In recent years, solar‐induced chlorophyll fluorescence (SIF) retrieved from spaceborne spectrometers has been extensively used as a proxy for terrestrial photosynthesis at relatively sparse temporal and spatial scales. The near‐infrared band of the recently launched TROPOspheric Monitoring Instrument (TROPOMI) features the required spectral resolution and signal‐to‐noise ratio to retrieve SIF in a spectral range devoid of atmospheric absorption features. We find that initial TROPOMI spectra meet high expectations for a substantially improved spatiotemporal resolution (up to 7‐km × 3.5‐km pixels with daily revisit), representing a step change in SIF remote sensing capabilities. Here, we will show novel retrievals of red and far-red retrievals from TROPOMI spectra and sensor inter-comparisons with OCO-2 as benchmark. Overall, we find an excellent performance of the TROPOMI sensor, which will enable a step-change in fluorescence research on the global scale

The high-performance imaging spectrometer HyPlant consists of two modules, one being a classical hyperspectral imaging module (380 - 2500 nm), while the second module resolves the red and far-red spectral window with very high spectral resolution enabling the retrieval of solar-induced fluorescence (FWHM 0.24 nm in the window of 670 - 780 nm). HyPlant is operational since 2012 and since then was continuously improved to serve as an airborne demonstrator for the FLEX satellite mission. HyPlant delivered hundreds of flight lines that were recorded over different ecosystems using various flight patterns that range from low altitude flights for spatially high resolution mapping as well as high-altitude flights to better understand the influence of the atmosphere on fluorescence retrieval.
With this presentation we will first review the technical improvements of HyPlant and their links to the technical refinement of the FLEX satellite mission. Special care is put on an accurate pre-processing of the high resolution data and the correction of an imperfect point spread function and stray light. Three retrieval methods to quantify solar-induced fluorescence were implemented and used for the different campaign objectives.
Secondly, we will give an overview on various multi-national campaign activities during which we quantified fluorescence emission and photosynthetic efficiency across scales. Leaf-level measurements were used to characterize the functional dynamics of changing photosynthesis in response to environmental constraints. Novel active and passive fluorescence instruments were developed and applied to quantify integrated canopy photosynthesis top-of-canopy. These ground based data are crucial to correctly retrieve and interpret HyPlant imagery, which provide for the first time large scale maps of fluorescence emission and photosynthesis demonstrating the rich information content of solar-induced fluorescence. Various stress experiments were conducted and HyPlant data were complemented with drone and unmanned aircraft data to better understand the dynamic nature of solar-induced fluorescence, photosynthetic efficiency and other canopy traits.
This presentation will give an overview on the data and the interpretation from the past five years since HyPlant is operational and was used in the frame of several multinational campaigns in preparation of the FLEX satellite mission. We will review and synthesis various publications for a holistic view on 5 years of airborne campaigns in preparation of the FLEX satellite mission.

