Source and model output layers used for the prediction and display of the probability of elevated concentrations of redox-sensitive constituents in groundwater in the Fox-Wolf-Peshtigo watershed in Wisconsin and Michigan

This USGS data release contains 1-kilometer resolution source-layer rasters used to predict redox conditions and contaminant concentrations in groundwater in the Fox-Wolf-Peshtigo watershed in Wisconsin and Michigan using random forest classification. The model output layers are 1-kilometer resolution rasters of the predicted probability of elevated concentrations of nitrate, iron, and arsenic. These data support the report Predicting Redox-sensitive Contaminant Concentrations in Groundwater Using Random Forest Classification. This data release supports the following publication: Tesoriero, A.J., Gronberg, J.M., Juckem, P.F., Miller, M.P., and Austin, B.P., 2017, Predicting redox-sensitive contaminant concentrations in groundwater using random forest classification: Water Resources Research, v. 53, https://doi.org/10.1002/2016WR020197.