HYPERPEACH - Modeling biochemical processes in orchards at leaf- and canopy-level using hyperspectral data

You are here

Context and objectives

This research aims to precisely monitor and model economically important plant production processes in peach (Prunus persica L.) orchards by means of hyperspectral reflectance data. The development of an accurate method for processing large quantities of hyperspectral data (data mining) has been research in a previous project (Hypercrunch), while this proposal serves as extension of the Hypercrunch research. The general objectives are (i) to validate leaf- and canopy-level spectra (field spectrometer), scaled to airborne-level using radiative transfer methods against AHS 160 high-spatial hyperspectral data, and (ii) to estimate leaf- and canopy biochemical variables in peach (Prunus persica L.) orchards using these hyperspectral data from various levels and sources (ground- and airborne radiometers). Other specific are: (1) To study radiative transfer methods at the leaf and canopy levels for quantitative estimates of leaf biochemistry (Ca+b, Cm, and Cw) from hyperspectral imagery collected over peach (Prunus persica L.) orchards, (2) to study the potential for N estimation using hyperspectral data from model-retrieved Ca+b and Cm in peach (Prunus persica L.) orchard canopies, (3) definition of digitally measurable in situ parameters necessary to model growth processes, (4) optimization of newly developed processing methods for the large quantity of hyperspectral measurement data using multivariate techniques, and (5) fundamental study of the interaction between solar energy and living plant material.

Project outcome

Expected scientific results

(Expected) Results and deliverables will focus specifically on the derivation of a usable spectral data set in terms of potential corrections (e.g., atmospheric, radiative modeling), model development to describe plant biochemical processes, based on spectral and in-situ inputs, and fine-tuning of radiative transfer models, data reduction, and band selection techniques. These latter aspects are geared towards addressing continuing challenges stemming from previous projects. Specifically, expected results and deliverables can be summarized by: (1) A well-defined hyperspectral multi-layer data set will be derived for the study site in Spain. This will include hyperspectral measurements at various levels (leaf, canopy, airborne), as well as derived and developed indices, and differenced data. These data sets will be based on radiative transfer modeling of spectrometer data, as well as AHS 160 imagery; (2) Ground-based measurements will characterize the on-site peach (Prunus persica L.) orchard in terms of chlorophyll, dry matter, water content, and LAI; (3) Radiative transfer models will have been validated using the acquired AHS 160 imagery, thereby providing methodology and confirmation for the use of radiometers in subsequent research that involves data scaling; (4) Quantitative models with biochemical/biophysical dependent variables will have been developed, based on not only spectral data, but also on in-situ knowledge, as independent variables; (5) Data reduction techniques (Hypercrunch project) will have been validated and further adapted. The defined objectives will have been addressed, thereby providing efficient methods to measure, and eventually steer, plant production processes using data from any number of available data scales/levels.