Wavelength-specific urban nighttime light modulates expressed sentiment across China
摘要
As cities brighten under rapidly expanding lighting infrastructure, residents face increasing sentimental risks at night. Although artificial light at night is linked to mental health risks, its wavelength-specific effects remain poorly understood. Here we integrate the satellite observations of artificial light at night and sentiment analysis of geotagged social media posts to quantify wavelength-specific associations between urban light exposure and expressed sentiment across China. We find that exposure to blue light (424–526 nm) increases negative sentiment by 15.9%, whereas moderate green light (506–612 nm) enhances positive sentiment by 5.0%. Spatial mapping further reveals pronounced disparities both within and between cities, with elevated sentiment risk concentrated in commercial areas and eastern cities of China. In particular, optimizing correlated color temperature reduces sentiment risk by 89.7% compared with intensity reduction alone. These findings provide evidence-based guidance for urban lighting design to promote public mental wellbeing through spectral optimization, not just brightness control.