In developing management practices to reduce chemical leaching below the root zone, tillage and irrigation management are important considerations. Two studies were performed to evaluate the movement of bromide in tilled and non-tilled soils under sprinkler versus flood irrigation. In each study, bromide was applied either with an irrigation or presprayed to the soil surface followed by periodic soil sampling to monitor the bromide movement. Tillage was observed to reduce the mean depth of chemical penetration under both irrigation treatments and reduce the spatial variation of bromide concentration under flood irrigation. For example, after 30 days of periodic flood irrigation, 25% of the applied bromide remained in the upper 0.2m of a tilled soil while in the companion non-tilled soil virtually no bromide remained above this depth. The most rapid bromide movement was observed in non-tilled, flood irrigated soil, particularly when the solute was added with the irrigation. We speculate that the tillage effect of reduced leaching results from the alteration of pore continuity and creation of diffusional sinks and not increased evaporative water loss in the tilled soil. The Root Zone Water Quality Model was calibrated using site-specific hydraulic property measurements and used to predict the solute movement. The model predictions were fairly accurate for the sprinkler irrigated soil but less satisfactory for the flood irrigation studies. In comparing the effect on chemical leaching of the treatments imposed, we found that tillage and the timing of the chemical application had greater impact on reducing leaching than did the method of irrigation.

Experimental validation of solute transport models is often met with unsatisfying or ambiguous results. Aside from the inherent difficulty in conducting these experiments, the lack of a standard set of methodologies employed in these studies may be an impediment. This study was conducted to determine if the method employed in the sampling of soil solutes could influence the results of model validation attempts. A narrow pulse of KBr was uniformly applied to two 25-m(2) plots, one tilled and the other undisturbed, on a level Ulm clay loam (fine, montmorillonitic, mesic Ustollic Haplargid). Solute movement was then monitored for the next 388 d using replicated ceramic cup solution samplers at depths of 0.15, 0.30, 0.80, 1.20, 1.60, and 2.00 m and by the periodic removal of replicate soil cores in 0.05-m increments to a maximum depth of 3.70 m. Overall, the methods yielded consistent information regarding the mean convective transport of Br-. Both methodologies indicated deeper and more rapid movement of Br- in the nontilled soil, although the shallow solution samplers appear to miss the most rapidly moving solute. By the conclusion of the experiment, the difference between the centers of mass of the Br - plumes in the tilled and nontilled soil was at least 0.6 m. The soil core and solution sampler methods appear to yield inconsistent measures of the solute dispersion but we attribute the differences to the effects of incomplete mass recovery and, we argue, follow naturally from the instrument window of the sampling methodology relative to the observed soil heterogeneity.

Analytical solutions for the advection‐dispersion equation (ADE) usually assume that boundary and initial conditions are orthogonal to the principal axes of the dispersion tensor. However, this is not always the case in field studies or modeling scenarios. Using the method of Green's functions, a generalized analytical solution of the three‐dimensional ADE set in an arbitrary Cartesian coordinate system (i.e., not orthogonal to the principal flow direction) is given for the solute resident concentration in a semi‐infinite porous medium with an arbitrary initial condition, surface boundary condition, and sink/source terms. Two particular solutions for a rectangular surface flux boundary condition and a buried parallelopiped, respectively, are derived from the general solution. The corresponding frequency domain solutions are also given which provide a more efficient method of computation for generating two‐and three‐dimensional grids via use of a two‐dimensional fast Fourier transform algorithm (as an alternative to the two‐dimensional numerical integration required to calculate the concentration using the real space solution). When the arbitrary coordinate system is orthogonal to the principal axes, it is shown that these particular solutions are the same as previously published results.

Simulation of transient solute transport under field conditions has been limited by the requirement that local water velocities must be represented explicitly by a hydrologic model. In this paper, that requirement is avoided by using a solute transfer function model to characterize solute dispersion explicitly in terms of field scale or area-averaged water flow past a depth of interest. The model is constructed by assuming that the solute travel-time probability density function to a depth of interest is an invariant property of the soil when expressed as a function of cumulative drainage, except for differences in water-storage volume at the time solute is applied. Both water storage and water flux at the field scale may be calculated by the water-balance equation, provided that a relationship can be developed between water flux at depth z and water-storage volume from z = 0 to z. The model is illustrated by simulating transient area-averaged solute concentrations as a function of time both for a field experiment involving chemical leaching under rainfall, and for a hypothetical process involving simusoidal water input and time-dependent solute input concentration. The model preserves mass balance and produced a good representation of the field data, using independently measured parameters to calibrate the transfer function and hydrologic function.

