Proceedings of the 2020 Mississippi Water Resources Conference

Year: 2020 Authors: Adams R., Miller B., Kress W.H., Minsley B., Rigby J.R.

The U.S. Geological Survey (USGS) is currently conducting a multi-year analysis and recharacterization of the Mississippi Embayment Regional Aquifer System (MERAS) groundwater flow model focusing on the water resources within the Mississippi Alluvial Plain (MAP). Part of this recharacterization was the evaluation of the existing model based on uncertainty and data worth analysis. These data quality measurements indicated that the MERAS model was sensitive to groundwater-surface water exchange, but this component was poorly constrained and the confidence in the model forecast was low.

To increase the density of data within the models' most sensitive rivers and streams, the USGS completed 900 kilometers (km) of waterborne resistivity surveys within Mississippi to characterize streambed lithology. This technique characterizes streambed itself and the near surface (upper 15-30 meters) of the streambed that controls the recharge to the alluvial aquifer. These data can be used to map changes in the lithology of the streambed and identify areas of potential groundwater-surface water exchange.

To map these sediments, electrical resistivity data was collected using a resistivity meter connected to floating multi-electrode cables. Information about the spatial location of each data point, depth of the water column, and electrical properties of the water column were also collected.

Five rivers in the Mississippi Delta region were the focus of this study: the Bogue Phalia, Quiver, Sunflower, Tallahatchie, and Yazoo Rivers. These rivers flow over a variety of fluvial and deltaic deposits. While streambed sediments show a strong correlation with surficial geology, changes in the vertical extent of those geologic features had a strong impact on the aquifer recharge potential expected for a given water body.

The inverted waterborne resistivity data were transformed to hydraulic conductivity using relationships derived from geophysical logs collected within the study area. Estimated hydraulic conductivity values generated from downhole nuclear magnetic resonance (NMR) data were compared to electromagnetic induction logs to generate a relationship between electrical resistivity and hydraulic conductivity.

The resistivity-derived estimates of hydraulic conductivity show a significant increase in magnitude and spatial variability as compared to the estimates derived from groundwater model parameter estimation. Some amount of this change was expected due to the increased sampling density and smaller footprint of the resistivity surveys. The remainder of the difference between the two estimates is likely due to the incision of river channels into and beneath the shallow 5-10 m confining unit that overlays a large portion of the MAP within Mississippi.