Alternative Tessellations for the Identification of Urban Employment Subcenters: A Comparison of Triangles, Squares, and Hexagons

Despite the widespread use of equilateral square and hexagonal grids as spatial units in metropolitan spatial structure studies, their impact on the identification, selection, and categorization of employment subcenters remains poorly understood. This study looked at three contiguous regular polygons that tessellate the Euclidean plane—squares, hexagons, and triangles—and compared them using five indicators: employment density, employment-to-land ratio, number of subcenters, discrete compactness, and corner connections in subcenters at a fine scale. The goal was to find out what shape and size of grid works best for finding employment subcenters by looking at the employment density, employment-to-land ratio, number of subcenters, compactness, and corner connections of the identified employment subcenters at different scale sizes. The scale view technique in conjunction with the grid approach and the Local Moran’s I (LMI) approach were used in the identification, selection, and categorization of the employment subcenters. The analysis has revealed that, in addition to the shape and size of the grids, the edge effects—i.e., either corner or edge connections within the grids—have had a significant impact on the emergence of the number of subcenters, their employment-to-land ratio, density, and compactness. In addition, square grids have demonstrated an advantage in identifying and selecting employment subcenters compared to hexagonal and triangular grids. We make advancements by contributing to the techniques used to identify subcenters over time and by revealing the most appropriate shape and size of the grid for identifying and selecting employment subcenters across different scales.