SPATIALLY DISTRIBUTED REGIONAL RECHARGE-RATE ESTIMATION TO GUIDE AN AQUIFER-SENSITIVITY ASSESSMENT FOR MID-CONTINENTAL GLACIAL ENVIRONMENTS, USA

A map showing the distribution of groundwater recharge for the glaciated portion of Indiana, USA, was facilitated by a multiple regression model that, in turn, was based on a governing conceptual model that describes recharge as controlled by topography, geology, and land cover. The proportion of baseflow to runoff in stream hydrographs (n=189), determined using a local-minimums approach with a recession slope test, was used as the measure of recharge in this study. Partial correlation coefficients were used to identify and rank the most statistically influential (p=0.001) independent variables in the study domain, as follows: the elevation of the near-surface water table, the proportion of sand in the subsurface (as determined by lithologic logs of water-well records), and the thickness of near-surface clay. Other factors that appear to explain the spatial distribution of recharge include: the cumulative upstream area occupied by water bodies (streams, lakes, ponds), the density of wetland areas, and the cumulative upstream area occupied by forest. The model has a multiple adjusted R² value of 0.53. The model-estimated recharge rates vary between 1 and 34.5 cm/year (mean of 14 cm/year), averaging 12.5% of precipitation. The greatest deviations in the model residuals are overestimates in the north-northwestern part of the state where sandy soils are common. Limited pockets of this region are over-estimated by up to 10 cm/year; however, the median residual for the entire glaciated portion of Indiana (north of the Wisconsin glacial limit) is a slight underestimate of -0.04 cm/year. Groundwater-quality data collected from domestic wells, public water supplies, and commercial/industrial locations were used to identify settings that have experienced groundwater contamination. To assess the sensitivity of near-surface aquifers to contamination, locations with detectable contamination were compared to the groundwater-recharge data to classify the region into low, medium, high, and very high aquifer sensitivity.