Master of Science (MS), Wright State University, 2006, Geological Sciences

Ground water flow and contaminant transport patterns are largely controlled by the distribution of high- and low-permeability sediments. Therefore, an accurate description of the aquifer architecture is paramount to producing a representative ground water model. Models of contaminant fate and transport in the aquifer near Woburn, Massachusetts, have previously been created by others using a deterministic approach. As a complement to these prior studies, the proportions, geometry, and juxtaposition of the different lithofacies of the aquifer were statistically characterized for developing stochastic models for the aquifer system. The descriptions of lithology from boreholes were separated into eleven categories based primarily on grain size. Hydraulic conductivity values were available for some of the categories and their frequency distributions were analyzed. However, it was not possible to conclusively divide the categories into facies based on permeability because of the overlap in the values. As a result, three classifications (termed A, B, and C) were devised to explore the effect of different classifications. The classifications were designed to represent both the worst- and best-case scenarios with respect to the volumetric proportion of low-permeability facies. In each classification, the study area was divided into three sections: the northern section, the central section, and the southern section. The proportion of low-permeability facies was found to be highest in northern section and lowest in central section. The vertical range of the low-permeability facies was characterized using the transition probability models while the variogram model characterized the lateral range of the low-permeability facies. The results of the stochastic characterization were utilized with a sequential indicator simulator code to produce visualizations under each classification. Using previous results of Ritzi et al., (2000) from transport studies of contamination in simulati (open full item for complete abstract)

Committee: Robert Ritzi (Advisor) Subjects: Geology

Master of Science, University of Akron, 2018, Geography-Geographic Information Sciences

Urban soils are often polluted with lead (Pb) from historical deposition from air pollution and other industrial sources. Lead in soil may pose a significant health risk to children under the age of six years. As community gardens become more common in urban areas in the United States, exposure of children to soil Pb is increased through direct contact with soil and the consumption of plants that are cultivated in urban gardens. The aim of this project was therefore to assess the spatial distribution of soil Pb in the city of Akron (Ohio), and to assess the potential for plants cultivated in urban gardens to absorb soil Pb. This study used four interpolation methods to interpolate soil Pb values at unmeasured locations and a comparison was made between the four methods to assess how best predictions were made for unmeasured locations. The four interpolation techniques used were Ordinary Kriging, Co-kriging, Empirical Bayesian Kriging (EBK) and Inverse Distance Weighting. A survey of urban gardens was also conducted to ascertain the type of plants gardeners most frequently cultivated and whether they adopted measures in reducing plant soil Pb uptake. Results from the study showed that the output surface from the four methods were relatively similar, and most surfaces within the city had soil Pb exceeding 600ppm. The EBK model gave the best predicted surface when the four methods were compared. The study also found that whilst soil Pb was high for most places within the city, gardeners replaced old soils with new clean soil and the majority of the crops they cultivated were fruit vegetables. Therefore the potential for plants cultivated in these gardens to absorb soil Pb is minimal.

Committee: Shanon Donnelly Dr. (Advisor); Linda Barrett Dr. (Committee Member); Teresa Cutright Dr. (Committee Member) Subjects: Geographic Information Science

Master of Science, The Ohio State University, 2013, Environment and Natural Resources

Given the rapid changes our environment experiences, soil resource data on various spatial and temporal scales are critical to monitor and mitigate soil degradation. Realistic and useful soil data can be obtained through digital soil mapping. The prediction models are based on deterministic relationships between soil attributes and one or more landscape variables. This concept argues that the relationship between soil and its environment provides the information necessary to infer what soil might occur at a given point by assessing the environmental conditions there. It is known that environmental processes occur at natural scales, describing the hierarchical organization of the environment. Although the scale for observation should be related to the object or function under investigation, there is no clear method to determine the optimal scale. The appropriate scale is identified by the data and the process being studied. In hierarchical systems, there is often a range of scales over which a process remains stable as scale changes. These ranges are separated by transition points, a scale or range of scales where the importance of explanatory variables on the process changes. By changing the raster cell size of digital terrain models, various scales of data can be created to match the relative scale of the studied process. Statistical analyses such as correlation and model performance can then be used to explore the relationships between environmental variables, processes and scale. Five methods were used to determine the most appropriate spatial resolution for environmental variables: (a) significant differences in cumulative distribution functions, (b) using the RMS slope measure, (c) using the maximum RMS measure, (d) using the peak variable value, and (e) using the peak local variance. Analysis of the CFDs revealed that altitude and aspect do not change significantly with resolution, and a general decrease in slope and an increase in TWI as data reso (open full item for complete abstract)

Committee: Brian Slater PhD (Advisor); Kristin Jaeger PhD (Committee Member); Edward McCoy PhD (Committee Member) Subjects: Geographic Information Science; Soil Sciences

