|
The
South Carolina Institute of Archaeology and
Anthropology (SCIAA) and the Earth Sciences
and Resources Institute of the University of
South Carolina (ESRI-USC) collaborated to
develop a predictive model for likely
archaeological sites using available data
sources. The study area covered 111 7.5'
Digital Orthophotographic Quadrangles or
roughly one-third of the area of the State
of South Carolina. The model made extensive
use of floodplain, hypsographic, and Digital
Elevation Model (DEM) data.
Throughout time, people have been closely tied to
their natural and cultural environments;
these environments were a significant
determinant in their choice of settlement.
An understanding of the natural and
cultural environment related to past peoples
and their settlements, therefore, allows the
archaeologist to make predictions about
similar, but as yet undiscovered sites.
For instance, historic sites have
traditionally been located along old roads. Gillam (2000) was able to predict historic sites with some
accuracy by using historic map data,
rectified to and overlaid on current USGS
7.5’ Digital Othrophotographic Quadrangles
(DOQ’s). This method requires the creation
of new data through the scanning and
rectifying of historic maps, location of a
number of ground control points for
rectification and finally, the digitization
of salient features.
Prehistoric settlements, however, are
usually associated with natural features,
the strongest being proximity to water. Other environmental variables that may be considered in the
location of prehistoric sites and combined
with proximity to water may include
topography (gentle slopes, small rises in
flood plains) or soil type.
This methodology is not new to the
archaeologist who would makes use of
a hard copy topographic map as the primary
tool to predict where good sites may be
located. In order to make these predictions
over a large area, covering many topographic
maps, the archaeologist would be greatly
aided by a GIS-based model. This article
documents the development and testing of a
parsimonious GIS-based model that predicts
the location of prehistoric settlements
using existing, publicly available data
only.
|
