The aim of the project is to develop a novel approach, which will allow us to use remotely sensed spectral information to infer leaf traits on a watershed scale. Relationships between chemical composition of leaf pigments and remotely sensed spectral traits are based on classical data from more than two decades ago. Technologies have progressed, resolution of biogeochemical analytics has increased, remote sensing detectors have improved, but these classical data and relationships have not yet been updated. In the proposed project, we plan to first, measure the spectral data (e.g., hemispherical, conical) of leaf and canopy traits in the field, second describe and identify individual leaf pigments using liquid chromatography coupled to light detectors, and then establish relationships between the two based on their absorption spectra.
More than two decades ago, the classical relationships between plant pigment chemical composition and remotely sensed spectral properties have been established. Thanks to technological advancements since then, precision of biogeochemical tools and remote sensing detectors improved significantly, but the relationships have not yet been updated. Today, chromatography combined with mass spectrometry allows the identification of yet unknown pigments, and other biosynthetic precursor and successor coloring molecules, even in complex mixtures. The aim of the proposed project is to improve correlation and relationships between chemical data and remote sensed traits, integrating analytical advances made in plant biogeochemistry and remote sensing. The expected overall outcome would be that remotely sensed spectral information could be used to infer plant traits on a watershed scale.
Monitor pigment development during the growing season for individual trees
Leaves of easily accessible beech trees in the Irchel Park will be sampled multiple times during the whole growing season over all phenological stages. During the days of sampling, a drone equipped with spectral sensor will determine the field spectral data. In the laboratory, we will measure spectral reflectance and pigment composition of the collected leaves. Further, spectral reflectance will be determined after removal of the wax layer for every third sample in order to determine the specular behavior and influence on the spectral data by the cuticle (eg wax layer). The data will be compiled and a first attempt will be made to compare known spectral correlations of chemical and remote sensing data (both, literature and model based), added by yet not included/assessed pigments. This will allow to derive new relationships, at least for one tree from one tree species.
Verify new relationships with several tree species
Samples will be collected and spectral data and chemical composition will be compared once (Laegern) or monthly (Irchel Park) during the growing season for beech and three additional tree species to determine transferability of obtained data to different growing seasons and different tree species with different pigment and wax compositions. Improvements of the relationships will be made and tree species-specific differences will be identified, expanded to Laegern, Irchel Park and potentially another URPP site. Ideally, we would be able to compare biogeochemical and spectral information also to genetic information from the same, identical leaf material data from Kentaro Shimizu.