In constitutive modeling, one of important tasks is to calibrate the model. To calibrate a model is to find out the values of the model parameters for a material whose stress-strain behavior is to be simulated by the model. Conventional approach is to find certain well-defined states ill certain tests where behavior of a material is controlled by those parameters and then the stress and strain and other history parameters at those states can be used to find them. However, as the model evolves more sophisticated, such as the Disturbed State Concept Model (DSC), in which a greater number of parameters are introduced to account for behavior of the material under various stress conditions, it is not possible to find an easy way to calibrate, mainly due to certain stress-strain states are difficult to be isolated out. In this study an optimization approach is proposed by using quasi-Newton method with BFGS up-dating scheme. Contrary to the conventional approach which determines parameter values by averaging values of laboratory tests or by simple data fitting of the assumed parameter relations, the optimization approach is to find the best agreement of the model simulation with the experimental observation, then gives a set of parameter values for the best agreement which is quantitatively measured by the least error residual. Weight is used in the optimization procedure to emphasize on better simulation agreement with the observation for certain stress path conditions. This weight can be decided based on the engineering judgment for certain practical problems. By using the DSC model to simulate stress-strain response of various laboratory tests of sands, and by using the DSC model in a finite element analysis to simulate dynamic soil-structure interaction response of a shaking table test for saturated soil, it is shown that the optimization approach yields closer agreement with the observation. Based on the proposed optimization approach, a computer program DSCOPT is developed for the DSC model. The program takes the laboratory test data as input and outputs the model parameter values by the conventional and optimized approaches, and graphics plots of the model simulation.

In an effort to determine the agronomic necessity of K fertilization of cotton (Gossypium spp.) in Arizona, a five-year study was initiated in 1991, with a single field study near Gila Bend. Subsequent sites selected ranged from western (Yuma) to eastern (Safford) Arizona which totaled 11 site-years. Both Upland (G. hirsutum L.) and Pima (G. barbadense L.) cottons were cultivated, using soil and foliar applications of K. In 1992, study sites included the Safford Ag. Center (SAC), Maricopa Ag. Center (MAC), and a fanner cooperator site at Coolidge. In 1993, the experiment stations (SAC and MAC) were continued and Yuma Valley was added. The 1994 sites included only the experiment stations (SAC and MAC). In 1995, SAC and MAC were again maintained and a third location, a fanner cooperator site at Buckeye was added. Results from the study (12 site-years) indicated no lint yield increases due to K fertilization in all locations with either Upland or Pima cotton However, in 1995 at Buckeye, the result revealed a significant yield reduction due to the K foliar treatments. There were, however, no significant differences among the soil-only or the soil- plus-foliar treated plots at this location. Potassium (K) release kinetics of clay samples from 10 agricultural representative soils of Arizona was determined by successive extraction using a Ca-saturated cation resin. A preseason physical and chemical characterization of the soils showed all soils contain smectite-mica minerals. Four mathematical models (power function, Elovich, parabolic diffusion and first-order) were used to describe the nonexchangeable K release reaction involving 700-hour cumulative extraction time. Comparison of the models using the coefficient of determination (r²) and the standard error of the estimate (SE) indicated that the Elovich and the power function equations overall displayed the best fit. The first-order, and for the most part, the parabolic diffusion equation did not describe the K release very well. The constants a (initial rate) and b (release rate) for the Elovich and the power function equations, are at least in the order of magnitude as those found by others in several previous studies.

Numerical simulations were used to study spatial averaging in low-induction-number frequency-domain electromagnetic induction (LIN FEM) instruments. Local ( LS) and cumulative (CS) sensitivity were used to analyze three different aspects of LIN FEM spatial sensitivity. LS is the variation in a measured property given a small change at a given location of the property of interest. CS contours are derived from LS and reveal the shape and the fraction of total instrument sensitivity enclosed within the contours. The first study re-evaluated the asymptotic approach to LIN FEM spatial sensitivity. Using this approach, LIN FEM measurements have often been assumed to represent electrical conductivity (sigma) at discreet depths that do not vary with the sigma of the ground. This assumption was tested using simulations of electromagnetic fields in environments with homogeneous and layered sigma distributions. When the induction number was greater than 0.01, the 1-D vertical CS distribution and the depth of investigation varied up to 20% over the range of sigma simulated. As sigma increased, CS contours and depth of investigation decreased in depth. In the second study a small perturbation approach was used to calculate CS distributions so that each distribution is unique to a given LS distribution. CS was summed from regions of high to low LS, and retained information on the magnitude and location of LS. As sigma increased, CS became focused around the highest LS values. The maximum reduction in depth of investigation was about 40% at the highest sigma investigated. In the final study, a series of small, electrically conductive perturbations was simulated in a three-dimensional, homogeneous environment. Three-dimensional LS varied markedly with a large difference between horizontal (HMD) and vertical (VMD) orientations of the transmitter and receiver dipoles. In some regions, the calculated magnetic field intensity with the perturbation was less than that calculated for the host without the perturbation. This occurred for both VMD and HMD orientations of the transmitter. CS contours were highly complex. One dimensional, vertical LS curves extracted from the three-dimensional data were very different from curves from infinite layer simulations.

