Geospatial modeling of pedestrian transportation networks: a case study from precolumbian Oaxaca, Mexico
Abbreviated Journal Title
J. Archaeol. Sci.
Oaxaca; GIS; Movement; Networks; Pathfinding; High performance; computing; Least cost path; LEAST-COST-PATH; VALLEY; DISTRIBUTIONS; LANDSCAPES; HISTORY; TERRAIN; LAW; Anthropology; Archaeology; Geosciences, Multidisciplinary
Using accumulated cost surfaces and various pathfinding techniques within Geographic Information Systems (GIS) software, archaeologists and other spatial scientists have developed increasingly sophisticated models of human movement. Despite their utility, these approaches can be limited because standard GIS software cannot model movement (1) from many origins to many destinations or (2) without specific origins and destinations. Absent these capabilities, it is particularly difficult to model networks of movement over a given tract of land if you are interested in obtaining a more general sense of movement dynamics, not specific site-to-site patterns. In this paper, we present an innovative way of modeling past movement that generates both natural-looking networks and also indicates the degree of traffic that may have existed on any particular segment of those networks. The "From Everywhere to Everywhere" (FETE) model generates networks based on topography and landcover. without requiring that origin and destination points be supplied in advance. We apply the FETE model to a case from the southern Mexican state of Oaxaca, a region that has extensive archaeological and ethnohistoric data sets that serve as a test of the efficacy of our technique. A comparison of the FETE output with known late precolumbian and early colonial movement corridors indicates that the method is effective and should be useful for modeling networks in other areas. (C) 2012 Elsevier Ltd. All rights reserved.
Journal of Archaeological Science
"Geospatial modeling of pedestrian transportation networks: a case study from precolumbian Oaxaca, Mexico" (2012). Faculty Bibliography 2010s. 3484.