Abstract: For more than forty years, scholars have explored the idea that cities conform to a relatively predictable mathematical growth pattern with respect to total occupied area and total population, which then also predictably relate to several other aspects of urban living such as the amount of critical infrastructure, the intensity of social interactions, trade volume, total income, gross domestic product, and even the number of patents issued--a proxy indicator for the level of innovation present. In other words, if you have information about at least one aspect of a city that has been examined in this way, you should be able to infer others via a set of straightforward equations that leverage occupied area, population, or both, allowing you to move between them as long as the relationship between total occupied area and total population is well understood. Where our current understanding of the pattern and its relationships can be significantly enhanced is in having higher quality data on occupied area and population for cities around the world. Will the pattern and its relationships still hold when the entire world is taken into account? Are the proposed equations and the current empirical estimates of their coefficients still valid? Will something new emerge? This presentation will focus on a study that sought to answer these questions by exploring the research space from the ground up, leveraging high performance computing and the best spatially-explicit global population datasets available to organically define urban regions, which has never been attempted before, and then derive the equations and associated coefficients that best explain them. Some of the results were anticipated, but others were quite surprising and potentially allow for a much more sophisticated understanding of human settlement patterns and social complexity across both space and time.
Bio: Devin White (Ph.D. 2007, University of Colorado) leads the Autonomous Sensing and Perception Department at Sandia National Laboratories in New Mexico and is a Research Assistant Professor of Anthropology at the University of Tennessee, Knoxville. His research interests include machine intelligence, photogrammetry, remote sensing, computer vision, imaging science, geographic information science, computational social science, high performance computing, modeling human movement and visibility across landscapes, complex adaptive systems, and Southwestern archaeology. Prior to joining Sandia, he was the Distinguished Scientist for Photogrammetry and Remote Sensing at Oak Ridge National Laboratory in Tennessee.