Extracting physiological info from a hyperspectral time series of a citrus orchard
The collection and interpretation of a hyperspectral time series for a mature citrus orchard in the Western Cape province of South Africa - using a field spectroradiometer - are presented. As a first analysis of this spectral data time series covering the 350-2500 nm range, vegetation, indices were computed. Spectral data were collected from 30 representative trees on a monthly basis, supplemented with climatology, orchard management practices, leaf and soil nutrient data. The hypothesis was tested that biophysical parameters derived at the canopy level from spectral observations can be related to known observed physiological and phenological changes. Most trends in the vegetation indices derived at the canopy level can be linked to observed changes at the leaf level and in the canopy structure, as elicited by the leaf area index (LAI) and the leaf angle distribution (LAD). The Modified Chlorophyll Absorption Ration Index (MCARI/OSAVI) was the index most closely related to leaf level trends in chlorophyll content. Additionally, we could observe a noticeable influence of sunburn and vegetative flush. Seasonal changes in the photochemical reflectance index (PRI) could be inversely linked to changes in the carotenoid-to-chlorophyll ratio. Canopy structure indices such as the Standardized Leaf Area Index Determining Index (sLAIDI) were found sensitive to changes in LAI, average leaf angle as well as to orchard management (pruning and harvest).