Can Vegetation Indices Serve as Proxies for Potential Sun-Induced Fluorescence (SIF)? A Fuzzy Simulation Approach on Airborne Imaging Spectroscopy Data

Abstract

In this study, we are testing a proxy for red and far-red Sun-induced fluorescence (SIF) using an integrated fuzzy logic modelling approach, termed as SIFfuzzy and SIFfuzzy-APAR. The SIF emitted from the core of the photosynthesis and observed at the top-of-canopy is regulated by three major controlling factors: (1) light interception and absorption by canopy plant cover; (2) escape fraction of SIF photons (fesc); (3) light use efficiency and non-photochemical quenching (NPQ) processes. In our study, we proposed and validated a fuzzy logic modelling approach that uses different combinations of spectral vegetation indices (SVIs) reflecting such controlling factors to approximate the potential SIF signals at 760 nm and 687 nm. The HyPlant derived and field validated SVIs (i.e., SR, NDVI, EVI, NDVIre, PRI) have been processed through the membership transformation in the first stage, and in the next stage the membership transformed maps have been processed through the Fuzzy Gamma simulation to calculate the SIFfuzzy. To test whether the inclusion of absorbed photosynthetic active radiation (APAR) increases the accuracy of the model, the SIFfuzzy was multiplied by APAR (SIFfuzzy-APAR). The agreement between the modelled SIFfuzzy and actual SIF airborne retrievals expressed by R2 ranged from 0.38 to 0.69 for SIF760 and from 0.85 to 0.92 for SIF687. The inclusion of APAR improved the R2 value between SIFfuzzy-APAR and actual SIF. This study showed, for the first time, that a diverse set of SVIs considered as proxies of different vegetation traits, such as biochemical, structural, and functional, can be successfully combined to work as a first-order proxy of SIF. The previous studies mainly included the far-red SIF whereas, in this study, we have also focused on red SIF along with far-red SIF. The analysis carried out at 1 m spatial resolution permits to better infer SIF behaviour at an ecosystem-relevant scale.