Plant Spectral Signatures as Growth Stress Indicators

Spectral properties of land covers are bearers of information and lie at the root of remote sensing technologies which are widely used in natural resources management, change detection, crop assessment, ecosystems preservation and many other world significant problems. In agriculture remote sensing is applied for monitoring of plant development, evaluating of physiological processes and growth conditions. Especially valuable are the spatio-temporal aspects of the remotely sensed data in detecting crop state differences and stress situations. The implementation of airborne and satellite data requires efficient algorithms for data analysis as well as explicit information about vegetation spectral behaviour under different conditions for higher reliability of the derived information. Our paper focuses on different techniques for handling data from multispectral and multitemporal measurements analysing plant spectral signatures with respect to stress factors. Plant spectral properties have been examined for their ability to serve as sustainable stress indicators. Reflectance features, vegetation indices, "red edge" position, transmittance and fluorescent emission characteristics have been studied. The obtained results indicate that growth conditions cause statistically significant variations of plant spectral signatures that allow not only to discriminate between stressed and healthy vegetation but also to quantitatively describe the stress impact in terms of crop agronomical parameters and yield