A QUANTITATIVE APPROACH TO RECONSTRUCTING GREAT PLAINS CLIMATE FROM SOIL PROPERTIES

Soils that formed in alluvial and eolian sediments during the late-Quaternary are commonly preserved in the stratigraphic record on the Great Plains. Because pollen and tree rings rarely provide reliable signals on the Great Plains, soil stratigraphy has emerged as an effective technique for reconstructing environmental change in the region. However, no quantitative method for reconstructing past climate, such as those that exist for pollen assemblages and tree-ring growth patterns, exists for buried soils. Consequently, the Great Plains paleoclimate record exhibits large gaps, and climate change models that rely on such data for validating predictions of future climate in the region display large uncertainties (Kutzbach et al. 1998, Webb et al. 2004, National Research Council 2006). Based upon the results of a pilot study that correlated climate variables from a water budget analysis to properties of modern soils in Kansas (Zung and Feddema 2011), we focused on three pedogenic processes strongly linked to climate: melanization, lessivage, and calcification. Quantitative variables were developed to represent the climate factors that drive these processes (e.g. precipitation, moisture deficit, moisture surplus) and the soil properties that result. Methods for standardizing soil properties for confounding state factors such as parent material and time will be discussed. Multiple regression and discriminant function analysis of modern soil data from the NCSS Soil Characterization Database and historical climate data from the NCDC Traditional Climate Divisional Database was used to develop quantitative models for reconstructing climate from soil properties. The resulting models will be presented, and considerations for applying and testing these models on buried soils will be discussed.