GIS-BASED MODELING OF CULTURAL DYNAMICS IN THE CENTRAL GREAT PLAINS

A GIS-based model for predicting potential archaeological site locations on the High Plains of western Kansas has been developed. The raster data model is used to assign a probability score of containing archaeological material to every 30 m2 land parcel. The probability score is based on local environmental conditions. Environmental variables relevant to cultural activity, such as distance to water, distance to rock outcrop, and distance to overlook, are combined with archaeological site data and entered into a multiple regression analysis. The regression analysis relates the presence or absence of an archaeological site to the environmental conditions at each location. Because the model focuses on hunter-gather lifeways, no socio-cultural variables are used. Once the regression model is computed for the sample locations, the equation is re-entered into the GIS. The GIS then applies the model to the entire study area and computes a probability score describing the potential of each 30 m2 land parcel to contain archaeological material based upon the environmental characteristics of each parcel. Archaeological data were provided by the Kansas State Historical Society (KSHS) site database, which represents a detailed compilation of recorded archaeological sites in Kansas. Environmental data were derived from both public sources (USGS and State of Kansas) and from prior work completed at the University of Kansas and Kansas Geological Survey (playa-lunette system distribution, locations of springs). Model testing and validation use additional samples of archaeological sites from the KSHS database not used in the original model development, as well as data emerging from on-going archaeological surveys in the area.