References
Rascher U. et al. (2015) Sun-induced fluorescence - a new probe of photosynthesis: First maps from the imaging spectrometer HyPlant. Global Change Biology, 21, 4673–4684; doi: 10.1111/gcb.13017.
Rossini M. et al. (2015) Red and far red Sun-induced chlorophyll fluorescence as a measure of plant photosynthesis. Geophysical Research Letters, 42, 1632-1639.
Simmer C. et al. (2015) Monitoring and modeling the terrestrial system from pores to catchments – the Transregional Collaborative Research Center on Patterns in the Soil-Vegetation-Atmosphere System. BAMS – Bulletin of the American Meteorological Society, 96, 1765-1787.
Wieneke S., Ahrends H., Damm A., Pinto F., Stadler A., Rossini M. & Rascher U. (2016) Airborne based spectroscopy of red and far-red sun-induced chlorophyll fluorescence: Implications for improved estimates of gross primary productivity. Remote Sensing of Environment, 184, 654-667, doi: 10.1016/j.rse.2016.07.025
Drusch M. et al. (2017) The FLuorescence EXplorer mission concept - ESA’s Earth Explorer 8. IEEE Transactions on Geoscience and Remote Sensing, 55, 1273-1284, doi: 10.1109/TGRS.2016.2621820.
Middleton E.M.et al. (2017) The 2013 FLEX – US Airborne Campaign at the Parker Tract Loblolly Pine Plantation in North Carolina, USA. Remote Sensing, 9, article no. 612, doi:10.3390/rs9060612.
Colombo R. et al. (2018) Variability of sun-induced chlorophyll fluorescence according to stand age-related processes in a managed Loblolly pine forest. Global Change Biology, 24, 2980-2996, doi: 10.1111/gcb.14097.
Gerhards M., Schlerf M., Rascher U., Udelhoven T., Juszczak R., Alberti G., Miglietta F., & Inoue Y (2018) Remote sensing of water stress symptoms based on airborne optical and thermal images. Remote Sensing, 10, article no. 1139, doi: 10.3390/rs10071139.
Liu X. et al. (2018) Downscaling of solar-induced chlorophyll fluorescence from canopy level to photosystem level using a random forest model. Remote Sensing of Environment, doi: 10.1016/j.rse.2018.05.035.
von Hebel C. et al. (2018) Understanding soil and plant interaction by combining ground-based quantitative electromagnetic induction and airborne hyperspectral data. Geophysical Research Letters, doi: 10.1029/2018GL078658.
Damm A., et al. (2018) Remote sensing of plant-water relations: An overview and future perspectives. Journal of Plant Physiology, 227, 3-19, doi: 10.1016/j.jplph.2018.04.012

Mapping and monitoring the spatial and temporal patterns of terrestrial gross primary productivity (GPP) through the use of satellite remote sensing is of paramount interest for vegetation, ecosystem and climate science. While the rate of terrestrial photosynthesis cannot be directly measured from space, research in the last years has demonstrated that sun-induced chlorophyll fluorescence (SIF) retrieved from satellite spectrometers can be a highly valuable proxy. However, the SIF signal is very small and notoriously difficult to measure, requiring high spectral resolution and signal-to-noise ratio, which generally comes at the expense of revisit frequency and spatial detail. As a result, current SIF datasets have a spatial resolution on the order of 0.5° (approximately 50 km), which is too coarse for many applications in land science. The situation has changed in 2018 with the operational arrival of the TROPOspheric Monitoring Instrument (TROPOMI) on-board of the European Sentinel-5 Precursor (S5P), from which SIF can be retrieved from a ground pixel footprint of around 7 km by 3.5 km at nadir. While this offers promising possibilities for the future, many domains of applications, such as forest biomass accumulation, agricultural production or plant stress monitoring during droughts, would also require an archive of various years of data before adopting the technology.

The global objective of this work is therefore to explore the feasibility of using coarse SIF retrievals to reconstruct a valid high spatial resolution archive for TROPOMI. To do so, we leverage on a proven methodology to downscale SIF retrieved at 0.5° spatial resolution from the GOME-2 instrument. This technique relies on using a light use efficiency (LUE) modelling approach to establishing a semi-empirical relationship between SIF and various explanatory variables also derived from remote sensing, such as vegetation indices (VIs), land surface temperature (LST) and evapotranspiration (ET). These explanatory variables are first aggregated from their fine spatial resolution (e.g. 0.05°) to that of the GOME-2 dataset. Then, the non-linear LUE relationship is calibrated independently for each time-slice based on various spatial samples within an adaptive moving window. Third, once properly calibrated the relationship is applied on the original explanatory variables to result in an estimate of SIF at fine spatial resolution.

While this methodology has been demonstrated in the past using a deprecated GOME2 SIF retrieval and a single set of explanatory variables (NDVI, LST and ET) from the previous collection of products from the MODIS instrument, the present work moves a step further by exploring different SIF products and enhanced explanatory variables from the newest MODIS collection, including NDWI, EVI, NIRv and LST at different times of the day. The optimal configuration is selected based on a confrontation with yet another satellite: NASA’s OCO-2 mission. OCO-2 takes very sparse and discontinuous spatial samples that provide SIF values over a small footprint of ~3 km2. An exhaustive SIF validation dataset was thus compiled based on near nadir OCO-2 observations across the globe during the period of 2015-2017 to ensure the robustness of the downscaled SIF. Finally, the optimized downscaled SIF is confronted with the first global retrievals from TROPOMI in 2018.

