Modeling solar-induced fluorescence of forest with heterogeneous distribution of damaged foliage by extending the stochastic radiative transfer theory

Abstract

Solar Induced chlorophyll Fluorescence (SIF) has been used as a novel proxy of photosynthetic activity, which carries information about plant physiological state from the remote sensing observations. It is of great interest and potential to use SIF to detect forest stresses, but the approach requires accurate modeling of the SIF emission within stressed forests. However, the existing radiative transfer approaches of SIF generally ignore the within-crown heterogeneity caused by pests. To account for within-canopy scattering with both high accuracy and efficiency, FluorESRT was proposed as a new model for simulating the SIF of forests with heterogeneous distribution of damaged foliage, which was the synergy of the SIF radiative transfer and the Stochastic Radiative Transfer (SRT) model for forest with vertical distribution of damage. The performance of FluorESRT for both homogeneous and heterogeneous cases were well validated not only by the one-dimensional (1D) model SCOPE but also by the three-dimensional (3D) model DART. As an analytically simple approach, FluorESRT can evaluate the sensitivity of canopy SIF and apparent reflectance to the level of pest damage. The impact of structural properties of damaged foliage on canopy fluorescence, as well as the response of hyperspectral signals on the pest damage was analyzed. According to our study, hyperspectral vegetation indices with the consideration of SIF showed higher sensitivity to pest damage in the early stage when the response of broad band vegetation indices was hardly detectable. The results indicated that the potential of FluorESRT to simulate SIF and hyperspectral apparent reflectance should play a critical role in early monitoring of pest damage.