Quantification of spatially varying hydrogeologic properties for a predictive model of groundwater flow in the Ogallala aquifer, northern Texas Panhandle
— Alan R. Dutton, Robert E. Mace, and Robert C. Reedy

The Ogallala aquifer, an extensive unconfined aquifer, is the main source of agricultural and drinking water across the Texas Panhandle and the U.S. High Plains. After 50 yrs of water production at rates in excess of recharge, less than half of the original saturated thickness remains in parts of the aquifer in the northern Texas Panhandle. Numerical models of groundwater flow in the Ogallala are tools with which to manage the aquifer; at least 15 numerical groundwater-flow models have been developed for different parts of the Ogallala aquifer in Texas. Regional models of groundwater flow in the heterogeneous Ogallala aquifer should draw on as many spatially varying data as possible. Besides the basic geometry of the aquifer (top and bottom and lateral boundaries), model input for hydraulic conductivity and recharge can be delimited using spatial attributes. Contouring hydraulic conductivity on the basis of trends in depositional systems and percentage sand and gravel in the aquifer represents both geological heterogeneity and a degree of lateral continuity not captured in simple variogram geostatistics. Recharge rates can be predicted on the basis of long-term precipitation trends and locally adjusted where soils are more or less permeable than average. The combination of mapping hydraulic conductivity using depositional systems and recharge using precipitation and soil factors minimized model calibration error to less than 5 % of the hydraulic head change across the study area. Reaching a low model calibration error with negligible parameter adjustment raises confidence in the applicability of a groundwater-flow model to aquifer management.