Interpretation of multispectral satellite data as a tool for detecting archaeological artifacts (Navkur Plain and Karamleis Plain, Iraq)

Abstract

Contemporary studies of geographical space, including archaeological research, incorporate multiple spatial digital data. Such data provide an opportunity to extend research to large areas, and to objectify studies on the basis of quantitative data thus obtained and gaining access to the hard-to-reach study area. Examples of such data are satellite images at various spatial resolutions and in a wide spectrum of electromagnetic radiation (visible, infrared, and microwave). The authors made an attempt to use satellite images to analyze the areas of probable location of the Battle of Gaugamela (the Navkur Plain and the Karamleis Plain in Iraq). The photointerpretation was performed, enhanced by the multivariate processing of the multispectral image. The aim of the work was indicating the most likely places where the camp and the battle were located based on the visual interpretation of an array of satellite data. The adopted methodology of precise allocation of interpretative values to remote sensing materials for every detected artifact provided an opportunity to accumulate an extensive amount of information. It also provided the basis for a synthetic analysis regarding the methods of image processing on the one hand and the dates of recording on the other. It turned out that the season in which the photos are recorded is very important—although the best data for analysis turned out to be the autumn data (38% of all recognized artifacts), the use of data from three seasons increased the total number of indicated artifacts by as much as about 50% (the so-called unique detections). In addition, advanced image processing (such as principal component analysis and decorrelation stretch) turned out to be important, as it increased the number of areal artifacts by 31% compared to the interpretation of only photos in natural (true) color composite and false color composite (with near-infrared). The conducted analyses have confirmed the usefulness of high-resolution satellite data for archaeological applications, and the detected and described anomalies visible in satellite images are excellent material for selecting sites for detailed field research.