Sand hills in the Savannah River valley in Jasper County (South Carolina, USA) are interpreted as the remnants of parabolic eolian dunes composed of sand derived from the Savannah River and stabilized by vegetation under prevailing climate conditions. Optically stimulated luminescence ages reveal that most of the dunes were active ca. 40 to 19 ka ago, coincident with the last glacial maximum (LGM) through early deglaciation. Modern surface winds are not sufficient for sustained eolian sand transport. When the dunes were active, winds blew at velocities of at least 4 m/s from west to east, and some vegetation was present. The ratio of annual precipitation to potential evapotranspiration (P:PE) was less than the modern ratio of 1.23 and may have been < 0.30, caused by stronger winds (which would have resulted in greater evaporation) and/or reduced precipitation. The Savannah River dunes are part of a larger assemblage of eolian dunes that were active in the eastern United States during and immediately after the LGM, suggesting that eolian sediment behavior in this region has been controlled by regional forcing mechanisms during the Quaternary.

Geological maps of South Carolina, covering >6800 km2, confirm the existence of eight preserved Pleistocene shorelines above current sea level: Marietta (+ 42.6 m), Wicomico (+ 27.4 m), Penholoway (+ 21.3 m), Ladson (+ 17.4 m), Ten Mile Hill (+ 10.7 m), Pamlico (+ 6.7 m), Princess Anne (+ 5.2 m), and Silver Bluff (+ 3 m). Current geochronologic data suggest that these eight shorelines correlate with Marine Oxygen Isotope Stages (MIS) as follows: Marietta—older than MIS 77; Wicomico—MIS 55–45; Penholoway—MIS 19 or 17; Ladson—MIS 11; Ten Mile Hill—MIS 7; Pamlico—MIS 5; Princess Anne—MIS 5; and Silver Bluff—MIS 5 or 3. Except for the MIS 5e Pamlico, and possibly the MIS 11 Ladson, the South Carolina elevations are higher than predicted by isotope proxy-based reconstructions. The <4 m of total relief from the Pamlico to the Silver Bluff shoreline in South Carolina, for which other reconstructions suggest an expected relief of ~ 80 m, illustrates the lack of match. Our results suggest that processes affecting either post-depositional changes in shoreline elevations or the creation of proxy sea-level estimates must be considered before using paleo sea-level position on continental margins.

No abstract