Skip to main content

Remote Estimating Vegetation Health

Final Report Summary - VEGHEALTH (Remote Estimating Vegetation Health)

Life on Earth is driven by photosynthesis, producing both oxygen and organic matter. Photosynthesis is one of the earliest biological processes to evolve, and the pigment systems in modern photosynthetic bacteria, algae, and higher plants appeared at least 2.5 billion years ago and date from this period. The length and stability of this record, and the various roles of pigments in photosynthesis, protection and defense, demonstrate the functional importance of their composition, associated protein complexes, and chloroplast structure.

Significance of the pigments for life on Earth provides the rationale for improving our capability to remotely monitor them. Developing methods to quantify pigment content and composition from remotely sensed data provides a capability that advances understanding of photosynthetic processes (e.g. light regulation, photooxidation, chlorophyll fluorescence) and insight into detection and monitoring of vegetation condition (e.g. environmental stressors), thus, vegetation health and vigor.

Monitoring plant physiological status via measuring leaf and plant reflectance offer a number of distinct advantages over traditional destructive approaches - simplicity, sensitivity, reliability and a high throughput. Remote sensing techniques allow monitoring of vegetation status at field, regional and global scales.

Overall goal of the project was to develop the scientific foundation for remote monitoring vegetation health - pigment content and composition. It focused on development a prototype "Sentinel-based Vegetation Pigment Product" that accurately estimates chlorophyll, carotenoid and anthocyanin contents, and easy to implement.

For non-destructive estimating foliar pigment content (chlorophylls, carotenoids and anthocyanin), a model that relates reflectance in three spectral bands with foliar pigment contents was validated using measured reflectance spectra of leaves of tree and crop species, and reflectance simulated by radiative transfer model. It was shown that using developed model, simple low cost radiometers were able accurately estimate foliar pigment contents and composition in the field. For the first time generic, unified approach for accurate foliar pigment estimation not requiring re-parameterization for different species was developed and validated.

We took part in development of a new radiative transfer model PROSPECT-D. For the first time the model was able estimating contents of all three foliar pigments – chlorophylls, carotenoids and anthocyanins. In this project, we performed a validation of the model on various experimental datasets. The results showed that the model allowed accurate prediction of leaf pigments. This was particularly the case for leaf carotenoids that are sensitive indicators of plant stress.

For proximal sensing in the field, three band model was applied for estimating crop chlorophyll and carotenoids contents, green leaf area index, biomass, yield and gross primary production. The models for estimating biophysical characteristics were tested over crops (maize, soybean, potato, and wheat) in Israel, Holland, and Japan and grasslands in Italy. The results showed that the models did not require re-parameterization for different crop and grassland species studied. These findings lay a strong foundation for the development of generic, universal algorithms for estimating crop health.

For satellite observations, in situ hyperspectral reflectance and ground truth data, collected across eight years (2001–2012) at three AmeriFlux sites in Nebraska USA over C3 and C4 crops, were used for model calibration. The hyperspectral reflectance was resampled to simulate the spectral bands of sensors aboard operational satellites (Aqua and Terra: MODIS, Landsat: TM/ETM+), a legacy satellite (Envisat: MERIS), as well as Sentinel-2, Sentinel-3, and Venus. The models were validated using MODIS, TM/ETM+, and MERIS satellite data. The models using the red edge bands of MERIS, Sentinel-2, Sentinel-3, and Venus allowed accurate chlorophyll, green LAI and gross primary production estimation over areas containing different crops with no re-parameterization, which is crucial for remote sensing of vegetation biophysical parameters. Using ESA MERIS satellite data gave us an opportunity to test the prototype “Sentinel-based Vegetation Pigment Product” on available satellite data before Sentinel-2 and Sentinel-3 were launched.

Quantitative remote sensing techniques for estimating pigment content in crops and grasslands developed in this project are of great interest to people from diverse scientific backgrounds. Physicists, modelers, meteorologists greatly need these results to include the estimates in the models of climate and weather. Biologists need these results for the photosynthesis and production models. Agronomists need the results to understand the role of pigments in crop production. Decision makers need the results of local, regional and global monitoring of crop and grassland conditions as a basis for making decisions.

Among the main goals of Horizon 2020 sustainable growth project are protecting the environment, climate action, food security, sustainable agriculture and preventing biodiversity loss. The project contributes directly to the enhancement of the quality of monitoring terrestrial vegetation, quantification of the carbon budget, food security and sustainable agriculture by estimating the quality of vegetation using the satellite data. The project helps also to address major concerns shared by all Europeans such as climate change and will tackle societal challenges by helping to bridge the gap between research and the use of its results, by bringing objective information about the vegetation quality directly to the users. Moreover, this project provides a new direction and add value to future research projects in the ERA, specifically the Earth Explorer mission FLEX. Pigment quantification is a critical issue for FLEX mission.