Quantifying continental-scale distribution of transmissivity using inversion of groundwater level data

Transmissivity and permeability are crucial variables for large-scale groundwater model studies. Current large-scale datasets of permeability are based on calibrated groundwater models and high resolution lithology maps. We aim to extend continental-scale estimates of transmissivity of the shallow subsurface by using large databases of waterlevel data in combination with inverse analytical and numerical models. We use a dataset of 4000 stable long-term waterlevel records from the United States in combination with new datasets of perennial streams and watershed delineations and existing estimates of groundwater recharge. Transmissivity was calculated using simple analytical solutions and numerical models of groundwater flow that include a more realistic exchange with surface water bodies and evapotranspiration. We perform an uncertainty analysis to account for uncertainty in recharge, waterlevels and elevation data. We discuss how well we can predict transmissivity using these large-scale datasets and how the calibrated values of transmissivity correlate with lithology and geological setting.