Geographic information systems (GIS), also known as computer mapping technologies, are digital databases that store, manipulate, capture, analyze, create, and display spatially referenced data. The spatial component of these systems make them readily applicable to archaeological research, where archaealogists record location information for everything from individual artifacts to sites in a regional survey. Many projects also record environmental data and such modern features as roads and villages. All of these data are spatially referenced and often best understood when stored as maps, an easy and relatively inexpensive task for a GIS.
Data for building a GIS can come from a number of sources. Remotely sensed data, such as satellite images, aerial photography, and geographic positioning systems, along with more traditional sources, such as paper maps and archaeological databases, can be used to construct a GIS. Unlike paper maps, which store multiple data themes on a single sheet, GIS store data themes in individual layers: a layer for roads, a layer for elevation, a layer for Iron Age sites. Because these map layers are referenced to a common coordinate system, they can be electronically overlaid to create any combination of data themes. The advantage of this type of system is obvious: maps can be quickly and easily revised as new data become available without the expense of traditional cartography.
GIS come in two basic types, vector and raster. Because of the differences in how they manage spatial data, each has its own particular strengths and weaknesses. Vector-based GIS are easily understood because, like paper maps, they store their data as points, lines, and polygons. These entities form the basic data units of a vector GIS, for which spatial information must be explicitly encoded. For example, if a road network is put into a vector GIS as a series of lines, specific X and Y coordinate data must be encoded for each line before the spatial relationships between the various road segments have meaning.
The strengths of vector-based systems lie principally in their familiar format, accuracy, high-quality cartographic output, connectivity to powerful databases, and the relatively small amounts of computer space they require. These strengths make vector-based systems ideal for managing data across large regions and producing detailed and accurate maps. For example, archaeological resource management on a state or national level would be an ideal application for a vector-based system, especially if publication-quality maps are needed. The weaknesses of a vector GIS stem primarily from the necessity to encode spatial data explicitly. This means that continuously changing surfaces, such as elevation and slope, must be generalized as elevation contours or polygons representing slope categories. These are ultimately unsatisfactory for many types of archaeological modeling.
In a raster-based GIS, data are stored in a grid of columns and rows, much like a spreadsheet, that represent X and Y coordinates in the real world. The intersection of each row and column is known as a cell. Each cell represents a specific area in the real world and contains a Z value, or number that can represent anything from elevation values, to sites, to soil types. Unlike vector systems, the basic data unit in a raster GIS is a spatial unit (the cell) for which entity information must be explicitly encoded. For example, if a road network is put into a raster GIS, the X and Y coordinates are implicit because they are determined by their position in the grid; however, Z values corresponding to the various components of the network must be explicitly encoded in each cell for them to have meaning as roads.
The strengths of raster systems are found principally in their ability to manage data that are continuous across a surface, in their simple data structure, and in their ability to use remotely sensed data, such as satellite images, which are stored in a raster format. These strengths make the creation of complex mathematical models possible. For instance, with a raster GIS, it is possible to reconstruct large portions of the paleoenvironment, especially terrain-related data, as continuous surfaces. Rather than representing elevation as a series of contours, each cell in a raster GIS contains a Z value corresponding to the elevation at that location. Because these surfaces are constructed of numbers in cells, they can be subjected to sophisticated mathematical manipulation to create new surfaces. Any equation that can be devised, such as surface I + surface 2 = surface 3, can be carried out in a raster GIS. The disadvantages of a raster GIS lie principally in file size and cartographic output. Raster files use large amounts of computer disk space. Even grids representing relatively small areas will contain tens of thousands of cells, and grids with hundreds of thousands of cells are common. Also, because of the raster structure, the traditional cartographic output of points, lines, and polygons is less accurate and of generally poorer quality.
An example of the types of analysis possible with a GIS can be drawn from the Tell el-῾Umeiri regional survey in Jordan. The ceramic record from this region indicates a substantial increase in human activity with the change from Iron I to Iron II. To investigate how this increased activity may have modified subsistence strategies, a raster GIS containing several environmental variables was constructed. Eight of the variables were selected and probability models based on a logistic regression were created. This modeling process involved a comparison of the sites from each period with a random sample of 250 nonsite locations. The environmental variables were then weighted, according to their relative importance as revealed by the regression, and summed to create models of the environmental signature for sites from each period. The models illustrated that during the Iron I period the areas considered environmentally suitable for settlement were limited, but that the pressures of increased human activities forced a change in subsistence strategies during the Iron II period, leading to an expansion into areas previously considered unsuitable.
[See also Computer Recording, Analysis, and Interpretation.]
- Allen, Kathleen M. S., et al., eds. Interpreting Space: GIS and Archaeology. London, 1990. .
- Burrough, P. A. Principles of Geographical Information Systems for Land Resources Assessment. Oxford, 1986. .
- Kvamme, Kenneth L. “Geographical Information Systems in Regional Archaeological Research and Data Management.” In Archaeological Method and Theory, vol. 1, edited by Michael B. Schiffer, pp. 139–203. Tucson, 1989. .
- Kvamme, Kenneth L. “A Predictive Site Location Model on the High Plains: An Example with an Independent Test.” Plains Anthropologist 37 (1992): 19–40. .
- Monmonier, Mark. How to Lie with Maps. Chicago, 1991. .
Gary L. Christopherson