College of Agriculture & Natural Resources
Permanent URI for this communityhttp://hdl.handle.net/1903/1598
The collections in this community comprise faculty research works, as well as graduate theses and dissertations.
Browse
4 results
Search Results
Item Release, Survival, And Removal of Bovine Manure-Borne Indicator Bacteria Under Simulated Rainfall(2017) Stocker, Matthew Daniel; Hill, Robert L; Environmental Science and Technology; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The effects of simulated rainfall intensities and its interactions with manure consistency and weathering on the release, survival, and removal of fecal indicator bacteria, Escherichia coli and enterococci, from land-applied dairy manure were evaluated. Rainfall intensity had significant effects on the number of bacteria in the soil following rainfall. Bacteria concentrations in soil decreased with increased soil depths and the topmost centimeter of soil accounted for the greatest proportion of bacteria. Escherichia coli persisted longer than enterococci once removed from manure. Manure consistency was not a significant factor in the removal of bacteria when manure was fresh, but as manure weathering progressed, consistency became a significant factor. The Vadas-Kleinman-Sharpley model was preferred over the exponential model for simulating the removal of manure-borne bacteria. Results of this work will be useful for improving predictions of the human health risks associated with manure-borne pathogenic microorganisms.Item Modeling Nitrogen, Phosphorus and Water Dynamics in Greenhouse and Nursery Production Systems(2011) Majsztrik, John Christopher; Lea-Cox, John D.; Plant Science and Landscape Architecture (PSLA); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Nutrient and sediment runoff from the six states and Washington, DC that form the Chesapeake Bay watershed is a major cause of environmental degradation in the Bay and its tributaries. Agriculture contributes a substantial portion of these non-point source loads that reach the Bay from its tributaries. Research in this area has traditionally focused on agronomic farm contributions, with limited research on the nursery and greenhouse industry. This research presents the first known attempt to model operation-specific information, validated by published research data, where multiple variables are assessed simultaneously. This research provides growers and researchers with a tool to assess and understand the cultural and environmental impact of current practices, and predict the impact of improving those practices. Separate models were developed for greenhouse, container-nursery and field-nursery operations, since specific production variables and management practices vary. Each model allows for simple entry of production input variables, which interface with the Stella modeling layer. Each model was first calibrated with one published research study, and subsequently validated with another peer-reviewed study, with multiple independent runs for each model. Validation results for all three models showed consistent agreement between model outputs and published results, increasing confidence that models accurately process all input data. Verified models were then used to run a number of what-if scenarios, based upon a database of production practices that was gathered from 48 nursery and greenhouse operations in Maryland. This database provided a detailed analysis of current practices in Maryland, and adds significantly to our understanding of various operational practices in these horticultural industries. Results of the what-if scenarios highlighted model sensitivities and provided answers to hypotheses developed from the analysis of the management database. Some model functions, such as denitrification, would greatly benefit from additional research and further model modification. Models were designed to be easily adapted to local conditions for use throughout the U.S. and potentially other parts of the world.Item Evaluation of SWAT Model Applicability for Waterbody Impairment Identification and TMDL Analysis(2007-10-30) Sexton, Aisha M; Shirmohammadi, Adel; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The U.S. EPA's Total Maximum Daily Load (TMDL) program has encountered hindrances in its implementation partly because of its strong dependence on mathematical models to set limitations on the release of impairing substances. The uncertainty associated with predictions of such models is often not formally quantified and typically assigned as an arbitrary safety factor to the margin of safety (MOS) portion of TMDL allocations. AVSWAT-X, a semi-distributed, watershed-scale model, was evaluated to determine its applicability to identify the impairment status and tabulate a nutrient TMDL for a waterbody located in the Piedmont physiographic region of Maryland. The methodology for tabulating the nutrient TMDL is an enhancement over current methods used in Maryland. The mean-value first-order reliability method (MFORM) was used to calculate variance in output variables with respect to input parameter variance and the MOS value was derived based on the level confidence in meeting the water quality standard. A calibration, validation and an uncertainty analysis was conducted on the AVSWAT-X model. Monthly results indicated that AVSWAT-X is a good predictor of streamflow, a moderate (at best) predictor of nutrient loading and a poor predictor of sediment loading. Improved performance was observed on an annual basis for nitrate and sediment loadings, indicating the most appropriate use of SWAT for long-term simulations. The most pronounced reason for discrepancies in model performance was the use of the SCS curve number method to tabulate surface runoff. Uncertainty results indicated that input parameters that are highly sensitive may not necessarily contribute the largest amount of uncertainty to model output. The largest amount of variance in output variables occurred during wet periods. Predicted sediment output had the largest amount of variability around its mean, followed by nitrate, phosphate, and streamflow as indicated by average annual coefficients of variation of 28%, 19%, 17%, and 15%, respectively. The methodology used in this study to quantify the nitrate TMDL and the MOS associated with it, was a useful tool and an improvement over current methods of nutrient TMDL analysis in Maryland. Overall, AVSWAT-X is a moderate to good model for estimating waterbody impairment and conducting TMDL analysis of waterbodies impaired by nutrients.Item A Comparison Of Artificial Neural Networks And Statistical Regression With Biological Resources Applications(2006-08-07) Resop, Jonathan Patrick; Montas, Hubert J; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Artificial neural networks (ANNs) have been increasingly used as a model for streamflow forecasting, time series prediction, and other applications. The high interest in ANNs comes from their ability to approximate complex nonlinear functions. However, the "black-box" nature of ANN models makes it difficult for researchers to design network structure or to physically interpret the variables involved. Recent investigations in ANN research have found connections linking ANNs and statistics-based regression modeling. By comparing the two modeling structures, new insight can be gained on the functionality of ANNs. This study investigates two primary relationships between ANN and statistical models: the potential equivalence between feed-forward neural networks (FNN) and multiple polynomial regression (MPR) models and the potential equivalence between recurrent neural networks (RNN) and auto-regressive moving average (ARMA) models. Equivalence is determined through both formal and empirical methods. The real-world phenomenon of streamflow forecasting is used to verify the equivalences found. Results indicate that both FNNs and RNNs can be designed to replicate many regression equations. It was also found that the optimal number of hidden nodes in an ANN is directly dependant on the order of the underlying physical equation being modeled. These simple relationships can be expanded to more complex models in future research.