Angular normalization of GOME‐2 Sun‐induced chlorophyll fluorescence observation as a better proxy of vegetation productivity

Abstract Sun‐induced chlorophyll fluorescence (SIF) has been regarded as a promising proxy for gross primary productivity (GPP) over land. Considerable uncertainties in GPP estimation using remotely sensed SIF exist due to variations in the Sun‐satellite view observation geometry that could induce unwanted variations in SIF observation. In this study, we normalize the far‐red Global Ozone Monitoring Experiment‐2 SIF observations on sunny days to hot spot direction (SIF h ) to represent sunlit leaves and compute a weighted sum of SIF (SIF t ) from sunlit and shaded leaves to represent the canopy. We found that SIF h is better correlated with sunlit GPP simulated by a process‐based ecosystem model and SIF t is better correlated with the simulated total GPP than the original SIF observations. The coefficient of determination ( R 2 ) are increased by 0.04 ± 0.03, and 0.07 ± 0.04 on a global average using SIF h and SIF t , respectively. The most significant increases of the R 2 (0.09 ± 0.04 for SIF t and 0.05 ± 0.03 for SIF h ) appear in deciduous broadleaf forests. Key Points The sunlit leaf gross primary productivity (GPP) is better correlated to SIF in the hot spot direction The GPP in the peak growth season tends to be underestimated if SIF is not angularly corrected The dense a forest, the significant the angular correction is for SIF as a proxy of the total GPP