RADIUM IN THE JORDAN SANDSTONE, MINNESOTA: DISTRIBUTION AND GEOCHEMICAL CONTROLS IN AN AQUIFER RECHARGED BY MELTWATER

Preliminary results from new water sampling of the Cambrian Jordan sandstone aquifer (CJDN), along with Minnesota Department of Health data, provide new insights on the distribution of and controls on radium (Ra) in this major water resource in the midwestern USA. Ra levels above the 5 pCi L^-1 (0.185 Bq L^-1) maximum contaminant level occur in a 100-km-long N-S corridor from the Minneapolis-St. Paul area to Medford, MN. Several trends constrain Ra-controlling mechanisms: (1) Elevated Ra occurs in confined, anoxic, slower-circulating portions of the aquifer, whereas portions receiving recent recharge indicated by ³H, elevated Cl^-, NO₃^-, and/or oxic conditions exhibit lower Ra. (2) High Ra is dominated by ²²⁶Ra, with ²²⁸Ra/²²⁶Ra <1 (median 0.76; often < 0.5 in high-Ra samples), a carbonate-like ²²⁸Ra/²²⁶Ra value lower than other sandstone aquifers. (3) The distribution of elevated Ra and low ²²⁸Ra/²²⁶Ra is similar between CJDN and overlying Ordovician aquifers in Minnesota: Prairie du Chien Group dolomite, St. Peter Sandstone, and Galena Group carbonate, despite lithologic differences. (4) Likewise, the fresh, Ca-Mg-HCO₃^- composition of waters in all four aquifers is similar, which implies that water-rock interaction with the quartz sandstone matrix of CJDN is less significant than equilibrium with secondary carbonate. (5) Chemical and isotopic indicators ((Ca+Mg)/Na, ⁸⁷Sr/⁸⁶Sr, ²²⁸Ra/²²⁶Ra) suggest that Ra sources in CJDN differ significantly from the underlying, high-Ra Mt. Simon-Hinckley aquifer. The corridor of elevated Ra coincides roughly with the extent of Wisconsin glaciation and aligns with a body of Pleistocene recharge in a deep, confined section of the aquifer in Iowa (Siegel 1991, Geology v19 p433); this might imply that recharge of meltwater influenced the present-day Ra distribution. Also, present-day U concentrations in CJDN are associated with redox conditions, indicated by higher U in the shallow subcropping and faulted NE portion of the aquifer. These data are consistent with a hypothesized mechanism in which (a) an influx of fresh, oxic meltwater could have mobilized U to oxic-anoxic interfaces in different locations than today; (b) subsequent ingrowth contributed Ra to secondary carbonate; and (c) Ra is contributed to groundwater by carbonate dissolution. However, other primary, diagenetic, or glacial influences on Ra cannot be ruled out.