Photochemical reflectance index (PRI) can be used to improve the relationship between gross primary productivity (GPP) and sun-induced chlorophyll fluorescence (SIF)

Despite much progress has been made in recent years, the estimation of terrestrial gross primary productivity (GPP) remains highly uncertain. Recent studies have demonstrated the usefulness of sun-induced chlorophyll fluorescence (SIF) from plants for estimating GPP. In this study, the relationships of GPP to SIF and Photochemical Reflectance Index (PRI) under different environmental conditions are established using carbon flux data at 61 sites distributed globally and coincident SIF data from OCO-2 and PRI data from MODIS. We analyzed OCO-2 SIF data at both 757 nm and 771 nm against instantaneous and daily GPP values derived from the tower flux data. Our results for these four cases are similar, and those with SIF at 757 nm and daily GPP show that: (1) a significant relationship is found between GPP and observed SIF (SIFobs) for all vegetation types (R2 = 0.53, RMSE = 4.28 μmol m−2 s−1 p < .001); (2) NIRv, a product of the normalized difference vegetation index (NDVI) and near-infrared reflectance, can be used as an approximation of the canopy-escaping probability of emitted SIF to convert observed SIF at one angle to the hemispherical SIF emission that is better correlated to GPP. The relationship between GPP and SIFobs/NIRv was considerably improved from that between GPP and SIFobs (R2 = 0.67, RMSE = 3.39 μmol m−2 s−1, p < .001); (3) The remaining scatter in the relationship between GPP and SIFobs/NIRv is significantly negatively correlated to PRI (R2 = 0.28–0.34, p < .001), and PRI can be used to improve the relationship between GPP and SIFobs/NIRv, (R2 = 0.73, RMSE = 3.04 μmol m−2 s−1, p < .001); and (4) After the use of PRI in the regression, the variability of the slope of GPP against SIF among different plant functional types is greatly reduced from 33% to 19%. As PRI is an indicator of the non-photochemical quenching (NPQ), a pathway that is in parallel of photochemical quenching and chlorophyll fluorescence, these improvements suggest that the information of NPQ from PRI can be used in addition to SIF for better GPP estimation.