GIS and archaeology : investigating source data and site patterning
Using a Geographic Information System (GIS), locational analysis was performed for prehistoric sites recorded during a 1985 surface survey conducted in Henry County, Indiana. Two sensitivity models were developed to identify areas more likely to contain substantial archaeological resources. Both models were based on environmental data derived largely from soil survey information. An intuitive model was created and "blindly" applied to the study area. This model did not interpret the distribution of sites very well. During development of an alternative model, the 1985 survey data was more thoroughly investigated. Site locations were found to be correlated with Soil Conservation Service drainage categories. In upland areas, sites with ten or more artifacts clustered around pockets of very poorly drained Millgrove loam soils. In lowland areas, sites with ten or more artifacts exhibited a preference for well drained soils. Before and during analysis, the integrity of source data was investigated. A United States Geological Survey 7.5-minute digital elevation model was found to be unsuitable for analysis within the study area. Mapping errors were discovered within the 1985 survey data. Global Positioning System (GPS) technology, which can increase the spatial integrity of survey data, was demonstrated and used to register and adjust source data. A mapping-grade GPS base station was established at Ball State University.