MS, University of Cincinnati, 2001, Engineering : Environmental Engineering

The widespread arsenic contamination of groundwater in Bangladesh has been recognized as posing a serious health problem to millions of people in the region. Understanding of the complex lateral and vertical variability in its countrywide distribution is an important step towards better spatial estimation and also improved understanding of the controlling physical processes. A geostatistical characterization of arsenic followed by classical and indicator variogram analysis was performed in an attempt to study the partitioning of variability across scales and thereby identify the dominant physical process controls on arsenic. The results suggest that 'large-scale' (greater than tens of km approx.) geological and physical features control a significant fraction of the spatial variability in shallow wells (55%) as well as in the deeper wells (88%). The indicator variogram analysis also suggested the possibility of deeper aquifers acting as potential safe water sources in the country. The usefulness of countrywide spatial contamination maps is brought into question. The residual, small-scale variability appears to be unrelated to the widespread application of groundwater towards irrigation in the country. Experiments on the sampling pattern of wells in the arsenic datasets revealed that the variability in the arsenic distribution is proportional to the degree of spatial clustering of sampled tube wells. The result suggests that the nugget value in the variograms is not governed by purely random variability and instead, is controlled by distinct 'small-scale' processes. It is proposed that the prevalence of higher arsenic concentrations of arsenic in shallow wells is because of these 'small-scale' processes (less than tens of km approx.) exerting a greater degree of control at shallower depths in the sediments. A possibility of the right-skewedness of arsenic distribution leading to completely random variations in the arsenic concentrations yielding an increased number o (open full item for complete abstract)

Committee: Dr. Shafiqul Islam (Advisor) Subjects: Engineering, Environmental

Master of Science, The Ohio State University, 2011, Geodetic Science and Surveying

Airborne traverse data require spatial interpolation to generate continuous Digital Elevation Models (DEMs). DEMs in turn are important resources for geographical and environmental studies. In this study we use ice thickness data and surface elevation data collected by the NASA DC-8 aircraft as a part of the IceBridge mission from October and November, 2009. The ice thickness and surface elevation data downloaded from the National Snow and Ice Data Center (NSIDC) were processed for removal of anomalous data using a GIS. We developed a locally adaptive interpolation algorithm, which segments the data into several local regions based on the statistical properties of the data. We produced ice thickness, surface elevation and subglacial DEMs for Pine Island and Thwaites glaciers using the processed dataset and our locally adaptive algorithm. We made a comparative study between different interpolation techniques to determine the suitable interpolation technique for use with airborne traverse data. Finally, we validated our DEMs through visual comparison with the RADARSAT imageries and by comparison with the existing DEMs available for the Pine Island and Thwaites glaciers. Our DEMs can be used by glaciologists for climate change studies, as it captures the geographic features better than existing DEMs available for this region.

Committee: Dr. Rongxing Li PhD (Advisor); Dr. Kenneth C. Jezek PhD (Committee Member); Dr. Carolyn J. Merry PhD (Committee Member) Subjects: Civil Engineering; Climate Change; Computer Engineering; Earth; Geographic Information Science; Geophysics; Remote Sensing

Master of Science (MS), Ohio University, 1982, Industrial and Manufacturing Systems Engineering (Engineering)

Optimizing exploratory drilling locations

Committee: Donald Scheck (Advisor) Subjects: Engineering, Industrial

PHD, Kent State University, 2008, College of Business and Entrepreneurship, Ambassador Crawford / Department of Management and Information Systems

How important is manufacturing to the U.S. economy? This dissertation examines that question in three separate essays. The first essay compares the spatial distributions of the manufacturing and service sectors across U.S. counties, and analyzes the changing patterns over the 1990-2003 study period. The Getis-Ord Gi* statistic is used to identify clusters of economic activity. The data show a great deal of co-clustering, with growth in both sectors over time. In some cases, growth in the service sector cluster appears to predate growth in the manufacturing cluster at the same location. One cluster in Northeast Ohio is further analyzed, using input-output analysis to assess the forward and backward linkages between industries. Five key industries are identified, all in the manufacturing sector. While scholars and policy makers discuss the merits or drawbacks of a dwindling manufacturing sector, the hardship of lost manufacturing jobs was felt most strongly by workers in the Great Lakes states. The second essay uses a geostatistical technique called kriging to perform a space-time analysis of the extent of the displacement in this region. Kriging smoothes the data over space, enhancing visualization. In addition, this methodology allows for the stochastic interpolation of missing data points. Kaldor's laws posit that manufacturing is the engine of economic growth. Do these laws imply that a shift to services hurts the economy? The third essay is an empirical investigation of the differential growth rates in services and manufacturing, and their effect on state income growth. The model that is estimated takes account of the spatial autocorrelation in the data. The findings suggest that when the service sector expands faster than manufacturing or at the expense of manufacturing, economic growth is negatively affected.

Committee: Marvin Troutt PhD (Committee Chair); Felix Offodile PhD (Committee Member); Jay Lee PhD (Committee Member); Pervaiz Alam PhD (Committee Member) Subjects: Economics; Geography; Management; Statistics