Simulating spatially distributed solar-induced chlorophyll fluorescence using a BEPS-SCOPE coupling framework

Remotely sensed solar-induced chlorophyll fluorescence (SIF) has been increasingly used to probe photosynthesis and model the gross primary productivity (GPP). Although SIF at the top of canopy (TOC) can be simulated using the coupled photosynthesis-fluorescence model SCOPE (Soil Canopy Observation, Photochemistry and Energy fluxes), simulating spatially distributed TOC SIF usually requires extensive calculations, entailing some challenges when applying the model to the regional and the global scales. This study puts forward a coupling framework that combines SIF and global terrestrial biosphere models (TBMs). The theory for fluorescence emissions and the fluorescence radiative transfer algorithm described in the SCOPE model were integrated with the two-leaf -based BEPS (Boreal Ecosystem Productivity Simulator) model. To simplify the fluorescence radiative transfer physics, we put forward a canopy-averaged leaf-level fluorescence to represent the fluorescence emitted from sunlit and shaded leaf groups and performed a sensitivity analysis to assess the determining factors in upscaling fluorescence from leaf scale to canopy scale. We found that the relationship between the leaf and canopy fluorescence at 740 nm was mainly affected by LAI. Although brown pigments and leaf inclination angle demonstrate some impacts on the scaling process, an LAI-based coefficient can well characterize the upscaling from leaf to canopy scale. Since our BEPS-SCOPE coupling model deploys the sunlit-shaded leaf separation strategy, we expect that it can efficiently characterize the nonlinear responses of photosynthesis and the associated fluorescence to environmental factors. The performance of our model was evaluated at both site and global scales, which demonstrated a good performance for most plant functional types (PFTs) except for needleleaf types that have a more clumped nature. Apart from these limitations, the presented model can contribute to efficiently simulating SIF at regional and global scales, and has the potential to reduce uncertainties in GPP estimation.