Open Retail Centre Boundaries for the UK and Their Utility for Functional Analysis and Policy

This paper presents an updated methodology for delineating retail centre boundaries across the United Kingdom, building upon Macdonald et al. (2022). The approach integrates open source Foursquare Places point-of-interest (POI) data with OpenStreetMap land use polygons, employing hexagonal spatial aggregation (H3) and network-based connected component identification. Key improvements include Leiden community detection to partition large connected city or town centres into distinct subcentres and automated detection of car-oriented retail parks based on the presence of characteristic big-box chain brands. The centre boundary utility is demonstrated by a case study focused on night-time economy. This shows how consistent spatial definitions enable meaningful integration of heterogeneous data sources, including event listings and opening hours data, supporting nuanced assessment of urban evening economies across multiple dimensions of activity.