GEOPHYSICAL DELINEATION OF SALINE GROUNDWATER PLUME PATTERNS NEAR SALINE LAKES IN SEMIARID DUNE ENVIRONMENTS

Although theoretical models of hydraulics near lakes in ambient groundwater flow are available, these models are being tested because the 3D salinity distributions in groundwater are commonly unavailable. In general, conceptual models indicate low groundwater salinity upgradient and elevated lake salinity downgradient from lakes that concentrate solutes by evaporation. Standard invasive groundwater sampling methods rarely reach the necessary sampling density for 3D plume characterization, especially if numerous lakes are studied. However, salinity variations in groundwater strongly correlate with electrical conductivity (EC). In cases where geophysical signals from saline water can be discriminated from variations due to sediments, the plume geometry can be inferred from electrical methods.

The relative homogeneity of eolian materials and a rolling landscape in dune environment makes the Nebraska Sandhills area ideal for delineation of spatial plume patterns using geophysical ER and EM methods supplemented by sparse invasive measurements (wells and direct-push holes). Using direct current ER techniques guided by EM reconnaissance, we elucidate patterns of electrically conductive areas (plumes) and geometry near 4 lakes (0.2 -120 g/L range of TDS and 100-1000 m in diameter).

We assessed locations of the center of plume mass from the shore line, and mass variance in longitudinal and transverse directions. Commonly, TDS are on the order of 10 g/L , and much lower than lake TDS due to various mechanisms. The most conductive regions of the aquifer and largest saline concentrations are commonly attached to the lakes and generally are associated with the regional downgradient flow direction. Local topography and gradients contribute to the plume characteristics that correspond to certain conceptual steady-state and transient groundwater models. An updated conceptual model is discussed based on the observed field data.