From illumination expansion to ecological response: artificial light at night and detectability of forest vegetation responses across India

Artificial night-time light (NTL) is widely used as a satellite-derived proxy for anthropogenic activity, but its ecological implications at regional scales remain uncertain. This study examines the spatio-temporal expansion of NTL across India during 2012-2024 and evaluates its association with forest vegetation productivity, greenness, and phenology. VIIRS-derived NTL was analysed with satellite-based gross primary productivity (GPP), enhanced vegetation index (EVI), normalized difference vegetation index (NDVI), forest-type information, and phenological metrics. Total illuminated area increased from approximately 0.58 to 2.51 million km2, representing a more than fourfold expansion, while mean radiance increased from 0.46 to 1.23 nW cm^-2 sr^-1. Approximately 46-53% of illumination expansion occurred within 2 km and about 80% within 5 km of forest boundaries. NTL also showed strong correspondence with electricity consumption and urban population (r ≈ 0.92-0.96). At the annual aggregate scale, NTL showed strong raw correlations with EVI (r ≈ 0.83) and GPP (r = 0.74), but these relationships were strongly influenced by shared temporal trends. After detrending, correlations were - 0.008 for EVI, - 0.019 for NDVI, and 0.029 for GPP. Alternative detrending methods, lagged analyses, spatial-scale sensitivity, and phenological assessments showed no consistent moderate regionalscale response. Effect-size and statistical-power analyses further indicated that the overall dataset had reasonable sensitivity to a predefined moderate association (|r| = 0.30), while forest-type analyses had lower sensitivity. Overall, the results indicate that a moderate, spatially coherent annual-scale NTL-vegetation response is unlikely within the sensitivity of the available observations, while smaller, localized, seasonal, species-specific, or long-term ecological effects cannot be excluded.