Using GIS to Explore Legacy Spatial Data at Isthmia

0 downloads 0 Views 5MB Size Report
(GIS), which combine digital databases with scalable electronic maps to .... Timothy E. Gregory, Director of the Ohio State University Excavations at Isthmia.
Using GIS to Explore Legacy Spatial Data at Isthmia Louise M. Steele and Jon M. Frey, Ph.D Department of Anthropology, Michigan State University 1

2

1

Department of Art, Art History, and Design, Michigan State Univeristy

2

As is the case at many excavations that had their start before the “digital revolution,” the Ohio State University Excavations at Isthmia must face the challenge of updating its traditional paper-based recording systems in order to take advantage of the many digital tools that are available for archaeological data analysis. This is most clearly illustrated in the case of Geographic Information Systems (GIS), which combine digital databases with scalable electronic maps to enable statistically significant visual analyses of the types of complex datasets common to archaeological sites. In order to serve as an effective tool for archaeological analysis, a GIS must rely on precisely mapped features in combination with archaeological evidence in electronic form. Yet at Isthmia, these forms of information are “locked up” in handwritten notes and drawings contained in a variety of notebooks, maps and plans. This poster outlines the process by which that analog information is being made digitally acessible and therefore useful for spatial analysis.

The Problem This project builds on the work initiated last season to create a unified, scalable, digital model of the Sanctuary of Poseidon at Isthmia. Previously, features at the site were represented in a number of actual state plans and less accurate overall site plans that could not be directly related to one another. Moreover, even the most accurate plans were created using an outdated projection system and datum that could not be easily related to GPS data collected in a more universally relatable sytem. This shortcoming was remedied by means of an intensive aerial survey that produced a centimeter accurate, geospatially referenced orthomosaic of the site, which could serve as a base map onto which all prior plans of the sanctuary and its larger regional setting could be projected.

Artifact Data

The Process Georeferencing the Plans

Adding the Sketches

In the first stage, digitized copies of hand-drawn actual state plans are modified to achieve a “best fit” with the newly created orthophoto mosaic of the site. Notice the small distortions in the semi-transparent state plan.

Next, these plans as well as the real world features that remain visible since the time of excavation are used to geospatially reference the sketches of excvated trenches with their sketch-specific coordinate systems.

Plotting the Points

Because all information from previous decades was documented on paper, data about the objects themselves must also be converted to electronic format. As is the case when plotting the spatial data, this step in the process is also done manually. In this case though, the Isthmia Excavation makes use of its installation of the Archaeological Resource Cataloging System (ARCS), to organize an online, crowd sourced effort to transcribe the essential details about each object from digitized copies of original inventory catalogue cards. Once this stage is complete, information can be exported from ARCS in a format that is easily integrated into the site’s GIS.

Figure 6: Screen caputre of the data transcription tool in the Isthmia ARCS

Next Stages Much remains to be done with other field journals both in the area of the Fortress as well as other more systematically excavated areas of the site. Yet already, this project has identified the following items in need of more attention:

Figure 1: (Left) Field journal with informal grid system; Figure 2: (Center) Field journal with artifact location and catalogue number Figure 3: (Right) Plan of the Tower 14 region of the Fortress at Ithmia

Having succeeded in creating a unified digital basemap of the site, it became possible to focus on the associated problem of locating catalogued artifacts in and around the sanctuary. Lacking a systematic grid system for their site, archaeologists at Isthmia have traditionally relied upon locally generated plans to map the locations of objects and features uncovered in excavations. These plans, which are accompanied by an informal grid system, appear in the opening pages of most field journals and are referenced throughout the discussion that follows. Of course, these sketch plans differ in quality from journal to journal and lack the overall accuracy of an actual state plan -- a factor that must be addressed when attempting to generate reliable data for statistically significant spatial data. Actual references to the findspots of objects in the field journals are of two types. Occasionally, for artifacts understood to be important at the moment of discovery, excavators provided coordinates referring to a specific plan in that field journal. More commonly, objects recognized as significant at a later stage in the cleaning and recording process are simply associated with a stratagraphic unit. While this was not always the case, in both scenarios, the final catalogue number of the object is supposed to be wrtiten into the field journal as a part of the cataloging process. Thus, as a first step, researchers are focusing on identifying all references to catalogued artifacts in the excavation journals, then estimating each object’s place of discovery, in order to create the first overall plan of artifact scat-

Acknowledgements This project could not have been possible without the support of the following individuals and organizations: • Timothy E. Gregory, Director of the Ohio State University Excavations at Isthmia • Betsy Gebhard, Director of the University of Chicago Excavations at Isthmia • James Herbst, Architect, American School of Classical Studies at Athens Excavations at Corinth • Amanda Tickner, GIS Librarian, MSU Libraries • MATRIX Center for Digital Humanities and Social Sciences • MSU College of Arts and Letters Undergraduate Research Initative I would also like to thank Dr. Frey for his mentorship and advising during this project.

Finally, artufacts that have been located according to sketch-specific coordinate systems are plotted into the GIS. Note the systematic use of a naming convention, which often must be imposed on inconsistent data.

Preliminary Results

There can be no doubt that this is a very time-consuming process. In order to include the greatest amount of detail, information about each artifact must be aggregated from a variety of sources. Evidence concerning objects first discussed in a field journal is augmented with details written onto the catalogue cards and ocasionally artifact drawings, photographs or other forms of process metadata. This would suggest that, into the near future, few aspects of this workflow will be able to be automated.At the same time, even though this project is only in its preliminary stages, the results are already promising. A comparison of the location of coins that have been assigned a data reveals: • The eastern and southwestern parts of the ffortifications exhibit a fairly uniform distribution of artifacts • In the area of Tower 15, Byzantine (6th-13th c AD) coinage far outnumbers that of other artifacts. unlike the eastern and southwestern parts of the fortifications, which exhibit a generally uniform distribution of artifacts • In the area of Tower 14, Frankish (13-15th c AD) and Venetian (14th-16th c AD) coins are more commonly found, especially in the area of a one-room building appended to the western face of the Byzantine fortification. Figs. 4 and 5: Comparison of the Tower 14/Tower 15 ara on the west side of the Fortress(left) with the Tower 8/ Tower 9/ Tower 10 area on the south (right). Note the concentration of the predominately Byzantine coins near Tower 15 and the concentration of Venetian and Frankish coins near Tower 14. This periodization is not seen, however, in the Tower 10 area.

• Because recording inventoried objects in the midst of an active field season was a haphazard process, references to many artifacts did not find their way back into notebooks. Thus, it will be important to examine all inventoried object cards to extract any geospatially signifcant evidence that has not yet been recognized. • Because the locations of some artifacts were described with greater precision than others, it will be necessary to develop a more effective way to indicate visually a relative degree of confidence in the locations that make up the artifact scatters on the plan. What is the best way to represent the location of an artifact that can only be identified as part of a stratigraphic unit? • Because some excavators were more careful to record any artifact of potential significance, the resulting plots of objects may be less an indication of actual artifact densities than individual practices of collection and recording. It may be worthwhile to adopt techniques developed among survey archaeologists to address inter-observer error. • Because most GIS systems only offer a two-dimensional approximation of a three-dimensional reality, it is important to consider how to best represent stratigraphic layers in examining spatial relationships. At present, the scatter of artifacts across the site is potentially misleading, as it fails to distinguish between areas that were truly more “artifact-rich” and areas that were simply excavated more deeply through more strata of human activity.

Figure 7: Mosaic of all field journal sketches in the Tower 14 area of the Fortress at Isthmia

Contact Information: [email protected] louisemsteele.com