Model Evaluation

Literature and Link

Project Staff

Project Home

Project Overview

Data Acquisition

Model  Development

Willamette Industries, Inc. is a diversified, integrated forest products company with a number of manufacturing facilities located in several states and overseas. In South Carolina, Willamette owns approximately 108,000 acres on 375 properties covering 111 7.5’ DOQ’s.  The company, aware of its stewardship responsibilities towards natural and cultural resources on its property (Willamette website, 2001), contracted with the South Carolina Institute of Archaeology and Anthropology (SCIAA), University of South Carolina, to:

  1.      Assess known archaeological sites located on its lands and

2.      Identify tracts which may contain significant cultural resources through predictive modeling

The objectives of the work were to provide Willamette with an initial data set on which to base immediate management decisions while planning for the future.  This work was to be completed within 6 months on a limited modeling budget.

Site assessment and Identification of Known Sites

This project did not fall under the strict legal guidelines of Section 106 of the National Historic Preservation, which stipulates a very formal method of survey and evaluation.  This allowed SCIAA to loosely define the archaeological significance of sites on Willamette holdings, and for the purposes of this study, "significance" was defined on three levels (Table 1). 

Table 1: Definition of Site Significance in the Project

Willamette's landholdings were examined in relation to known archaeological sites listed in the South Carolina Archaeological Site Files using GIS data layers provided by

Willamette and those maintained at SCIAA.  Based on these data, 58 known sites were noted on, in whole or in part, Willamette holdings.  Many of these sites were recorded years ago using  standards that are difficult to assess within the parameters described above. SCIAA used a modified version of the formal evaluation procedure required by law to evaluate the sites based on National Register of Historic Places criteria.  An example of a formal site form is available on the SCIAA web site (SCIAA, 2001). Known sites on Willamette holdings were identified as Level I, Level II, or Level III by examining site form data.  Site form data fall under major headings that include general information, environment and location, site characteristics, archaeological components, methods employed during discovery, and data recovered.

The Need for a Predictive Model for Archaeological Sites

In addition to the known sites on Willamette property, many other sites have yet to be identified.  Predictive modeling would give Willamette the option to avoid areas where significant resources are likely to exist until field investigation could confirm their presence.

The premise behind modeling is that historic and prehistoric peoples were closely tied to their natural and cultural environment, and that these environments were a significant determinant in their choice of site location.  Predictive modeling examines soils, distance to water, and slope as potential natural variables, and subsistence systems, transportation systems and previous settlement as potential cultural variables.  The intended outcome was a spatial depiction of probability zones outlining geographic areas with a high likelihood of Level I or Level II resources, correlated with individual company tracts of land.

GIS systems allow the overlaying of spatial data from different sources, using different structures and resolutions, thus providing a tool for modeling spatial data (Goodchild, 1993).  Willamette’s desire to be good stewards presented SCIAA with an ideal opportunity to develop and test a simple predictive model on a large scale (111 USGS 7.5” quadrangles) against the 58 known archaeological sites on Willamette property.  SCIAA partnered with the Earth Sciences and Resources Institute of the University of South Carolina (ESRI-USC) to build the GIS-based model through its competency in the area of GIS modeling of hydrologic and conservation systems (Rine and Covington, 1999; Covington et al, 2000).  SCIAA model parameters (Table 2) were used as a starting point for the process.

Table 2: Initial modeling parameters for spatial predictive model to determine archaeological sites.