Abstract A detailed geologic framework model was utilized for groundwater analysis using a fully three-dimensional variably saturated flow model. The geologic framework model, which was developed by a team of glacial geologists from federal and state geological surveys, was fully three-dimensional and did not contain the usual (unrealistic) assumption of widespread aquifer layers separated by leaky aquitard layers of equal extent. The goal of the analysis was to explore the implications of the new generation of geologic framework models for regional groundwater flow, and particularly, groundwater–surface water interactions. A transient numerical simulation, using infiltration at the ground surface as a boundary condition, revealed rich flow complexity, including: (1) widespread, yet patchy, recharge areas with rates that vary through several orders of magnitude, with the recharge rates being statistically correlated to hydraulic conductivity of the vadose zone sediments, elevation, and ground surface slope; (2) the predominance of local flow systems, resulting in an abundance of seepage zones along the sides of the incised (postglacial) stream valleys, and other manifestations of the high water table and strong groundwater–surface water interaction, such as kettle lakes and wetlands; and (3) existence of partially confined aquifers owing to partial burial of deltaic deposits by moraines and lake-bottom deposits having slow vertical permeability. Taken together, these findings support the need for, and value of, high-resolution geologic framework models and the potential fruitful outcome of strong collaboration between glacial geologists and groundwater modelers.

Eolian sand deposited in lakes downwind of coastal sand dunes can record a history of paleoclimatic fluctuations. The eolian sand signals from sediment within the Grand Mere Lakes, Michigan, which are downwind of sand dunes along southeastern coastal Lake Michigan, record the same sunspot, climate history, and lake-level fluctuations observed elsewhere along the east-central Lake Michigan coastline. Sediment cores were extracted from the Grand Mere Lakes in Berrien County, Michigan, and analyzed for variations in weight percentage of sand with depth, the sand signal, at 1 cm sampling intervals. Radiocarbon dates obtained from terrestrial macrofossils within the cores were used to develop age-depth models, from which sedimentation rates were derived, both for the varying sedimentary facies and the entire core. Spectral analyses of the sand signal data using both multi-taper and REDFIT methods indicate multiple periodicities that correspond to those from other regional and global studies, including Lake Michigan lake-level fluctuations, Lake Michigan coastal dune formation, and solar cycles. The common periodicities between the Grand Mere Lakes sand data and other studies suggest the sand-signal data set is not random, and is best explained as a record of paleo dune mobility. The appearance of the 80–110 year Gleissberg solar cycle in the data suggests that the storminess recorded by the eolian sand was influenced by periodic variability in extratropical cyclones across the Lake Michigan basin which, in turn, reflects variability in circulation patterns driven by the North Atlantic Oscillation, the variability of which has been associated with solar cycles.