Most of the existing data on vadose zone field scale solute transport have been obtained from experiments conducted under transient, nonmonotonic water flow. However, the majority of the theoretical analyses of these experiments have used models which assume monotonic steady state water flow and uniform water content for the entire profile. In this study, transport of nonreactive solutes under nonmonotonic, transient water flow is analyzed numerically. The effect of hysteresis on solute transport is evaluated by making the soil hydraulic properties hysteretic using the procedure of Kool and Parker (1987). The effect of profile heterogeneity on solute transport is analyzed by assuming that the medium is scale heterogeneous in a vertical direction, with a random scale factor. Results of the simulations show that under transient water flow, analysis of solute transport data with a steady state water flow model may considerably overestimate the effective vertical pore water velocity. Under nonmonotonic water flow, when the hysteretic characteristics of the soil are important, transient flow models which neglect hysteresis can also seriously overestimate the solute velocity. In addition, failure to account for profile heterogeneity will also overestimate the solute velocity, because both hysteresis and profile heterogeneity change the water content profile and concurrently retard solute transport relative to the movement predicted if the soil water system is considered as homogeneous and nonhysteretic. Analyses of the computed breakthrough curves suggest that direct stimates of the amount of water drained below a given depth may improve the goodness of fit of the solution of the convection dispersion equation with constant effective parameters to the breakthrough curves obtained under transient conditions. The fitted parameters, however, are depth dependent and the resulting effective solute velocity is smaller than the steady state pore water velocity.

The solute concentrations measured in the field experiment of G. L. Butters et al. (this issue) are used to compare two models of vadose zone solute transport: the deterministic one‐dimensional convection‐dispersion model, which represents solute transport far from the source of solute entry, and the stochastic‐convective lognormal transfer function model, which represents solute transport near the source. The stochastic‐convective model provided an excellent representation of the spreading of the solute pulse to a depth of 3 m after calibration at 0.3 m. Conversely, the deterministic model dramatically underpredicted solute spreading beyond 0.3 m after calibration. An analysis of the area‐averaged solute concentration revealed a nearly linear scale effect in the dispersivity to a depth of at least 14.8 m. A change in the growth pattern of dispersion observed in the breakthrough curve at 4.5 m was attributed to a soil texture change near 3 m, which caused the apparent dispersivity of the pulse to decrease between 3.0 and 4.5 m, after which it increased significantly between 4.5 m and the final profile sampling between 0 and 25 m.

Modeling the field scale movement of chemicals in unsaturated soil is of intense interest to both the public and private sectors and has become an area of active theoretical research in a number of environmentally based disciplines. However, the experimental data needed to validate existing solute transport models and to inspire the development of more refined approaches is very limited. In this research study, the movement of a mobile tracer (Br−) was monitored as it moved through the unsaturated zone beneath the soil surface of a 0.64 ha loamy sand field. Under flux‐controlled, steady state water flow achieved by bidaily sprinkler irrigation, a narrow pulse of 58.9 mol/m−3 NaBr(aq) was applied uniformly to the field and subsequently leached downward while monitored by vacuum solution samplers replicated 16 times at each of 6 depths between 0.3 and 3.0 m and 6 times at the 4.5‐m depth. Six deep soil cores to a maximum of 25 m were taken to characterize the final field average bromide depth profile after the pulse had passed the 4.5‐m depth. Although the mass recovery of the area‐averaged pulse was near 100% at all depths, the coefficient of variation (CV) of mass recovery between samplers at a given depth was near 50%. Lateral variations in apparent vertical solute velocity or in solute transport volume were considerable, with CVs near 50% in the shallow monitoring depths. However, variations in transport volume with depth at a given site were also large, even though the solution samplers for different depths were displaced laterally by only 0.3–0.6 m at different sites. The mean vertical velocity of the area‐averaged solute pulse was significantly less than the ratio of the average net water flux to the average volumetric water content, until approximately 1.8 m. The difference between the two average velocities near the surface was large enough (nearly a factor of 2) to suggest that transient effects from the bidaily irrigations were influencing solute transport.