Monitoring and modeling wetland chloride concentrations in relationship to oil and gas development

Extraction of oil and gas via unconventional methods is becoming an important aspect of energy production worldwide. Studying the effects of this development in countries where these technologies are being widely used may provide other countries, where development may be proposed, with some insight in terms of concerns associated with development. A fairly recent expansion of unconventional oil and gas development in North America provides such an opportunity. Rapid increases in energy development in North America have caught the attention of managers and scientists as a potential stressor for wildlife and their habitats. Of particular concern in the Northern Great Plains of the U.S. is the potential for chloride-rich produced water associated with unconventional oil and gas development to alter the water chemistry of wetlands. We describe a landscape scale modeling approach designed to examine the relationship between potential chloride contamination in wetlands and patterns of oil and gas development. We used a spatial Bayesian hierarchical modeling approach to assess multiple models explaining chloride concentrations in wetlands. These models included effects related to oil and gas wells (e.g. age of wells, number of wells) and surficial geology (e.g. glacial till, outwash). We found that the model containing the number of wells and the surficial geology surrounding a wetland best explained variation in chloride concentrations. Our spatial predictions showed regions of localized high chloride concentrations. Given the spatiotemporal variability of regional wetland water chemistry, we do not regard our results as predictions of contamination, but rather as a way to identify locations that may require more intensive sampling or further investigation. We suggest that an approach like the one outlined here could easily be extended to more of an adaptive monitoring approach to answer questions about chloride contamination risk that are of interest to managers.