Stephen R. Workman

SIMULATION OF DAILY AND MONTHLY STREAM DISCHARGE FROM SMALL WATERSHEDS USING THE SWAT MODEL

The Soil and Water Assessment Tool (SWAT) was evaluated and parameter sensitivities were determined while modeling daily streamflows in a small central Kentucky watershed over a two-year period. Streamflow data from 1996 were used to calibrate the model and streamflow data from 1995 were used for evaluation. The model adequately predicted the trends in daily streamflow during this period although Nash-Sutcliffe R 2 values were –0.04 and 0.19 for 1995 and 1996, respectively. The model poorly predicted the timing of some peak flow values and recession rates during the last half of 1995. Excluding daily peak flow values from August to December improved the daily R 2 to 0.15, which was similar to the 1996 daily R 2 value. The Nash-Sutcliffe R 2 for monthly total flows were 0.58 for 1995 and 0.89 for 1996 which were similar to values found in the literature. Since very little information was available on the sensitivity of the SWAT model to various inputs, a sensitivity analysis/calibration procedure was designed to evaluate parameters that were thought to influence stream discharge predictions. These parameters included, drainage area, slope length, channel length, saturated hydraulic conductivity, and available water capacity. Minimization of the average absolute deviation between observed and simulated streamflows identified optimum values/ranges for each parameter. Saturated hydraulic conductivity, alpha baseflow factor, drainage area, channel length, and channel width were the most sensitive parameters in modeling the karst influenced watershed. The sensitivity analysis process confirmed die trace studies in the karst watershed that a much larger area contributes to streamflow than can be described by the topographic boundaries. Overall, the results indicate that the SWAT model can be an effective tool for describing monthly runoff from small watersheds in central Kentucky that have developed on karst hydrology however calibration data are necessary to account for solution channels draining into or out of the topographic watershed.

VEGETATED FILTER STRIP REMOVAL OF CATTLE MANURE CONSTITUENTS IN RUNOFF

Pasture runoff can contribute to elevated concentrations of nutrients, solids, and bacteria in downstream waters. The objective of this study was to determine the effects of vegetative filter strip (VFS) length on concentrations and transport of nitrogen, phosphorus, solids and fecal coliform in runoff from plots treated with cattle manure. Three plots with dimensions of 2.4 ×30.5 m were used. The upper 12.2 m of each plot was treated with cattle manure, while the lower 18.3 m acted as a VFS. Runoff produced by rainfall simulators was sampled at VFS lengths of 0, 6.1, 12.2, and 18.3 m and analyzed for total Kjeldahl nitrogen (N), ammonia N, nitrate N, total phosphorus (P), ortho-P, fecal coliforms, total suspended solids and other parameters. The VFS significantly reduced concentrations and mass transport of incoming solids, fecal coliform, and most nutrient forms, particularly P. The relationships among VFS length, concentration and mass transport were well-represented by first-order exponential decay functions. Approximately 75% of incoming total Kjeldahl N, total P, ortho-P, and total suspended solids was removed within the first 6.1 m of the filter strips. Runoff concentrations of fecal coliform concentrations entering the filter strips were as high as 2 ×107 FC/100 mL; after a filter length of 6.1 m, however, the runoff exhibited no measurable concentration of fecal coliforms. This experiment suggests that even relatively short filter strips can markedly improve quality of runoff from grassed areas receiving cattle manure.

Statistical procedures for evaluating daily and monthly hydrologic model predictions

The overall study objective was to evaluate the applicability of different qualitative and quantitative methods for comparing daily and monthly SWAT computer model hydrologic streamflow predictions to observed data, and to recommend statistical methods for use in future model evaluations. Statistical methods were tested using daily streamflows and monthly equivalent runoff depths. The statistical techniques included linear regression, Nash-Sutcliffe efficiency, nonparametric tests, t-test, objective functions, autocorrelation, and cross-correlation. None of the methods specifically applied to the non-normal distribution and dependence between data points for the daily predicted and observed data. Of the tested methods, median objective functions, sign test, autocorrelation, and cross-correlation were most applicable for the daily data. The robust coefficient of determination (CD*) and robust modeling efficiency (EF*) objective functions were the preferred methods for daily model results due to the ease of comparing these values with a fixed ideal reference value of one. Predicted and observed monthly totals were more normally distributed, and there was less dependence between individual monthly totals than was observed for the corresponding predicted and observed daily values. More statistical methods were available for comparing SWAT model-predicted and observed monthly totals. The 1995 monthly SWAT model predictions and observed data had a regression Rr 2 of 0.70, a Nash-Sutcliffe efficiency of 0.41, and the t-test failed to reject the equal data means hypothesis. The Nash-Sutcliffe coefficient and the Rr 2 coefficient were the preferred methods for monthly results due to the ability to compare these coefficients to a set ideal value of one.