Beyond constituting a baseline for the TROPOMI SIF retrievals, this work could also be useful for the future FLuorescence EXplorer (FLEX) satellite mission, such as to produce a prior SIF data set that can be used to optimize FLEX retrievals.

The passive measure of sun induced chlorophyll fluorescence, based on the atmospheric oxygen absorption lines, is affected by many uncertainties such as sensor noise, atmospheric influences or mathematical assumptions. To assess the uncertainty and to calibrate the retrieval method in real measurement conditions an artificial fluorescence reference was envisioned.
This reference is designed to emit a known amount of red and near-infrared light, mimicking the actual spectral shape of the plants chlorophyll fluorescence using two distinct light emitting diodes with peak emissions at 680 nm and 750 nm. Not only the radiance of the diodes, also is their overlapping spectral shape chosen to follow the natural fluorescence. This serves as useful tool for the calibration and validation of fluorescence retrieval of hyperspectral sensors in the field. But by deploying a large grid of light emitting diodes, several square meters of artificial reference surface serve in a larger scale as a calibration and validation target for airborne sensors. We present the electrical design and the construction of two 3 by 4 m active reference panels and discuss their calibration. An exemplary field deployment for the validation of the fluorescence retrieval of the airborne HyPlant sensor, during the ESA Atmoflex 2018 campaign is shown. Preliminary results and future developments are discussed.

Agricultural production undergoes increasing pressure from anthropogenically-induced and natural changes, including rising population, conversion of food (cereals) into biofuels, increased protein demands and climatic extremes. Through a fleet of Earth observation (EO) satellites, Europe is determined to keep its fingers on the pulse of its agricultural lands. Among the most innovative vegetation monitoring concepts is the FLEX (FLuorescence EXplorer), which was selected as ESA’s 8th Earth Explorer (EE8) and is by design timely and ground-breaking: it is the first mission concept specifically dedicated to monitor the ‘breathing’ of terrestrial vegetation. FLEX will fly in tandem with ESA’s Sentinel-3 satellite (S3) and will globally measure the spectral emission of Sun-Induced chlorophyll Fluorescence (SIF) from terrestrial vegetation. Together with S3, these two new-generation European EO missions offer immense possibilities to increase knowledge of the basic functioning of the Earth’s vegetation, notably the photosynthetic process of plants resulting in carbon fixation. To do so, a diversity of essential vegetation products from both missions will be derived. These products will eventually be implemented into a prototype FLEX-S3 assimilation system to infer unbiased estimations of photosynthesis, i.e., uncoupled from confounding canopy and atmospheric effects.
In this work, first results will be presented on developed biophysical variables retrieval algorithms and obtained products using simulated FLEX-FLORIS scenes and real S3-OLCI images. When it comes to developing operational retrieval algorithms, biophysical variables retrieval algorithms should be accurate, fast, and deliver associated uncertainties. In all generality, retrieval methods can be classified into one of the following families: (1) parametric regression, (2) non-parametric regression, (3) radiative transfer model inversion, and (4) hybrid methods. A systematic qualitative and quantitative comparison of multiple methods of these families have been conducted in Verrelst et al., (2015a, 2015b). It was concluded that the non-parametric methods and hybrid methods show most promising features, especially those based on machine learning regression algorithms (MLRAs). Specifically, Gaussian processes regression (GPR) is concluded being among the most competitive MLRA (Camps-Valls, 2016): GPR provides a probabilistic approach for learning generic regression problems with kernels, while maintaining very good numerical performance and stability with sparse training data.
By training GPR models with simulated data coming from a radiative transfer model (RTM), hybrid retrieval models have been developed. The SCOPE RTM is used to generate a training database, which is then used to develop the models for the retrieval of the following biophysical variables: (1) leaf area index (LAI), (2) leaf chlorophyll content (LCC) and (3) fraction of absorbed photosynthetically active radiation (FAPAR). For both FLEX-FLORIS and S3-OLCI data sources, the three vegetation products have been successfully validated against simulated reference data as generated by the FLEX End-to-End simulator. At the same time, absolute and relative uncertainty layers provide information about the confidence of the per-pixel retrieval. By subsequently merging both FLEX and S3 data sources into a synergy product, retrieval accuracies of the vegetation products further improved. For S3-OLCI, generated models were also applied to images of the S3-SYN surface reflectance product acquired over Europe. Uncertainty maps reveal that retrievals are obtained below an uncertainty of 20% for the large majority of pixels. Finally, this work will close with some visions towards assimilating these products with FLEX fluorescence retrievals into a photosynthesis assimilation system.

