What was the purpose of FLEX-EU
The overall objectives of the FLEX-EU activity were as follows:
- Acquire and process high-quality hyperspectral datasets of fluorescence in conjunction with extended correlative data
- Perform initial analyses of data quality and generate first estimates of fluorescence
- Explore if the fluorescence signal scales proportionally when the fraction soil cover changes
- Assess how changes in energy dissipation pathways caused by different factors modulate changes in photosynthetic rates and fluorescence
- Explore if major stress or soil amendment practices do have a detectable effect on fluorescence
- Understand if fluorescence enables the detection of differences in canopy gas exchange of different crop species and at different times of the day
- Provide feedback for existing state-of-the-art fluorescence models. The dataset shall be utilised in the framework of this activity to test and evaluate different modelling approaches that simulate and retrieve top-of-canopy fluorescence.
- Additional objectives added in the course of the activity: Explore the functional relationship between tree height and age and sun-induced fluorescence.
What was the outcome of FLEX-EU
Two experimental sites, the forest area of Bílý Kříž (Czech Republic) and the agricultural, anthopogenical influenced sites around Jülich (Germany), were mapped with the HyPlant sensor for the second and third year, respectively. Those maps are part of a continuous data set to better understand the spatio-temporal variations of fluorescence across ecosystems. The maps of red and far-red fluorescence were calculated with the most recent, improved Fraunhofer Line Discrimination (iFLD) method for the agricultural sites, while the forest maps were calculated using the Singular Vector Decomposition (SVD) method. Additionally, a set of classic vegetation indices were calculated from atmospherically corrected reflectance data.
A 'mixed pixel experiment' was designed to investigate the role of fractional cover on the fluorescence signal using experimental data collected with HyPlant. The results of this experiment indicated that the far-red fluorescence signal scales linearly with vegetation fractional cover. Moreover, we demonstrated that a simple unmixing model, with two classes, is suitable to model fluorescence in mixed pixels. These results may have implications for large-scale mapping in the context of the FLEX mission.
Photosynthetic energy dissipation was altered by applying two physical and two chemical agents on a homogeneous grassland (turf) site:
(i) A highly reflective powder (Kaolin) was used to reduce the absorbed Photosynthetic Active Radiation (APAR) by increasing leaf surface reflectance. This simulates the building of reflective leaf substances that is known to be an ecological acclimation to high light conditions. The expected changes in reflectance signal in the spectral range of the APAR was clearly detectable in canopy reflectance. However, the reduced APAR only led to the expected reduction in the red fluorescence. The far-red fluorescence signal cannot be interpreted due to the signal’s the high dependence on the changed reflectivity of the vegetation surface.
(ii). An anti-transpirant agent (Vapor Gard) was used to produce a waxy layer on the leaves that sealed stomata and as consequence reduced transpiration. The treatment simulates the stomatal closure during drought periods. Chamber gas-exchange measurements revealed a reduction of CO2 uptake and transpiration a day after the Vapor Gard treatment, proving the closure of the stomata. Ground and airborne fluorescence measurements revealed a slight but measurable reduction of the far-red fluorescence signal due to the reduced photosynthetic activity the day after the treatment.
(iii) Two chemical agents were applied to block electron transport at photosystem I (Paraquat) and at photosystem II (Dicuran). Both treatments should result in an immediate and great increase of the fluorescence signal due to the limited electron transport of the photosystems. Unfortunately, the Paraquat treatment proved to be too severe, and hours after application clear signs of destruction of the photosynthetic machinery occurred. No clear dynamics of the fluorescence emission could be derived from this experiment
(iv) The experiment with Dicuran was very successful. A reduction of the electron transport at photosystem II resulted in a great increase of fluorescence emission (more than twofold). This increase was reproducibly detectable from ground an airborne measurements. The increase in fluorescence was coupled to a decrease in photosynthetic CO2 uptake rate. Different doses of Dicuran were applied to investigate quantitatively the relationship between photosynthetic electron transport rate and intensity of the fluorescence emission. For all doses, an increase of fluorescence was detected; however, the increase of the absolute fluorescence signal was not directly (linearly) related to the doses of the application. Fluorescence measurements from the ground could reveal that areas treated with lower doses recovered faster than areas treated with higher doses. This result could not be confirmed from the airborne measurements.
|Geographic Site||Agricultural area around Jülich, Germany
Forest sites in Czech Republic
|Field of Application||Vegetation and Forest Monitoring|
|Dataset Size||5.9 TB|
Data Citation Users, who, in their research, use ESA Earth Observation data that have been assigned a Digital Object Identifier (DOI), are asked to use it when citing the data source in their publications:
Digital Object Identifier: European Space Agency, FLEX-EU 2014 https://doi.org/10.5270/ESA-20835d4