ANALYZING URBAN DYNAMICS BEFORE AND AFTER EXPANDING THE TRANSPORTATION NETWORK USING NTL DATA: A CASE STUDY OF UTSUNOMIYA LRT

抄録

 This study analyzes urban dynamics before and after the introduction of the Light Rail Transit (LRT) system in Utsunomiya, Japan, using Nighttime Light (NTL) data. NTL data, which captures artificial light intensity on the Earth’s surface, enables high-frequency monitoring of urban activity. Since land use data in Japan is updated only every five years, it is challenging to detect timely urban changes following infrastructure development. To address this limitation, we use monthly NTL data to assess both localized and citywide transformations associated with the LRT project.

 First, we conduct a localized analysis of NTL values within a 500-meter radius of newly constructed LRT stations, comparing changes across four key phases: pre- and post-announcement, construction, completion, and operation. This reveals station-level variations and the timing of urban activity shifts. Second, we apply Random Forest models trained on global land use datasets (GHSL, CORINE, NLCD) to predict annual land use changes in Utsunomiya using NTL, population, surface temperature, and Sentinel-2 imagery as explanatory variables. The dual-scale analysis demonstrates the utility of NTL data in capturing both micro- and macro-scale urban dynamics and provides a methodological basis for continuous land use monitoring in Japan.