A simple leaf-scale radiative transfer model by spectral invariant theory

Abstract:

Radiative transfer modeling is widely used to simulate leaf spectra by quantifying photon transfer processes of reflection, transmission and absorption within a plant leaf. Recent advances in spectral invariant theory offer a unique and efficient approach for modeling the radiative transfer processes. It has been successfully applied for characterizing canopy-scale optical spectra but its potential remains underexplored for applications at the leaf scale. In this study, we developed a leaf-scale optical model based on the spectral invariant properties (SIP) of plant leaves. Similar to the canopy-scale model, the SIP model decouples leaf-scale radiative transfer process into two parts: wavelength-dependent contribution from leaf chemical components and wavelength-independent contribution from leaf structures, with two spectral invariant parameters (i.e., a photon recollision probability p and a scattering asymmetry parameter q). We implemented the SIP model by parameterizing p and q with a measurable leaf trait, the leaf mass per area (LMA). We evaluated the SIP model performance using two in situ datasets (i.e., LOPEX and ANGERS) and another widely used leaf spectral model-PROSPECT. Leaf spectra simulated by the SIP model agreed well with in situ datasets, and with simulations of the PROSPECT model. Benchmarking SIP simulations with the in situ datasets, the root mean squared error (RMSE), bias, and coefficient of determination (R²) are 0.026, 0.035, 0.95 and 0.037, 0.049, 0.91 for leaf reflectance and transmittance, respectively. The results also show that the SIP model can be used to accurately estimate important leaf functional traits such as leaf chlorophyll content, equivalent water thickness and LMA. The SIP model provides an alternative and efficient way of accurately modeling leaf optical spectra and retrieving key leaf functional traits from hyperspectral measurements.