SUITABILITY OF A GPS COLLAR FOR GRAZING STUDIES

The traditional means of tracking animal location in a field is by visual observation. Not only is this method labor intensive, it is also prone to error as the observer can alter cattle movement, observation periods are often too short to obtain confidence in general daily behavior patterns, and observer fatigue becomes an issue. In the 1990s, the University of Kentucky began using GPS collars on cattle to track their position with the goal of incorporating this information into cattle management practices. One of the key unanswered questions regarding the GPS collars is the accuracy of the position data recorded by the collar. The objective of this work was to assess the capabilities and limitations of using GPS collars to track animal movement in grazed watersheds. Static tests were conducted in an open field, under trees, and near fence lines to ascertain the impacts of various field features on collar performance. Dynamic tests were carried out to examine the errors associated with the collars while operated under real-world conditions. Results from these tests indicate that the collars generally provide data with horizontal accuracies of 4 to 5 m. This information will assist researchers in the development of experiments based on collar capabilities and limitations.

EVAPORATION REDUCTION POTENTIAL IN AN UNDISTURBED SOIL IRRIGATED WITH SURFACE DRIP AND SAND TUBE IRRIGATION

The efficiency of drip irrigation is highly dependent on evaporation losses occurring from the constantly saturated soil beneath emitters. Advent of subsurface drip irrigation is in part an approach to curb this inefficiency. An irrigation method, Sand Tube Irrigation (STI), is proposed to increase the efficiency of “Normal” surface applied drip Irrigation (NI method) on permanent tree crops without the need for burying the irrigation tubing. The sand tube consists of removing a soil core beneath the emitter and filling the void with coarse sand. A weighing lysimeter was constructed in the laboratory and instrumented to directly measure temporal evaporation from large, undisturbed soil columns, 0.7 m in diameter and 0.8 m in height. Experiments were performed on six replicated soil monoliths to compare the two methods. The results indicated that, for four consecutive days after irrigation, there was a significant difference at the 95% confidence level between evaporation occurring from the NI and STI methods. After four days of evaporation, comparison of water contents indicated that a higher amount of water existed between the depths of 0.2 to 0.55 m in the STI versus the NI method. Although drainage occurred from the macropore structure of the undisturbed soil monoliths, the STI method showed potential in retaining more water in the micropore structure of the lower depths, that would be available for plant use rather than potential evaporation.

Comparison of Daily Water Table Depth Prediction by Four Simulation Models

The Agricultural Drainage And Pesticide Transport (ADAPT) model was compared to the water management simulation models DRAINMOD, SWATREN, and PREFLO. SWATREN and PREFLO are one-dimensional finite-difference models while ADAPT and DRAINMOD are one-dimensional mass balance models. ADAPT, an extension of the computer model GLEAMS, also provides chemical transport information. All four models were tested against field data from Aurora, North Carolina. Observed water table depth data were collected during 1973 through 1977 from a water table management field experiment with three subsurface drain spacing treatments of 7.5, 15, and 30 m.

Hydrologic Properties of Pervious Concrete

Pervious concrete is concrete made by eliminating most or all of the fine aggregate (sand) in the concrete mix, which allows interconnected void spaces to be formed in the hardened product. These interconnected void spaces allow the concrete to transmit water at relatively high rates. The main objective of this project was to conduct research on the potential application of pervious concrete in agricultural settings, specifically for use in animal feed lots, manure storage pads, animal manure and bedding compost facilities, or floor systems in animal buildings. Laboratory tests were conducted on replicated samples of pervious concrete formed from two rock sources (river gravel and limestone) for coarse aggregates and different size fractions to determine hydrologic relationships. Linear relationships were found between density and porosity, density and permeability, porosity and permeability, and porosity and specific yield. The results suggest that properties such as permeability, porosity, and specific yield are not significantly affected by different aggregate types. However, density and porosity can be effective methods for predicting porosity, specific yield, and permeability. In addition, t-tests were conducted to determine the effect of aggregate types on the solid/liquid separation properties of the pervious concrete after adding composted beef cattle manure and bedding to the surface of the specimens. The amount of composted beef cattle manure and bedding retained within the specimens was significantly less (p = 0.012) when samples constructed of #8 river gravel were used rather than the other aggregates. The #8 river gravel also had significantly less reduction in permeability compared to other aggregates. Although the #8 river gravel had a different effect on the compost retained and the reduction in permeability for the specimens, all four aggregates exhibited a significant reduction in the permeability after the compost was applied.

