The predictive modeling approach is based on the premise that late Pleistocene people, like their modern counterparts, depended on important environmental factors in the selection of locations used for occupation. Geographic Information Systems data allow us to simultaneously consider various kinds of information (geologic, soils, cultural) that affect the probability of finding sites or strata of the appropriate age. While the area within which this project is searching is large, the use of modeling techniques, all conforming to the same scale and projection, may reduce the job to a manageable size and make surveys on the ground more focused. Use of GIS does not preclude use of traditional literature searches for information about excavated or named sites, and it builds on what is known in the anthropological literature about landscape characteristics that people choose for their sites.

     Michele Punke defended her Master's thesis, a computer modeling exercise focused on a portion of this Sea Grant project corridor, in May, 2001. The title of her thesis is Predictive Locational Modeling of Late Pleistocene Archaeological Sites on the Southern Oregon Coast Using a Geographic Information System (GIS) and is available from the Oregon State University Valley Library.

     Example of figure from thesis project Predictive Locational Modeling of Late Pleistocene Archaeological Sites on the Southern Oregon Coast Using a Geographic Information System (GIS).