Fast and accurate monitoring of economic dynamics is essential for evaluating economic policies and informing effective strategies, especially when timely official gross domestic product (GDP) statistics are unavailable. In this study, the capability of time-series nightlight satellite data to model urban economic activity was assessed by integrating it with observed GDP statistics. Experimental results showed that, when nightlight information was elaborately fused with core urban land-use data, urban GDP dynamics could be modeled with satisfactory accuracy. Model-based estimates indicated that China’s average GDP growth rate was approximately 4% across cities, with pronounced fluctuations during the pandemic, especially in 2020 and 2022. The proposed approach further quantified that the actual GDP losses in China due to the pandemic were about 1.5% of the total GDP, and the simulated results suggested that many coastal and industrial cities faced even more severe impacts than major provincial cities. Overall, the findings demonstrate the value of nighttime remote sensing for monitoring urban economic dynamics and support the transferability of the proposed approach to other countries with limited or delayed GDP statistics.
Xu et al. (Mon,) studied this question.