Streambank Erosion Associated with Grazing Practices in the Humid Region

The effects of cattle grazing on stream stability have been well documented for the western portion of the U.S., but are lacking for the east. Stream and riparian damage resulting from grazing can include alterations in watershed hydrology, changes to stream morphology, soil compaction and erosion, destruction of vegetation, and water quality impairments. However, few studies have examined the successes of best management practices (BMPs) for mitigating these effects. The objective of this project was to assess the ability of two common BMPs to reduce streambank erosion along a central Kentucky stream. The project site consisted of two replications of three treatments: (1) an alternate water source and a fenced riparian area to exclude cattle from the stream except at a 3.7 m wide stream ford, (2) an alternate water source with free stream access, and (3) free stream access without an alternate water source (i.e., control). Fifty permanent cross-sections were established throughout the project site. Each cross-section was surveyed monthly from April 2002 until November 2003. Results from the project indicated that the incorporation of an alternate water source and/or fenced riparian area did not significantly alter stream cross-sectional area over the treatment reaches. Rather than exhibiting a global effect, cattle activity resulted in streambank erosion in localized areas. As for the riparian exclosures, changes in cross-sectional area varied by location, indicating that localized site differences influenced the processes of aggradation and/or erosion. Hence, riparian recovery within the exclosures from pretreatment grazing practices may require decades, or even intervention (i.e., stream restoration), before a substantial reduction in streambank erosion is noted.

Atrazine and alachlor dissipation rates from field experiments

Chemical transport is being monitored in the root zone of three agricultural management systems at the Ohio Management Systems Evaluation Area (OMSEA). Atrazine and alachlor concentration data from soil cores taken to a depth of 0.9 m and partitioned into the increments of 0.0 to 0.15, 0.15 to 0.3, 0.45 to 0.6, and 0.75 to 0.9 m show the herbicides remained in the top 0.15 m of the profile during the 1991 and 1992 growing seasons. The slow movement of herbicides was partly due to below normal rainfall during the period. Since the herbicides have not been transported out of the soil profile, dissipation rates could be determined from the field observations.

MODELING SURFACE AND SUBSURFACE PESTICIDE TRANSPORT UNDER THREE FIELD CONDITIONS USING PRZM-3 AND GLEAMS

Contaminant transport models should be evaluated over a wide range of conditions to determine their limitations. The models PRZM and GLEAMS have been evaluated many times, but few studies are available in which predicted movement in runoff and percolate were simultaneously evaluated against field data. Studies of this type are essential because pesticide leaching and runoff are mutually dependent processes. For this reason, PRZM-3 and GLEAMS were evaluated for their ability to predict metribuzin concentrations in runoff, sediment, subsurface soil, and pan lysimeters under three field conditions (yard waste compost amended, no-till, and conventional-till) on a Lowell silt loam soil. Sensitive input parameters were either site specific (climatic, soil, and chemical) or calibrated (K-factor, C-factor, curve number). In general, both models under-predicted metribuzin concentration in runoff water, runoff sediment, subplow layer soil (15-75 cm), and pan lysimeter water (75 cm). Contrary to field data, both models predicted that a large percentage (> 50%) of metribuzin would move below the “mixing zone” (top 1 cm) during the first rainfall event after application. Relatively little metribuzin was predicted to move beyond the plow layer (top 15 cm) into the pan lysimeters or subsurface soil throughout the simulation period, possibly due to the lack of a macropore component in the models. High metribuzin concentrations in sediment (field data) indicated that relatively little metribuzin moved below the “mixing zone”, possibly because of hysteresis but much of the metribuzin that did move was quickly transported into the pan lysimeters, probably due to macropore flow. GLEAMS more accurately predicted pesticide concentration in sediment and PRZM predicted subsurface soil concentration somewhat more accurately than GLEAMS. Little difference in accuracy was detected between models on metribuzin concentration in runoff or metribuzin concentration in percolate. Although both models generally under-predicted metribuzin concentration in runoff, runoff transport (mass of metribuzin in runoff) for the study period was over-predicted by both models which emphasizes the importance of accurately predicting herbicide concentration and runoff volume soon after application when the surface pesticide concentrations are highest.