systems through reductions in seepage losses with canal linings, operational spills through systems irrigation scheduling, and interception of water losses and reuse. Potentials also exist in improving on‐farm water management through improved irrigation application systems, on‐farm irrigation scheduling, and other cultural products such as changes in cropping patterns. These on‐farm management options will affect net farm income due to increased costs of equipment and structures, increased labor requirements, and changes in crop revenue. But not all of these will have a negative economic input. For instance, efficient irrigation systems not only value drainage but tend to increase crop yield or decrease crop production inputs. Installation of structural system improvements and use of system scheduling of water delivery may result in decrease in operation and maintenance costs. Institutional and legal constraints, particularly water‐right ownership, may constrain efforts to minimize drain‐water production and toxic‐element problems. The long‐term economic viability of irrigated agriculture will require careful consideration of both off‐farm and on‐farm management options.

Concern over the potentially toxic effects of trace elements in agricultural drainage waters has been aroused by the discovery of selenium poisoning of wildlife at Kesterson Reservoir, California, and was heightened further through discovery of elevated levels of selenium in shallow ground water underlying extensive areas of San Joaquin Valley's west side. Recent investigations by the U.S. Department of the Interior's National Irrigation Water Quality Program (NIWQP) revealed that similar trace‐element problems exist in several other western states including Nevada, Utah, Wyoming, and Colorado, as well as the Tulare Lake Bed and Salton Sea in California. Management options assessed for trace elements discharged from irrigated lands include source control, drain‐water reuse, drain‐water treatment, and removal of contaminants, disposal, and institutional and jurisdictional control measures. A combination of source control and other drainage‐water‐management options has the potential to reduce the toxic‐element problem. An initial prognosis indicates that a status quo scenario will be unacceptable to the public. Agriculture will be increasingly challenged in its use of water and land resources and the perceived impacts on the quality of the environment. The effect of drainage reduction and other management options on reducing the discharge of trace elements is, to some extent, influenced by site‐specific conditions. The economic viability of agriculture will be severely tested in the most severely trace‐element‐impacted lands, water, and biota, and this may lead to changes in land use.

This study was conducted to examine the entrance resistance and head losses in the immediate vicinity of subsurface drains and to measure the change with time just after drain installation. Laboratory experiments were conducted on drain tubes and sandy loam soil excavated from the drain depth. Tubes were tested both with and without a thin fabric envelope. Theoretical and approximate methods of quantifying entrance resistance and head losses were compared to measured results. Changes in hydraulic conductivity, head loss, and entrance resistance in the immediate vicinity of the drains were examined using laboratory soil tanks. The hydraulic conductivity decreased over time. The effective radius, $r_e$, for the tube without a fabric envelope increased over the test period. The two drains equipped with fabric envelopes had fairly constant $r_e$ values throughout the test period. The resistance factor, $R$, for the tube with no envelope was not constant. The value of $R$ increased throughout the test period for the tubes with fabric envelopes. Experimentally determined entrance constant values were initially higher than the theoretical values and then decreased below the theoretical values with time.

The aquifers beneath irrigated lands in the southwestern United States are composed of alluvial deposits that, together with subsurface geologic conditions and water‐level elevations, largely determine the movement of deep‐percolation water below the water table. Significant factors in determining the rate and direction of ground‐water flow are the layering of alluvium and the contrast between the horizontal permeability of coarse‐grained deposits and the vertical permeability of interbedded, fine‐grained deposits, or confining beds. Once the physical or hydraulic component of the ground‐water‐flow system is understood, knowledge of geochemical processes can be used to predict the mobility of various trace inorganic chemical constituents in deep‐percolation water. Some of these constituents can move at the same rate as the ground water itself, while others are relatively immobile. The trace inorganic constituents of greatest concern relative to water‐quality standards are derived from natural geologic deposits, and most are present at oxyanions. These constituents have been mobilized by irrigation practices in some localities and are important in terms of ground‐water or surface‐water uses in the alluvial basins.