| dc.description.abstract | This dissertation presents advanced methods in data processing, statistical analyses, integration, and visualization of archaeogeophysical data to increase the accuracy of archaeological remote sensing interpretation and predictions. Three case studies are presented from an experimental controlled archaeological test site and two nineteenth century historic military archaeology sites at Paint Rock, Texas and Alcatraz Island, California. I demonstrate the ability of the Geonics EM-63 time-domain electromagnetic-induction metal detector to detect and localize historical metal artifacts at an experimental site and Paint Rock. Moreover, point pattern analysis spatial autocorrelation statistics were used to detect statistically significant patterns that spatially compacted the amplitude response of the data to improve the localization of artifacts of archaeological significance. The archaeological data was used to determine the spatial and temporal extent of the military camp at Paint Rock and conforms well to the historic record. A virtual ground-truthing was conducted at Alcatraz Island, where the results of a quantitative attribute analysis of ground-penetrating radar data was tested against the georectification of historic maps in order to determine the location, extent, and integrity of historic military features without excavation. These studies increased the information content of archaeogeophysical data via feedback with statistics, quantitative attributes, controlled experiments, excavation, and georectification modeling in order to increase the predictive capabilities of the methods to answer the most questions with the least amount of costly excavations. | en |
