Platespect: A new model for leaf fluorescence spectra

Models to simulate solar-induced chlorophyll fluorescence (SIF) are widely used to interpret fluorescence observations across scales. However, leaf fluorescence spectra models often mix-use the plate model and Kubelka-Munk (KM) model, which differ in their assumptions in the internal scattering within a leaf and brings in uncertainty in explaining SIF observations. Additionally, fluorescence photons are not conserved in spectral models due to their use of a sigmoid function to adjust the fluorescence emission spectrum dependent on excitation wavelength. To resolve these problems in SIF simulation, we present a new spectral model, Platespect. It is based on the plate model that can compute backward and forward leaf fluorescence spectra and also rescales the raw fluorescence emission spectrum to conserve fluorescence photons. We theoretically compared the fluorescence simulations from Platespect and Fluspect, which adopts the commonly used KM model, at the leaf and canopy scales; we also evaluated them with leaf-level backward fluorescence observations. At the leaf level, although Platespect predicted fluorescence magnitudes similar to those of Fluspect, it showed substantial differences in the backward and forward fluorescence spectra. Accounting for scattering among leaf plates in Fluspect helps reduce the difference. Platespect predicted a higher far-red fluorescence due to the rescaled fluorescence spectra emitted by longer wavelength light. When fitted against leaf-level observations, Platespect performed slightly better in the red fluorescence region, but all models showed a systematically biased fluorescence spectrum. Assessed at the canopy level, Platespect-based simulations predicted a higher SIF and higher sensitivity to leaf chlorophyll content. Our results highlight the necessity of better representing the scattering among plates, improving the raw fluorescence emission spectrum, and conserving emitted fluorescence photons to improve the simulation of SIF across scales.