ABSTRACT
Chapter 4 covers various methods for defining functional regions, within which encompassed areas are more closely connected than between. It begins with the proximal area method that assigns areas to their nearest facility, based on the commonly known “least-effect principle” in geography. As a result, a function region is made of areas sharing the same nearest facility. The Huff model accounts for the joint effects of facility sizes and their distances from residents and assigns areas to a facility being visited with the highest probability. Therefore, a function region includes areas being visited most often, or the “plurality rule.” The proximal area method or the Huff model is used for defining functional regions in absence of data on actual interactions between residents and facilities. The Dartmouth method, proposed for defining hospital service areas (HSAs), also uses the plurality rule but assigns a demand area to a facility based on actual origin–destination (OD) flow data. The network community detection approach represents the state-of-art development in delineating function regions by explicitly maximizing connections within derived regions and minimizing connections between them. Two spatialized network community detection methods, namely, ScLouvain and ScLeiden, are introduced, and both account for spatial constraints, such as area contiguity and threshold population size in derived regions. All methods are automated in ArcGIS tools, and their implementations are demonstrated in a case study of defining HSAs in Florida.