Concern over environmental quality has generated interest in the chemistry of Cd in soils. When Cd sorption and the influence of soil characteristics on the process are better understood, Cd contamination of ground water and plant availability may be assessed more accurately. Therefore, a series of experiments, including laboratory batch studies, soil thin layer chromatography studies and selective ion electrode procedures were conducted to evaluate the effect of soil properties on Cd sorption and mobility by fourteen different Aridisols and Andisols having a range of chemical properties. In addition, surface complexation models were used to model the amount of Cd adsorbed onto each of the humic acid and hydrous ferric oxide surfaces. Sorption isotherms were obtained using batch experiments by in which 25 ml of solution containing a total of 0.8, 1.6, 3.2, 6.4, and 12.8 ppm of Cd²⁺ to 0.5 g were added to soil samples. The results of the batch experiments indicate that in all the soils used the Cd sorption is best described by the Freundlich sorption isotherm. The maxima soil sorption capacities were significantly correlated with the free iron oxides. The Cd activities in the soils varied from 10⁻⁷·⁴⁴ to 10⁻⁴·⁸² M. It increased with increasing total Cd added and were inversely related to the soil pH The Cd mobility retardation factor (Rf) obtained from soil thin layer chromatography ranged from 0.25 to 0.95. It showed that Cd was slightly mobile in 64%, moderately mobile in 29% and very mobile in 7% of the soils. The Cd Rf indicates that Cd mobility would decrease with increasing amounts of iron oxide fractions; silt % and exchangeable Mg²⁺ in the soils. The non electrostatic and diffuse layer models results indicate that humic materials are an important factor in Cd sorption at pH values greater than 3 and hydrous ferric oxide surfaces are important at pH values greater than 7. The ion activity products of Cd(OH)₂ ranged from 2.3 x 10⁻²⁵ M to 5.6 x 10⁻¹⁶ M, while the solid activity coefficients of adsorbed Cd (SACCd2+) ranged from 3.9 x 10⁻⁶ to 4.6. The SAC(Cd₂₊) values were significantly correlated yet negative with silt %, CEC, Ald, Alo, O.C and iron fractions.

A High N fertilizers citrus rates may contribute to a rise in the appearance of groundwater NO₃ around groves in central Arizona. Nitrogen fertilizer recommendations need to be reevaluated for citrus due to recent changes in fertilizers, irrigation technology, and increasing groundwater NO₃. Studies were initiated to optimize fertilizer N management for microsprinkler-irrigated citrus. Navel oranges were planted in Jan. 1997 with varying N rates and frequency of liquid urea NH₄-NO₃ or granular controlled-release fertilizers (CRFs) in order to determine tree response, N partitioning, N balance, and residual soil N. Trees grew slowly in 1997 with no more than 6% of the fertilizer N taken up by trees, while < 94% of fertilizer N remained in the top 0.9 m of soil. Leaf N was higher in fertilized plots than in controls as trees grew rapidly in 1998 with < 25% of the fertilizer N taken up by the trees. The soil contained low NO₃ concentrations, suggesting most of the fertilizer N was leached. Nitrogen is not needed during the first season after planting and rates of 68 to 136 g N tree⁻¹ applied in monthly intervals during the second season maintain adequate tree growth and N reserves. Application of CRFs resulted in little tree response during the two year study. Soil inorganic N was highest in the surface 0.6 m in 1997. In 1998, ammonium sulfate treatments resulted in higher NO₃ leaching than plots that received CRFs. Eight-year old 'Redblush' grapefruit trees were treated with two N rates and three fertigation frequencies in a factorial arrangement. There no differences in trunk diameter, or fruit quality on trees between 1996 and 1998. Leaf N content was > 18 mg kg⁻¹ in all plots in 1996. Leaf N declined in Aug. 1997 and control plots had lower leaf N than the other treatments. Yields tended to be the greatest at high N rates with monthly fertigation. Soil N levels were proportional to fertilizer N inputs. In 1998, the high N treatment at weekly and monthly frequencies increased leaf N compared to the other treatments. Results suggest that N rates of 272 to 408 g tree⁻¹ year⁻¹ at weekly or monthly frequencies may be optimal for desert-grown mature grapefruit.

Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) in both near infrared (NIR) and mid infrared (MIR) has been previously shown to be effective in quantifying soil nitrogen (N) concentrations when calibrated using numerous field soil samples. However, such an approach provides samples that likely contain substantial correlations between physical and chemical properties. To address these concerns, the performance of DRIFTS coupled with PLS regression in NIR regions, 5,000-4,000 cm⁻¹ (2,000-2,500 nm) and 6,500-5,500 cm⁻¹ (1,540-1,820 nm), and the M1R region, 3,400-2,400 cm⁻¹ (2,940-4,170 nm), was assessed first through analysis of the concentration of mineral N (ammonium (NH₄⁺) (0-50 ppm) and nitrate (NO₃⁻) (0-200 ppm)) artificially incorporated into a series of silica sand and clay samples with a consistent particle size. The influence of the reduction of sand particles to silt was also analyzed quantitatively. Subsequently, the Pima clay loam soil was evaluated and the concentration ranges of 0-200 ppm NH₄⁺ and 180-1,000 ppm NO₃⁻ were added in soil samples. All three regions provided good measurement of NH₄⁺ but the MIR region was significantly more useful for NO₃⁻ measurement in sand. The detection limits for the measurement of mineral N in sand with particle sizes within 212-300 μm using the MIR region were 9 ppm NH₄⁺ (7 ppm NH₄-N) and 36 ppm NO₃⁻ (8 ppm NO₃-N). For silt (particles less than 53 mum), the most effective model was the MIR region for both NH₄⁺ and NO₃⁻ measurements, yielding the detection limits of 15 ppm NH₄⁺ (12 ppm NH₄-N) and 50 ppm NO₃⁻ (11 ppm NO₃-N). The MIR region also performed reasonably well with soil samples but both NIR regions provided poor results. The detection limits for NH₄⁺ and NO₃⁻ measurements in soil were 100 ppm NH₄⁺ (78 ppm NH₄-N) and 330 ppm NO₃⁻ (75 ppm NO₃-N) with the correlation coefficients (R²) of roughly 80% and 90%, respectively. The spectral range of 2,900-2,400 cm⁻¹ was the effective common range for mineral N measurement in sand, silt, and soil samples.

A free water surface finite element model was developed. The method was implemented with the Galerkin approach to solve the Laplace equation in the saturated region. It was developed in the object oriented Visual C ++ computer language to permit easy update and drawing of the adaptive mesh. For each time step, the new water table position was calculated based on flux across the water table, a Brooks-Corey equation mass balance for the unsaturated region, and an equation that calculates water table position for the saturated region. An equation was developed to calculate a drainage transfer coefficient, alpha, based on percentage of perforated area in the drain tube wall. The drainage transfer coefficient was incorporated into the finite element model as a Fourier boundary condition. To validate the finite element model, its results were compared with the Kirkham equation results for steady state recharge of three subsurface drainage systems. The finite element model was used to calibrate a semi-analytical frozen stream tube model for subsurface drainage of heterogeneous soils. The first step in the calibration procedure is to run the finite element model for steady state recharge and calculate the water table height divided by recharge rate (the stream tube resistance to flow) as a function of distance between drains. Least squares regression is used to fit a polynomial logarithmic equation, called the resistance function, to the stream tube resistance to flow vs. distance from the drain curve. A differential equation based on the principle of conservation of mass and application of Darcy's law to the frozen stream tube was solved to obtain an equation that calculates stream tube flow rate and final water table elevation as a function of the resistance function and initial water table elevation. An example was developed for a non-homogeneous subsurface drainage system to illustrate the use of the semi-analytical model to predict water table fall and discharge.

Irreversible sorption has been widely observed for hydrophobic organic contaminants in natural sediments. It is a complicated process and impacts sediment quality and contaminated sediment management. In this research, experiments were conducted to further understand the nature of irreversible sorption and the mechanism(s) controlling this process. The unique characteristics of the irreversible compartment were delineated with the sorption and desorption of five chlorinated benzenes in four natural sediments. It was observed that each chemical-sediment combination exhibited a fixed amount of maximum irreversible sorption capacity, qirrmax . However, the organic carbon based partition coefficient associated with the irreversible compartment was essentially constant for different chemical-sediment combinations--- Kirroc=105.42+/-0.17 . The desorption behaviors were modeled with an irreversible sorption model, in which a Langmurian-type expression (representing the irreversible fraction) was added to the linear model. For the sorption and desorption of different compounds in the same sediment qirrmax seems to be related to both Kow and solubility. For sorption and desorption of the same compound in different sediments qirrmax depends mainly on sediment organic carbon content (foc), but may also be affected by the origin of the sediments. The release from the irreversible compartment was not affected by a number of factors, such as extended desorption time, competitive sorption, caustic treatment of sediments, and external mechanic forces. Many observations in this research suggest that sorption and desorption of hydrophobic organic contaminants in natural sediments are not controlled by the same mechanism, and irreversible sorption may be partly due to the association of organic sorbate molecules to some highly insoluble and hydrophobic materials in sediments. Irreversible sorption greatly affects the fate and transport of contaminants in surface and subsurface environments---it causes contaminants to persist in soils and sediments, and it also reduces the efficiency of pump and treat and bioremediation. Nevertheless, the same process could also greatly reduce the ecological and human-health risks of contaminated sediments. Thus, it is necessary to establish risk-based sediment quality standards, which could be of enormous economical impact. Also, to quantitatively address the impact of irreversible sorption on sediment quality, a one-dimensional solute-transport model was developed. This simple model could provide valuable support for environmental decision making.

