Real surface vegetation functioning and early stress detection using visible-NIR-thermal sensor synergies: from UAS to future satellite applications

Under the current climate change conditions, the early stress detection of crops and worldwide vegetation are crucial to promote sustainable agriculture and ecosystem management. With the upcoming European Space Agency’s Fluorescence Explorer-Sentinel 3 (FLEX-S3) tandem mission, vegetation fluorescence and the auxiliary parameters/traits needed to interpret solar-induced vegetation fluorescence (SIF) will become available at 300x300 m spatial resolution. Today, a variety of SIF-specialized UAS systems exist to retrieve the canopy-emitted SIF over larger areas, e.g., as a reference for airborne imaging SIF sensors. However, they lack the complementary sensors needed for a correct interpretation of the highly dynamic fluorescence emission.  In this study we present the FluoCat system, a unique UAS system which can be mounted either in a UAV or cable-suspended mobile platform. On board the FluoCat are mounted: a high-spectral resolution Piccolo Doppio dual spectrometer system, a MAIA-S2 multispectral camera and a TeAx Thermal Capture Fusion camera, which can be triggered simultaneously according to a pre-set protocol. The FluoCat system mimics the FLEX-S3 sensor configuration, by using a multi-sensor system integrating the visible, NIR and thermal spectral regions providing complete datasets to assess the actual vegetation stress. In this context a field campaign was conducted in the experimental site ‘Las Tiesas’ in Barrax, Spain, with the aim to (1) apply sampling protocols to obtain spatially representative canopy reflectance and SIF measurements, and (2) provide accurate ground truth measurements for real (i.e., leaf) surface reflectance and effective surface fluorescence measurements, linkable to the real photosynthetic performance. Further we demonstrate the development of a sensor synergy product, combining canopy physiological and structural information to reveal real surface physiological stress-related energy emission. The ‘sunlit green fluorescence’ is a synergy product combining the top-of-canopy fluorescence and the fractional vegetation cover of the sunlit vegetation. This synergy product improved the estimation of the effective surface fluorescence flux, using the leaf fluorescence emission as reference, by reducing the errors from 36 % to 18 % (band 687 nm); and from 24 % to 6 % (band 760 nm). Real surface properties and products referring to the actual photosynthetic surface behavior are promising quantitative proxies to assess the impact of climate change and/or management practices on crop lands or even whole ecosystems. With this study we show how innovative proximal sensing platforms can help to develop new data processing schemes combining all required information for the quantitative assessment of vegetation health, even before visible damage occurs. The further processing and normalization of first-derived stress proxies such as SIF can generate further in-depth early stress detection, directly related to the photosynthetic light reactions, and further global carbon assessment. These developments are in direct support for the global monitoring of early vegetation stress under a changing global climate.