Urban intra-night artificial light dynamics revealed by multi-platform synergistic observation with different overpass times

The hourly dynamics of artificial light at night (ALAN) mirror the intra-night urban human activity rhythm, yet remain a challenge due to the static, snapshot-based nature of traditional remote sensing. In this study, we utilized multi-temporal nighttime light (NTL) imagery combined with building data to evaluate intra-night ALAN dynamics in Shanghai, China. We analyzed the migration of ALAN hotspots throughout the night, as well as NTL radiance variations across different building functions and heights. Then, we utilized principal component analysis (PCA) to identify the heterogeneity of ALAN dynamics at different spectral bands. Results reveal an obvious migration of ALAN hotspots from the commercial core in the evening to the industrial and logistic nodes and the main road network at night. This spatial shift can be attributed to functional heterogeneity, where business buildings record the highest initial radiance but decay by 74% before midnight, in contrast to the significant stability observed in industry and office buildings. Meanwhile, high-rise buildings record higher radiance and faster decay rates. In addition, the blue band exhibits the most significant magnitude of temporal decay. The decay rate of the blue band of business buildings during the period from 21:10 to 23:42 is around 85%. This disproportionate spectral decay eventually results in the evolution of initially diverse spectral characteristics into a uniform state. In conclusion, this research demonstrates that urban ALAN dynamics are fundamentally characterized by the complex interplay of socioeconomic function, vertical morphology, and lighting technology.