Building A Predictive Model For Paleoindian Archaeological Site Location Using Geographic Information Systems

This research is a multi step method to predict unknown Paleoindian archaeological site locations within Pine Bluffs, Wyoming, situated in the southeastern corner of the state, using a Geographical Information System (GIS). The GIS technology is being used to predict Paleoindian archaeological site locations and will help demonstrate the geographic similarities and differences between already known Paleoindian archaeological sites and random non-site locations in the Pine Bluffs region. Using GIS, one can note the similarities and differences between the Paleoindian sites and the surrounding landscape and, with the help of logistic regression analysis, one can predict the location of unknown Paleoindian sites.

Although archaeologists are using GIS technology at an accelerating rate, publication of their work has not kept pace. A state-of-the-art exploration the subject, GIS and Archaeological Site Location Modeling pulls together discussions of theory and methodology, scale, data, quantitative methods, and cultural resource management and uses location models and case studies to illustrate these concepts. This book, written by a distinguished group of international authors, reassesses the practice of predictive modeling as it now exists and examines how it has become useful in new ways. A guide to spatial procedures used in archaeology, the book provides a comprehensive treatment of predictive modeling. It draws together theoretical models and case studies and explains how modeling may be applied to future projects. The book illustrates the various aspects of academic and practical applications of predictive modeling. It also discusses the need to assess the reliability of the results and the implications of reliability assessment on the further development of predictive models. Of the books available on GIS, some touch on archaeological applications but few cover the topic in such depth. Both up to date and containing case studies from a wide range of geographical locations including Europe, the USA, and Australia, this book sets a baseline for future developments.

Dutch archaeology has experienced profound changes in recent years. This has led to an increasing use of archaeological predictive modelling, a technique that uses information about the location of known early human settlements to predict where additional settlements may have been located. Case Studies in Archaeological Predictive Modelling is the product of a decade of work by Philip Verhagen as a specialist in geographical information systems at RAAP Archeologisch Adviesbureau BV, one of the leading organizations in the field; the case studies presented here provide an overview of the field and point to potential future areas of research.

This book describes a novel machine-learning based approach to answer some traditional archaeological problems, relating to archaeological site detection and site locational preferences. Institutional data collected from six Swiss regions (Zurich, Aargau, Grisons, Vaud, Geneva and Fribourg) have been analyzed with an original conceptual framework based on the Random Forest algorithm. It is shown how the algorithm can assist in the modelling process in connection with heterogeneous, incomplete archaeological datasets and related cultural heritage information. Moreover, an in-depth review of past and more recent works of quantitative methods for archaeological predictive modelling is provided. The book guides the readers to set up their own protocol for: i) dealing with uncertain data, ii) predicting archaeological site location, iii) establishing environmental features importance, iv) and suggest a model validation procedure. It addresses both academics and professionals in archaeology and cultural heritage management, and offers a source of inspiration for future research directions in the field of digital humanities and computational archaeology.

The Greater Yellowstone Region was a destination for nomadic hunter-gatherers for at least 12,000 years. Archaeological sites representing the whole spectrum of time, cultures, and activities, have been found throughout the region. Within Yellowstone National Park a number of Paleoindian projectile points and other related cultural materials have been recorded, however, only a handful of buried Paleoindian sites have been identified and excavated. Considering the nature of the archaeological record in the area, some interesting questions surface about the value of the information recorded on the Paleoindian sites. In terms of Yellowstone National Park (YNP) Paleoindian archaeology, is it possible to use the existing Paleoindian sites to make inferences about the landscape choices of Paleoindian cultures? Can the relationship between the location of known Paleoindian sites and the environment be modeled using quantitative methods? If so, is it possible to use the information about land use patterns derived from a known set of sites to find additional, currently unknown, Paleoindian sites? This paper attempts to answer those questions through the development of an archaeological predictive model, focused on Paleoindian sites, for Yellowstone National Park. Utilizing Geographic Information Systems (GIS) and statistical software, a probability model has been created that relates the existence or nonexistence of Paleoindian cultural materials with sixteen selected environmental features. The model output classifies areas within YNP through a set of environmental characteristics favorable for finding Paleoindian cultural material.

Archaeologists have employed Geographic Information Systems (GIS) software in the generation of predictive models for over thirty years. In the interest of creating a state wide predictive model, the North Carolina Department of Transportation (NCDOT) commissioned a pilot study in seven counties (Cabarrus, Chatham, Forsyth, Granville, Guilford, Randolph and Wake) of the Piedmont region. The primary goal of this thesis was to quantitatively examine the applicability of the Piedmont model to the Coastal Plain environment, specifically Pitt County. This thesis has demonstrated that the Piedmont predictive model does translate well to the Coastal Plain. Additionally, the predictive power of a model employing a generalized archaeological database (the Coastal Plain Model) was tested against a model employing a time period specific archaeological database (Coastal Archaic and Coastal Woodland models, respectively). The Coastal Archaic and Coastal Woodland models proved to have more predictive power than the Coastal Plain. A third research question analyzes the settlement decisions of archaic and woodland groups which are inferred from statistical data.

The search for archaeological materials dating to 15,000 yr BP along the southern Oregon coast is a formidable task. Using ethnographic, theoretical, and archaeological data, landscape resources which would have influenced land-use and occupation location decisions in the past are highlighted. Additionally, environmental data pertaining to the late Pleistocene is examined to determine what landscape features may have been used by human groups 15,000 years ago and to determine how these landscape features may have changed since that time. These landscape resource features are included in the modeling project as independent variables. The dependent variable in this modeling project is relative probability that an area will contain archaeological materials dating to the time period of interest. Two predictive locational models are created to facilitate the search process. These models mathematically combine the independent variables using two separate approaches. The hierarchical decision rule model approach assumes that decision makers in the past would have viewed landscape features sequentially rather than simultaneously. The additive, or weighted-value, approach assumes that a number of conditional preference aspects were evaluated simultaneously and that different environmental variables had varying amounts of influence on the locational choices of prehistoric peoples. Integration of the data and mathematical model structures into a Geographic Information System (GIS) allows for spatial analysis of the landscape and the prediction of locations most likely to contain evidence of human activity dating to 15,000 years ago. The process involved with variable integration into the GIS is delineated and results of the modeling procedures are presented in spatial, map-based formats.

Analyzed and modeled the distribution of archaeological sites in the region known as the Black Bottom of southern Illinois by investigating settlement patterning using Geographic Information Systems to look at the relationship between archaeological site locations and environmental variables and by creating an archaeological site prediction model to identify areas with a high potential for yielding previously undocumented archaeological sites.