Enhancing phenology modeling through the integration of artificial light at night effects

Abstract

Spring vegetation phenology is closely influenced by photoperiod, and the presence of artificial light at Night (ALAN) therefore substantially impacts the phenological response of plants to climate change. How ALAN impacts spring phenology in relative to warming and what are the drivers regulate these impacts are not well understood. Here we focused on the extra-tropical terrestrial ecosystem (>30°N) of China where the highest urbanization has experienced using satellite images to extract the start of the growing season (SOS) from three independent datasets, as well as ALAN data from harmonized global nighttime light (NTL over 2001–2018. We found that ALAN caused earlier SOS both at the ecosystem level and for the major climate zones, and this advanced effect weakened at lower latitude regions and for the high-altitude ecosystems. Further, we discovered that the advanced effect of ALAN on SOS was strengthened in areas with lower chilling days and with the increased distance from the city center. We therefore derived a new model for the estimation of SOS including the effects of ALAN and the new model provided improved representation of SOS in terms of higher proportions of significant pixels between model estimates and observations, higher correlation coefficients, lower root mean square error, Akaike information criterion and higher Kling-Gupta efficiency. Our results highlight that the effects of ALAN on SOS were influenced by latitude, elevation, and winter chilling. Overall, our study sheds light on the impact of human activities on plant spring phenology and provides insights for predicting plant growth patterns under future urbanization and global climate change.