Exploratory data analysis and model development (in R) between Carbon Dioxide emissions and measurements of other power plant variables (Load, Stack Gas Pressure, Heat Input, Stack Temperature, SO2, and NOx). Performance of the final model (Heat Input and Stack Temperature) is generalizable across the several power plants and has good explanatory power.
Population pyramid showing age-gender (y-axis and blue/red colors respectively) relationships of enrolled patients in GCDD (in MS Excel). Black line represents age-gender relationships for the entire country of India
A high-resolution user-interactive tool for analyzing biomass feedstock supplies built using Microsoft Excel (BioFeedStAT®). The tool is used to determine quantities vs. distances (in 0.5-mile increments — actual road miles, not air miles) and transport costs of any combination of target feedstocks. Source data for BioFeedStAT® is a combination of US Forest Service Forest Inventory Data, USDA Cropland Data Layer, and USDA Census of Agriculture (available at www.biomass2.com/fsa/fsa.html)
Visualization of multiple patients in the Republic of Moldova followed longitudinally through time. Data for this visualization is collected by laboratories in Moldova and typed in to the University of Arkansas’s Qualtrics Survey system and then dynamically visualized by combining R scripts with Microsoft’s PowerBI system.
Sankey diagram (HTML/Javascript) showing the relationships between the drug susceptibility testing results and M. tuberculosis lineages. Isolate counts in each phenotype and genotype group are shown in parentheses published in https://www.ncbi.nlm.nih.gov/pubmed/25859997 )
Comparisons of population projections produced by the State of Mississippi and an independent audit/analysis designed to examine the unsupported population growth estimates used by Mississippi Gulf Region Water and Wastewater Plan to use $653 million in federal Hurricane Katrina recovery money for water and wastewater facilities (ArcGIS). Audit included forensic analysis of data analysis and comparisons of their published method to normal good practice in the field.
Online interactive viewer for Lyme Disease incidence across Maryland and Pennsylvania. The viewer was developed for a final logistic regression model and tied ArcGIS for Server Javascript API to logistic regression predictive equations. The application assisted users as they explore models by the usage of interactive sliders to vary disease rate thresholds and risk probabilities and examine the consequences (base model published in https://www.ncbi.nlm.nih.gov/pubmed/21644127)
Multi-Sector Sustainability Browser (MSSB), an interactive decision support tool (Javascript) containing information from the scientific literature and technical reports that must be considered when making decisions to support sustainability objectives in the key sustainability areas (Land Use, Buildings and Infrastructure, Transportation, and Materials Management). The MSSB is designed to assist communities in understanding the impacts that sustainable decision alternatives and actions made in the key sustainability areas can have on human health, the economy and the environment. (A description of the MSSB can be found in this Eric S. Hall, A Decision Support Tool for Sustainable Land Use, Transportation, Buildings/Infrastructure, and Materials Management, American Journal of Environmental Engineering, Vol. 7 No. 2, 2017, pp. 35-46. doi: 10.5923/j.ajee.20170702.02. article: http://article.sapub.org/10.5923.j.ajee.20170702.02.html)
Species Distribution Model (SDM) for the soil fungus Coccidioides spp. which causes coccidioidomycosis (ArcGIS & custom programming). The SDM combined: (1) geographic range, (2) habitats where cocci is found, and (3) environmental data. Geographic range and habitat data were combined to create pseudo occurrence/non-occurrence points. These points and environmental data were used as inputs to MAXimum ENTropy (MaxEnt) modelling program. Very few areas in North America were predicted to have range shrinkage in particular Washington state were predicted expand.
Potential invasive species to the Great Lakes were modeled using the Genetic Algorithm for Rule-Set Production which combined global environmental layers with locations invasive species have previously been found in a statistical model to predict where within the Great Lakes these species may find suitable habitat. All of Lake Erie and the shallow water areas of the other four Great Lakes are most vulnerable to invasion (published in EPA/600/R-08/066F)
GAP Model for the Ovenbird (Seiurus aurocapilla) in Arkansas (GRASS GIS). The current range within the state (left).The distribution of potential habitats within the state was based on habitats the species is known to occur (middle). The predicted distribution is the combination of those two maps (right).
Infographic (ArcGIS and MS Excel) showing Los Angeles, County (California) coccidioidomycosis (Valley Fever) cases over time (2001-2014). Graphic shows an outbreak (Jan 1994) and how the number of cases has remained stable since the earthquake.
Time series plots for data from the different power plants are reproduced below. X-axis is the date and Y-axis (and units) is shown within the label of the chart. Carbon Dioxide (percent) is always shown as a reference as the chart on the bottom. Data shown are raw (blue line), median (black line), and LOESS smoothed (red line).The first set show Heat Input, Stack Flow, Stack Pressure, Load, and Carbon Dioxide for all the power plants. The next set shows Nitrogen Oxides, Sulfur Dioxide, Stack Gas Temperature, and Carbon Dioxide.