Camps-Valls, G., Verrelst, J., Munoz-Mari, J., Laparra, V., Mateo-Jimenez, F., & Gomez-Dans, J. (2016). A survey on Gaussian processes for Earth-observation data analysis: A comprehensive investigation. IEEE Geoscience and Remote Sensing Magazine, 4(2), 58-78.

Verrelst, J., Camps-Valls, G., Muñoz-Marí, J., Rivera, J. P., Veroustraete, F., Clevers, J. G., & Moreno, J. (2015a). Optical remote sensing and the retrieval of terrestrial vegetation bio-geophysical properties–A review. ISPRS Journal of Photogrammetry and Remote Sensing, 108, 273-290.

Verrelst, J., Rivera, J. P., Veroustraete, F., Muñoz-Marí, J., Clevers, J. G., Camps-Valls, G., & Moreno, J. (2015b). Experimental Sentinel-2 LAI estimation using parametric, non-parametric and physical retrieval methods–A comparison. ISPRS Journal of Photogrammetry and Remote Sensing, 108, 260-272.

Plants developed various strategies to cope with changing environmental conditions, ranging from physiological to biochemical to morphological adaptation mechanisms. Comprehensive observational capability of plant traits representing physiological processes, leaf biochemistry, and canopy structure is pivotal to assess the status and future trajectories of ecosystem functioning under the ongoing pressure of environmental change. Established remote sensing approaches (e.g. based upon Sentinel-1, -2 and -3 observations) provide reliable information of plant’s biochemistry and structure. Complementary approaches are currently under development (e.g. in frame of European Space Agency’s upcoming Fluorescence Explorer (FLEX) mission) and will provide essential data to assess functional traits representing plant mediated gas exchange processes (e.g. net CO2 assimilation, transpiration).

The development of remote sensing approaches to estimate plant’s gas exchange requires combining remote sensing measurements with established observations of gas exchange (i.e. gas chambers at leaf level or eddy covariance (EC) systems at canopy level). This combination, however, is still an unsolved scientific challenge due to fundamentally differing sampling strategies. EC measurements represent a spatially varying area of a vertically integrated canopy, while remote sensing measurements represent the upper leaf layer of a fixed canopy region. Further, the relation of both observations varies in time and space and depends on how the upper canopy layer is coupled with or decoupled from the vertically integrated canopy.

This study applies concepts of 3D radiative transfer modeling, ecohydrological modelling, remote sensing, fluorescence spectroscopy, and micrometeorology to provide a tomographic perspective on the gas exchange of forest canopies. We analyze simulated 3D fields of transpiration and net CO2 assimilation to quantify the relation of estimates from the upper canopy (remote sensing perspective) with the vertical integral of a canopy (in situ perspective). We assess the impact of canopy structure and environmental conditions on such relationships and identify strategies to compensate for a footprint mismatch. We discuss the severity of a possible “remote sensing bias” for the assessment of complex 3D vegetation canopies. Our findings are a step towards harmonizing the fundamentally different footprints of remote sensing and in situ gas exchange measurements. Results are essential to define optimal validation strategies of functional vegetation traits stemming from current (i.e. Sentinel-2, 3 and -5p) and upcoming satellite (i.e. FLEX) mission products.