Night lights observations significantly improve the explainability of intra-annual vegetation growth globally.

Understanding the underlying mechanism of vegetation growth is of great significance to improve our knowledge of how vegetation growth responds to its surrounding environment, thereby benefiting the prediction of future vegetation growth and guiding environmental management. However, human impacts on vegetation growth, especially its intra-annual variability, still represent a knowledge gap. Night Lights (NL) have been demonstrated as an effective indicator to characterize human activities, but little is known about the potential improvement of intra-annual vegetation growth using seasonal NL observations. To address this gap, we investigated and quantified the explainability improvement of intra-annual vegetation growth by establishing a multiple linear regression model for vegetation growth (indicated by Normalized Difference Vegetation Index, NDVI) with human factor (indicated by NL observations here) and three climatic factors, i.e., temperature, water availability, and solar radiation using the Principal Components Regression (PCR) method. Results indicate that NL observations significantly improve our understanding of intra-annual vegetation growth globally. Model explainability, i.e., adjusted R metric of the PCR model, was comparatively improved by 54 % on average with a median value of 11 % when taking NL observations into consideration. Such improvement occurred in 82 % of the whole investigation pixels. We found that the improvement of model explanatory power was significant in regions where both NL and NDVI trends were large, except for the case where both of their trends were negative. At the country-level, the improvement of model explanatory power increases as GDP decreases, illustrating a greater improvement in a lower middle-income country than that in a high-income country. Our findings emphasize the importance of considering human activities (indicated by NL here) in vegetation growth, offering novel insights into the explanation of intra-annual vegetation growth.