Santa Rosa Island Archaeology: A GIS Predictive Model

Santa Rosa Island Archaeology: A GIS Predictive Model By Lauren DeOliveira Master of Science in Geographic Information Science More archaeological projects are utilizing Geographic Information Systems (GIS) to aid in intensive pedestrian survey endeavors. This thesis presents the results of two types of analysis, point pattern analysis and probability modeling. First, point pattern analysis using archaeological site locations on Santa Rosa Island, located within the California Channel Islands, was analyzed using centrography and density-based analysis. Specifically, mean center, standard distance, quadrat analysis, the statistical calculations of Variance Mean Ratio and Nearest Neighbor Analysis along with Kernal Density are used to identify patterns in the spatial distribution of known archaeological sites on Santa Rosa Island. The results of the Variance Mean Ratio concluded that clustering is present within the dataset and that the clustering did not occur by chance. Furthermore, Nearest Neighbor Analysis indicated a clustered distribution that is statistically significant, strengthening the initial conclusion of Variance Mean Ratio. Kernal Density analysis indicated that archaeological sites are clustered in the north, west and east coastal regions of Santa Rosa Island. This clustering pattern ultimately suggests that outside independent variables are influencing the location of archaeological sites on the island. I believe those independent variables are environmentally based. Second, two environmental variables, hydrology, and typography, are analyzed to predict the locations of archaeological sites on Santa Rosa Island. Site prediction using these variables was completed to aid the National Park Service, and future researchers, in cultural resource management by providing sensitivity maps to be used out in the field. This analysis concluded that high probability areas are within 419 meters of a visible fresh water resource and on a 0-6.8% slope. Moderate probability areas are within 839 meters of a visible fresh water resource and on a 6.8%-18.3% slope while low probability areas are further than 839 meters from a fresh water resource and have a slope greater than 18.3%. These sensitivity maps can ultimately be used to coordinate and execute targeted archaeological surveys on Santa Rosa Island with the hope of discovering and documenting unknown archaeological sites. By having the ability to use probability maps to complete targeted intensive pedestrian surveys, archaeologists have the chance to document unknown sites, further adding to the archaeological database and increasing the knowledge of settlement patterns and the movements of the Chumash on Santa Rosa Island. The results of this research build on our understanding of settlement patterns on Santa Rosa Island and California's Channel Islands as a whole.