The relationship between solar-induced fluorescence and gross primary productivity under different growth conditions: global analysis using satellite and biogeochemical model data

Solar-induced fluorescence (SIF) has become a milestone for estimating terrestrial gross primary productivity (GPP) at the global scale, and therefore it is important to understand how the relationship between SIF and GPP varies with environmental stresses and canopy structures. Our study investigated the impact of environmental conditions on the relationship between SIF and GPP at the global scale based on Global Ozone Monitoring Instrument 2 (GOME-2) SIF and Max Planck Institute for Biogeochemistry (MPI-BGC) GPP. GOME-2 SIF had strong correlations with gridded GPP data at instantaneous (R ² = 0.618, p < 0.001) and daily timescales (R ² = 0.621, p < 0.001) across the globe in eleven major biomes. Strong linear relationships were also consistently found for nine individual biomes at the daily timescale (R ² = 0.512–0.839, p < 0.001) but were not found for evergreen broadleaf forests (EBF) (R ² = 0.142, p < 0.001) and closed shrublands (R ² = 0.395, p < 0.001). Generally, high coefficients of determination (R ² > 0.8) were mainly distributed in southern Europe, southern and central Africa, northeastern and central China, North America, India and southern Russia, and the slope had a similar spatial distribution as the coefficients of determination. The relationship between SIF and GPP varied with air temperature (T Aᵢᵣ), vapour pressure deficit (VPD) and leaf area index (LAI). The slope of the SIF-GPP relationship had a maximal value when the air temperature ranged between approximately 10°C and 20°C, when the linear relationships between SIF and GPP were also the strongest. Our analysis based on the Soil-Canopy Observation of Photochemistry and Energy fluxes (SCOPE) model also indicated that the SIF-GPP relationship was complex and affected by T Aᵢᵣ, VPD and LAI. Our study can improve our understanding of the functional relationship between SIF and GPP and the information contained in remotely sensed SIF.