Using GIS (Geographic Information System) compile and create digital geographic themes (explanatory variables) for the High Plains aquifer. These themes will include both vulnerability (conditions at the surface, such as percent irrigation, percent and type of agriculture, soil type, precipitation, etc) and susceptibility (conditions within the sub-surface, such as geology, hydraulic conductivity, specific yield, etc). Historical water-quality data have been compiled (Litke, 2001) and current water-quality data are being collected and compiled.

Develop, calibrate, and verify multivariate logistic regression (LR) models for nitrate and atrazine concentrations in ground water of the High Plains aquifer. These LR models will assist in developing statistically significant relations between explanatory variables and nitrate and atrazine concentrations that exceed selected thresholds, such as Federal drinking water standards. The most statistically significant explanatory variables will be identified. Well-specific probabilities of detecting nitrate and atrazine will be calculated using LR model output and the logit transformation.

Using GIS data interpolation techniques of the calculated well-specific probabilities, probability maps will be created. These maps will spatially identify the predicted probability of detecting nitrate and atrazine concentrations in ground water across the entire High Plains aquifer.

Explanatory variables in final LR models will be adjusted to represent future or hypothetical changes in land-use or ground water management. This type of model forecasting may identify changes in ground-water-quality, resulting from these scenarios.