THE COLUMBIA PLATEAU DUST ENGINE DURING THE LAST GLACIAL MAXIMUM: TROUBLE WITH COLD STARTS

Global circulation models are used as tools in characterizing Quaternary climate change, but are occasionally contradicted when compared to eolian records. Simulations show that the large ice sheets of North America created high pressure from which a glacial anticyclone was generated. Anticyclonic winds coming off the ice sheet were apparently from the east, different from the winds that generated expansive dune and loess deposits during the last glacial maximum (LGM). Several studies have noted the mismatch between models and geologic data and have either dismissed the anticyclone or minimized its influence. Loess accumulations that span the LGM on the Columbia Plateau, Washington, record a hiatus or slowing in dust deposition rates that was likely due to anticyclonic winds. The Palouse loess, which covers approximately 10,000 sq km as thick as 75 m, has grain-size and thickness trends suggesting dominant dust-transporting winds were persistently from the southwest as far back as 75 ka. Luminescence ages of loess that span the LGM and beyond (18-40 ka) show a sharp decrease in deposition rates proximal to dust-producing basins, and a hiatus at loess sections closest to the ice sheet. This decrease in dust deposition occurs at a time when world-wide dust deposition was increasing. Paleoclimate proxies including paleosol features and opal phytolith assemblages confirm cold and dry conditions from 18 to 40 ka, represented by a well-developed calcium carbonate horizon and sagebrush steppe dominating the Columbia Plateau. Exhaustion of the sedimentary supply is an unlikely cause of the hiatus because loess was still accumulating at proximal sites. The most likely explanation for the hiatus relates to easterly winds generated by the anticyclone that squelched prevailing southwesterly winds, shutting down dust deposition near the ice sheet. The influence of easterly winds, however, may have been seasonal, and were not strong enough to result in reorganization of the eolian system. The luminescence ages combined with paleoecologic data suggest anticyclonic winds may have influenced the Columbia Plateau as far back as 40 ka, when ice volumes were increasing. This evidence potentially implicates the anticyclone as more than just a LGM phenomenon.