Data Science
In situ soil moisture release curves
The RainMan project explores the plant and ecosystem consequences of temporal repackaging of precipitation. The app shows how soil water content and soil water potential respond to different treatments. I also report Van Genuchten parameters that were derived from in situ sensors.
Timeseries analysis of VOCs
Volatile organic compounds (VOCs) of decomposing litter can signal the presence and activity of the microbial community. I designed an algorithm and app for detrending, modeling, and classifying 304 VOC timeseries.
Exploring a C. diffile gene expression database
The bacterial pathogen Clostridioides difficile is the leading cause of health-care related bacterial infections in the US, and the mechanisms of disease are poorly understood and actively investigated. I built an app for researchers to query and visualize gene expression data across published studies.
Postfire effect of rangeland restoration practices
A naturally-occurring fire burned a rangeland restoration experiment, offering a unique opportunity to assess how well herbicide, grazing, and native grass seeding could control cheatgrass invasion after a major disturbance. I implemented a series of Bayesian ANOVAs to assess biomass, count, and cover of Bromus tectorum that accounted for overdispersion and a complex experimental design. The paper and repository are openly accessible.
Predicting trends in US organic agriculture
Understanding future trends in organic agriculture sales can inform production decisions by growers; while the USDA conducts surveys at the state level, data are infrequent and sometimes missing. We developed Bayesian hierarchical regression models to estimate trends in the number of organic farms, area under organic production, and sales of organic produce, which included missing values and error estimates. The paper and repository are openly accessible.