A finite element method for the solution of two‐dimensional transient dispersive‐convective transport of nonconservative solute species in fractured porous media is presented. A two‐nodal point one‐dimensional transport element for fractures is developed which provides a number of advantages relative to conventional fracture representation by two‐dimensional continuum elements. To eliminate the oscillatory behavior of convective‐dominated transport which is a more likely occurrence in fracture, a very efficient one‐dimensional upstreaming method along with a two‐dimensional method is implemented. Validity of the numerical scheme is established by comparison with existing one‐ and two‐dimensional analytic solutions.

A nitrogen‐transformation model was formulated for simultaneous computation of various N transformations: nitrification, denitrification, mineralization, immobilization, ion exchange, and plant uptake. These processes were assumed to obey first‐order kinetics. Differential equations that describe the above processes are solved simultaneously by using the 4th‐order Runge‐Kutta method and the Adams‐Bashforth‐Moulton predictor‐corrector equations. The computer model was tested against data reported from several incubation studies. This model could be linked to a transport model to predict the temporal and spatial distribution of various N species in soil profiles.

Previously, a steady‐state conceptual N model was applied to corn (Zea mays L.) field trials at two University of California Field Stations. That model examined only the inputs and outputs of N into a soil‐plant‐water system without considerations of carryover inorganic soil N and mineralization of organic soil N. It predicted N leaching losses reasonably well in the corn plots at the Kearney site where residual soil N was low, but not for the Davis site where considerable mineralization of soil organic N occurs.This steady state model was extended so that it would be applicable to the transient conditions prevailing at the Davis site. The revisions included provisions for the presence of residual inorganic soil N and the mineralization of organic soil N as well as inputs of organic N such as manures and sewage sludges and their mineralization rates. The original water submodel was not modified. The model is described in the appendix and required input parameters with examples are listed.The revised model was tested with 2 years of corn field trials at Davis which consisted of four (NH4)2SO4 rates (0, 90, 180, and 360 kg N/ha) and three irrigation regimes (1/3, 3/3, and 5/3 of the corn evapotranspiration). Sensitivity analysis evaluated the importance of the harvested crop coefficient, the denitrification coefficient, and the mineralization rate constant for organic soil N, and aided in estimating the latter two experimentally unmeasured parameters.Computed results compared favorably with measured N content in the harvested (grain and stover) crop, the residual inorganic soil N after harvest, and the average annual N concentration in the soil solution beneath the root zone at the 3‐m depth. There was no leaching of N in the 1/3 ET treatment; from 0.5 to 12 kg N/ha per year N leaching losses in the 3/3 ET treatment; and from 13 to 144 kg N/ha N leaching losses in the 5/3 ET treatments. Some differences were noted in the annual leaching losses between the 1975 and 1976 corn crops in the 5/3 and 3/3 ET treatments mainly due to differences in seepage out of the root zone.It is concluded that the applicability and utility of this expanded model have been significantly enhanced over that of the previous steady‐state model.

The distribution coefficient of trichloroethylene (TCE) was obtained from field and theoretical methods. The field method was based on measuring TCE concentrations in the soil samples and in the adjacent ground water. The theoretical method was based on using the organic carbon content of the soil and the octanol/water partition coefficient for TCE. The average distribution coefficient for 19 field samples and four methods of calculation was 0.18 ml/g which is in agreement with literature data and octanol/water partition coefficients results. For soils containing greater than 0.1 percent organic carbon, the theoretical methods of calculating the distribution coefficient appear to be valid. For soils low in organic carbon content, laboratory determinations of the distribution coefficient can provide reasonable estimates for predicting actual migration rates. Field determinations of distribution coefficients are, however, preferred because they integrate the effect of various factors on partitioning of TCE.

--The electrical conductivity of a colloid-water-electrolyte system increases with the frequency of the applied alternating electric current. The phenomenon is referred to as conductivity dispersion. This paper reports on the effects of electrolyte type, electrolyte concentration, and water content on the dispersion characteristics of kaolinite, illite, and silty clay soils, with emphasis on the mechanisms governing the dispersion phenomenon. It was observed that magnitude of conductivity dispersion increases with a reduction in water content, electrolyte concentration, and cation-exchange capacity of the clay. The type of ions influence the electrical dispersion through their size and mobility. Frequency effect increases as the hydrated radius of the counterions associated with the clay surface increases. Conductivity dispersion is explained primarily in terms of counterion/co-ion ratio in the diffuse double layer. Increase in the ratio of counterions to co-ions is an indication of a larger contribution to conduction by counterions than by co-ions, which in turn results in a larger frequency effect. Although diffusion coupling has an important role in the electrical dispersion characteristics of clay-water-electrolyte systems, other coupling phenomena, particularly electro-osmotic coupling, plays a significant part.