A hybrid finite element-boundary element formulation is developed for the nonlinear seismic analysis of 2D soil-structure systems that contain fully or partially saturated soils. The proposed formulation aims at combining the sophistication, versatility and nonlinear capabilities of the FEM with the ability of the BEM to model rigorously energy radiation. The FE part of the formulation is used to model the effective stress problem in the near field, incorporating all material and geometric nonlinearities. The BE part is used to model the infinite domain as an elastic medium, by considering strain-dependent equivalent linear properties. The new method is used in a series of analyses and reproductions of published results to verify its validity, demonstrate its versatility and attain more insight on select problems. The new formulation is applied on two different types of structures, with main objective to provide a better understanding of their seismic behavior. The first study is a reanalysis of the Lower San Fernando dam under the 1971 earthquake. The study reproduces remarkably many of the known behavior characteristics and the mode of failure of the dam. In a subsequent analysis aimed at demonstrating the effects of foundation flexibility, the original dam response is compared to the response assuming a hypothetical flexible foundation. The results demonstrate a substantial reduction in the response of the flexible foundation case. The second study examines the dynamic response of waterfront retaining walls. Emphasis is placed on the effects of relative density and stiffness of the backfill and foundation materials. As expected, the study shows that a wall with a dense backfill sand endures the dynamic loading with minimum permanent deformation, whereas a wall with loose backfill sand may experience excessive deformations and liquefaction. Significant deformation is also observed for the model with flexible foundation, despite the reduction of wave energy through radiation damping. The developed FE-BE formulation allows a deformation-based design of soil-structure systems by accounting rigorously for complex wave propagation phenomena and material nonlinearities.

<p> Digital soil mapping (DSM) is a rapidly growing area of soil research that has great potential for enhancing soil survey activities and advancing knowledge of soil-landscape relationships. To date many successful studies have shown that geographic datasets can be used to model soil spatial variation. This thesis addresses two issues relevant to DSM, scale effects on digital elevation models, and predicting soil properties. The first issue examined was the effect of spatial extent on the calculation of geometric land surface parameters (LSP) (e.g. slope gradient). This is a significant issue as they represent some of the most common predictors used in DSM. To examine this issue two case studies were designed. The first evaluated the systematic effects of varying both grid and neighborhood size on LSP, while the second examined how the correlation between soil and LSP vary with grid and neighborhood size. Results of the first case study demonstrate that finer grid sizes were more sensitive to the scale of LSP calculation than larger grid sizes. While the magnitude of effect was diminished when comparing a high relief landscape to a low relief landscape, the shape and location of the effect was similar. Results of the second case study showed that the correlation between soil properties and slope curvatures were similarly optimized when varying the spatial extent, but that the effect was more sensitive to grid size than neighborhood size. Slope gradient also showed significant correlations with some of the soil properties, but was not sensitive to changes in grid or neighborhood size.</p><p> The second study attempted to predict numerous physical and chemical soil properties for several depth intervals (0-15, 15-60, 60-100, and 100-150-centimeters), using generalized linear models (GLM) and geographic datasets. The area examined was the Upper Gauley Watershed on the Monongahela National Forest, which covers approximately 82,500 acres (33,400 hectares). This watershed represents a complex landscape with contrasting geologic strata, deciduous and coniferous forests, and steep slopes. Given this landscape diversity it was still possible to fit GLM which explained on average 38 percent of the adjusted deviance for rock fragment content, and exchangeable calcium and magnesium, and phosphorus. Some of the most commonly selected environmental predictors were slope curvatures, lithology types, and relative slope position indices. This seems to validate the prominence of these variables in theoretical soil-landscape models. Had the correlation between the soil properties and slope curvatures not been optimized by varying the spatial extent, it is likely that another less suitable LSP would have been selected.</p>