POST-LGM WIND INTENSITIES FROM GRAIN SIZE PARTITIONING OF LATE QUATERNARY LOESS AT THE DRY CREEK ARCHAEOLOGICAL SITE, NENANA VALLEY, ALASKA

Continental loess is a rich record of paleoclimate data. Advances in analytical techniques over the last decade and a half have allowed for the use of loess deposits as detailed records of wind direction and intensity during the Quaternary. Numerical grain size partitioning is a technique pioneered in Asian dust records which mathematically decomposes multimodal distributions into separate components, each of which are assumed to be genetically distinct. Each genetically distinct component can then be analyzed separately, creating a detailed picture of changes in all depositional mechanisms contributing to a site through time. A high-resolution column of 57 loess samples was collected from the Dry Creek archaeological site in the Nenana River Valley in central Alaska in order to assess changes in local and regional atmospheric circulation patterns following the Last Glacial Maximum. Numerical grain-size partitioning using a mixed Weibull function was performed on grain-size distributions to obtain a record of relative wind intensity over the last ~15,000 years. Two grain-size components were identified, one with a mode in the coarse silt range (C1) and the other with a modal size ranging from medium- to very-coarse sand (C2). C1 dominates and provides a record of regional northerly winds carrying sediment from the Nenana River. These winds were strong during cold climatic intervals, namely the Older Dryas (14.2-14 ka) and two Holocene Neoglacial periods (7.5-4 ka, and ~2.5-1.6 ka), weak during the Allerød (14-13.3 ka) and Younger Dryas (12.9-11.7 ka), and variable during the Holocene Thermal Maximum (11.4-9.4 ka). Deposition of C2 was episodic and represents locally-derived sand entrained and deposited by southerly katabatic winds from the Alaska Range. These katabatic winds occurred frequently prior to 12 ka and after 4 ka. This study shows that numerical grain size partitioning is a powerful tool for reconstructing paleoclimate and that it may be successfully applied to Alaskan loess.