Photosynthetic carbon uptake is a fundamental metabolic process in plants, but has been challenging to measure at large spatial scales. The relationship between canopy-scale Solar-induced chlorophyll fluorescence (SIF) and gross primary productivity (GPP) measurements is complicated by the fact that the sensor only observes a fraction of all the SIF photons emitted by a canopy. The observed SIF radiance is less than the total SIF emitted by all leaves and emitted by all photosynthetic membranes. This corresponds to the two-scale escape ratio (fesc) of SIF photons from (a) the canopy level to (b) the leaf level, and finally to (c) the photosystem level. The questions are:
- For the canopy-level fesc from (a) to (b), how to quantify the reabsorption of SIF photons by other leaves within the canopy?
- For the leaf-level fesc from (b) to (c), how to quantify the reabsorption of SIF photons within the leaf before emitted from that leaf?
- Finally, we need to quantify the intrinsic fluorescence yield at the level of photosystems.
We develop the formulae for fesc at both the canopy and leaf scales using spectral invariant theory by recollision probability and escape probability across multiple scales from within-leaf to canopy.
Results/Conclusions
The two-scale fesc and fluorescence yield can be acquired by independent spectral analysis of scattered sunlight and by combining spectral invariant theory with measurements of the near-infrared reflectance of vegetation (NIRv). The soil and the wax layer lead to the main differences between the visible/near-infrared (VNIR) and SIF measurements at two scales. The canopy-level fesc in the NIR band can be described by NIRv and the fraction of absorbed photosynthetically active radiation (fPAR) with the formula, fesc = NIRv/fPAR. The canopy-level fesc mainly depends on the sun-canopy-sensor geometry, diffuse radiation ratio, canopy structure and leaf scattering albedo. The leaf-level fesc mainly depends on the leaf structure and leaf biochemical constituents, which leads to the asymmetric effect of SIF emissions on the leaf forward and backward side. Our NIRv-based approach explains variations in the canopy-level fesc with an R2 of 0.90 and an RMSE of 1.52% across a series of simulations where canopy structure, soil brightness, and sun-canopy-sensor geometry are varied. Correcting for the escape ratio of SIF in turn improves the linearity of the SIF-APAR relationship. The simplicity and robustness of our approach indicates that NIRv is immediately useful for improving the remote sensing of SIF at large scales.