Accurate prediction of grain yield is essential for enhancing food security, particularly in the context of climate change. Although remote sensing indices have been extensively utilized to monitor vegetation growth and estimate crop yields, there has been limited research comparing their effectiveness for predicting grain yield, especially across different growth stages. This study examined the performance of multi-source indices, such as normalized difference vegetation index (NDVI), near-infrared reflectance of vegetation (NIR V ), and solar-induced chlorophyll fluorescence (SIF), in predicting grain yield at various growth stages at Shangshan Rice Research Station in Zhejiang Province, China. The results indicated that SIF exhibited the strongest and most consistent correlation with grain yield ( R 2 = 0.34 to 0.75), followed by NIR V ( R 2 = 0.34 to 0.71). SIF also demonstrated advantages in capturing the dynamic changes of GPP during the reproductive period. During both the vegetative and reproductive stages, leaf area index (LAI) showed significant correlations with NDVI, NIR V , and SIF, whereas leaf chlorophyll concentration exhibited comparatively weaker associations with these indicators. These findings provide valuable insights for improving crop yield forecasts using remote sensing, thereby contributing to enhanced agricultural management and food security strategies under climate change.
Gu et al. (2026) studied this question.