Phenology Dynamics of Dryland Ecosystems Along the North Australian Tropical Transect Revealed by Satellite Solar-Induced Chlorophyll Fluorescence

Accurate phenological characterization of dryland ecosystems has remained a challenge due to the complex composition of plant functional types, each having distinct phenological dynamics, sensitivity to climate, and disturbance. Solar-Induced chlorophyll Fluorescence (SIF), a proxy for photosynthesis, offers potential to alleviate such challenge. We here explore this potential using dryland systems along the North Australian Tropical Transect with SIF derived from Orbiting Carbon Observatory-2. SIF identified the seasonal onset and senescence of Gross Primary Production at eddy covariance sites with improved accuracy over Enhanced Vegetation Index and Near-Infrared Reflectance of terrestrial Vegetation from Moderate Resolution Imaging Spectroradiometer, especially at inland xeric shrublands. At regional scale, SIF depicted both earlier onset and senescence across North Australian Tropical Transect. We hypothesized that SIF outperformed Enhanced Vegetation Index and Near-Infrared Reflectance of terrestrial Vegetation mainly because, unlike reflectance, it is not contaminated by background soil, and its total signal is contributed by mixed plant species in additive way. Plain Language Summary Australian dryland ecosystems are critical in regulating the global land carbon sink dynamics. However, it is challenging to accurately characterize their phenology from spaceborne measurements. On the one hand, tropical savannas and semiarid ecosystems (e.g., grasslands and shrublands) are typically composed of a complex mixture of species (woody trees and C-4 grasses) with each having distinct morphologies and physiological responses to climate condition; on the other hand, such ecosystems are highly sensitive to irregular rainfall events and are often subject to disturbances such as fires and storms. In this study, we utilized the North Australian Tropical Transect rainfall gradient as a natural laboratory to assess the ability of satellite solar-induced chlorophyll fluorescence to capture the phenological dynamics of dryland vegetation, in comparison with traditional reflectance-based vegetation indices, that is, Enhanced Vegetation Index and Near-Infrared Reflectance of terrestrial Vegetation. Results showed that satellite solar-induced chlorophyll fluorescence outperformed Enhanced Vegetation Index and Near-Infrared Reflectance of terrestrial Vegetation for characterizing seasonal onset and senescence along North Australian Tropical Transect and therefore had potential for improving large-scale mapping of phenology dynamics of dryland ecosystems over traditional remote sensing of reflectance-based vegetation indices.