Geographical information systems data implementation for analyses of settlement patterns of an early agricultural society

The use of geographic information systems (GIS) for handling locational data of objects and their spatial attributes conveys an unfounded sense of accuracy and an unquestioned perception of authority on whatever results are obtained from processing these data. Several researchers have already documented and warned about the dangers of overlooking limitations implicit to geographic data, of uncertainties and in some cases plain inaccuracies resulting from the fact that at best the data constitute an approximate model of reality. Awareness of data quality issues has lagged behind increasing usage of GIS for research in the spatial components of the social sciences. This situation has created the need to complement the application of spatial modeling techniques with a parallel emphasis on actual data limitations and their proper handling.; The overarching objectives of this dissertation are (a) to evaluate and prepare GIS data needed for research related to the environmental, socioeconomic, and cultural processes that shaped the observed settlement patterns of an early agricultural society, (b) to optimize the analytical and predictive potential of modeling these processes with enhanced data, (c) to support realistic expectations and to reduce the risk of incomplete or misleading results from such type of modeling, and (d) to perform pertaining data transformations to determine the intended landscape classification.; The instance of agricultural society chosen as focus of interest for this study is that of the ancient Maya because it constitutes a good example of an early civilization for which enough data is available to carry out the analyses mentioned. Environmental data sets are evaluated and prepared for at least one example of each of the raster and vector data models, namely digital elevation model (raster), soils (polygons), and archaeological sites' locations (points at the regional scale). Examples are described exploring the interactions between the environmental data sets enabled and the observed patterns of settlement.