Border identification and spatial evolution of metropolitan areas by surface network analysis using nighttime light data

Metropolitan areas are described as large population centers consisting of a central city and its surrounding suburbs and exurbs. Traditionally, metropolitan areas are delineated based on travel time to a fixed urban center, without considering the dynamic evolution of urban centers. As a result, long-term analyses of the spatial interactions of metropolitan areas within urban agglomerations may be biased. Therefore, we propose a surface network method to identify the evolving metropolitan areas. In this surface network, nodes and edges—representing urban centers and their connections—are identified from nighttime light (NTL) data, with edge weights representing travel times. We then divide the surface network into sub-networks based on edge weights, each of which is considered a metropolitan area. Using the Yangtze River Delta Region (YRDR) as a case study, we examined the development trajectories of its metropolitan areas over a 20-year period at five-year intervals, focusing on both agglomeration-oriented and diffusion-oriented pathways. We found a significant increase in the number and size of metropolitan areas within the YRDR, which experienced an evolutionary process of “agglomeration–diffusion–re-agglomeration” that closely aligns with the region's strategic development plan. This study provides complementary insights into the evolutionary dynamics of urban systems and may serve as a useful reference for regional-scale urban planning. • A surface network-based method was developed to identify metropolitan areas using nighttime light and commuting data. • From 2000 to 2020, Yangtze River Delta Metropolitan areas evolved through agglomeration, diffusion, and re-agglomeration. • Our findings provide insights for policymakers on the spatial evolution of metropolitan areas.