Jerba Archaeological Survey

The archaeological survey of the Island of Jerba, Tunisia was the first systematic study of the historical landscape of the island.  The goal of the project was to reconstruct the island’s settlement history role in trade and political formation from the earliest archaeological evidence for settlement in the Punic era to Early Historical times. The field survey was directed by Professor Renata Holod (University of Pennsylvania) and Dr. Elizabeth Fentress from 1996 to 2001.  Michael Frachetti was the project cartographer and GIS specialist.  One of the primary aims was to estimate the density of occupation over the whole of the island's history. To do this, we needed to establish reliable figures on which to base such an estimate.

Survey Methodology:
The survey data was managed and analyzed using a Geographic Information System (GIS), structured as a layered combination of map themes and a relational database. The basic map layers include topography and soils, hydrology, survey surface collection data, and site data. As there was no easily adaptable source of digital map data for the island, all of the primary map layers had to be scanned and digitized or drawn in AutoCAD and subsequently imported into Arcview. These primary themes were subsequently queried and analyzed to produce a range of more specific and detailed maps related to specific questions.

The site map itself was generated from the site database, which includes a Universal Transverse Mercator coordinate location for each site. Other primary data layers include broad ceramic typologies and their distribution (Punic, Roman, Mediaeval, Early Modem), as well as types of vegetation (olives, palms, gardens, scrub, etc.). In addition to the surface collection hectares, identifiable features (generally 'sites') were recorded using GPS and subsequently projected as point feature maps. Like the hectares, the characteristics of each site are represented by a database entry.

The site database was designed to identify individually and describe each site, indexed by its unique number. The site database is spatially linked to the hectare data through a hectare field reflecting the unit in which the site lay. The ceramic database represents a "one entry per  sherd" structure, indexed by the site or hectare number that corresponds to the survey provenience of the sherd. Thus, there may be multiple sherd entries for each locale, according to the nature of the material recovered.

This 'nested' structure allows us to query the characteristics of the cultural landscape by the hectare at the broadest scale, by site at the intermediate scale, and by sherd at the most detailed scale.


Discovery Highlights:

Survey data was managed and analyzed by using a Geographic Information System (GIS) and structured as a layered combination of map themes and a relational database. Basic map layers include topography and soils, hydrology, survey surface collection data, and site data. Other primary data layers include broad ceramic typologies and their distribution (Punic, Roman, Mediaeval, Early Modem), as well as types of vegetation (olives, palms, gardens, scrub, etc.).