ABSTRACT
This chapter provides a national picture of how individual cities and towns across the United States connect together in a functional way, in contrast to the formal political geography of cities, counties and states. This approach builds upon a tradition of studying polycentric urban regions, 'functional urban areas' and the concept of Gottman's 'megalopolis' more generally. The chapter shows how patterns of spatial interaction revealed through the mapping of large interaction datasets can be delineated algorithmically to produce a functional regionalization of the United States. In order to emphasize the revelatory power of a geovisualization approach to big data, the chapter also examines the underlying uncertainty inherent in the data. We then present the results of our algorithmic partitioning approach. The chapter uses the American Community Survey (ACS) data as an example owing to its volume, richness and apparent spatial granularity.
