RECONSTRUCTING CLIMATE ON THE GREAT PLAINS FROM BURIED SOILS: A QUANTITATIVE APPROACH

Abstract

The Great Plains, U.S.A. lack quantitative paleoclimatic data for the late Quaternary largely because two common sources of paleoclimatic data, tree ring and pollen records, are rare in the region. Sequences of buried soils, however, are commonly preserved in eolian and alluvial sediments on the Great Plains and have the potential to enhance the region's paleoclimate record. This research presents a study of buried soils preserved in the Caddo Canyons of central Oklahoma to highlight opportunities and considerations for using buried soils to reconstruct past climate. Results indicate that sequences of buried soils dating to the mid- and late-Holocene are commonly preserved in the canyons. Canyon geomorphology dictates the nature of the fill contained in the canyons, and the effect of geomorphology and microclimate on soil formation must be carefully considered when interpreting the buried soil and stable carbon isotope record from the canyons. In an effort to capitalize on the rich paleoenvironmental record that buried soils can provide, this study presents the Buried Soil Reconstruction Model (BuSCR), a method for reconstructing paleoclimate based on properties of buried soils. The model was developed based on a study of modern analogue soils and climate on the Great Plains. BuSCR reconstructs mean annual precipitation (MAP), moisture index (Im), and mean annual temperature (MAT) with statistically significant results (r2 = 0.4, p < .0001) and low mean average errors. While error increases on the edges of the Great Plains climate envelope, application of BuSCR to a series of buried soils across the Great Plains, including soils from the Caddo Canyons, shows that it corroborates both paleoenvironmental reconstructions using other proxies (e.g. dune activation histories) and model-simulated hindcasts. In particular, BuSCR reconstructions corroborate model simulations of a -25% MAP anomaly during the Medieval Warm Period and a drastic reduction in MAP and Im across the Great Plains during the Altithermal. These results indicate that the BuSCR model, with further testing and if applied widely to buried soils across the Great Plains, could provide quantitative reconstructions of past climate that fill a current hole in the North American paleoclimate database.