Quantitative responses of satellite-derived night-time light signals to urban depopulation during Chinese New Year

Abstract

ABSTRACTAs an efficient proxy measure for various kinds of human activity, remotely sensed data of night-time light (NTL) have been widely studied. The connection between NTL and human population dynamics over fine time/space scales, however, is less understood. In this study, using NTL data derived from the Visible Infrared Imaging Radiometer Suite (VIIRS) and population dynamics sensed by the largest Chinese social media platform, we performed a nationwide analysis regarding whether NTL signals respond in a timely manner to provisional urban depopulation during the Chinese New Year holiday. Our results show that nearly three quarters of urban pixels with decreased NTL brightness show a marked decline in population, and more than half of the pixels showing reduced population can be identified through dimmed night-time lights. Moreover, our findings indicate that there might be a log-linear positive correlation between the magnitude of urban depopulation and the reduced brightness of NTL at both the pixel and city levels. Our findings suggest the overall usefulness of current VIIRS NTL data for investigating population dynamics at short time scales in terms of spatial extent and magnitude.