Development of Robust Hyperspectral Indices for Detection of Deviations of Normal Plant State

This research was conducted to assess the potential of hyperspectral indices to detect iron deficiency in capital-intensive multi-annual crop systems. A well-defined hyperspectral multi-layer dataset was constructed for a peach orchard in Zaragoza, Spain, consisting of hyperspectral measurements at various monitoring levels (leaf, crown, airborne). Trees were subjected to four different treatments of iron application (0 g / tree, 60 g / tree, 90 g / tree, and 120 g / tree). Groundbased measurements were used to characterise the on-site peach (Prunus persica L.) orchard in terms of chlorophyll, dry matter, water content, and leaf area index (LAI). Indices were extracted from the spectral profiles by means of band reduction techniques based on logistic regression and narrow-waveband ratioing involving all possible two-band combinations. Physiological interpretations extracted from leaf-level experiments were extrapolated to crownand airborne level. It was concluded from leaf level measurements that a decrease in leaf chlorophyll concentration resulted due to iron deficiency. The results suggested that spectral bands and narrow waveband ratio vegetation indices, selected via multivariate logistic regression classification, were able to distinguish iron untreated and iron treated trees (C-values>0.8). Moreover, the most appropriate indices obtained in this manner fulfilled the expectations by being highly correlated (R>0.6) to the measured chlorophyll concentrations. The visible part of the spectrum, mostly dominated by the amount of pigments (e.g. chlorophyll, carotenoids), provided the most discriminative spectral region (505 740 nm) in this study. The discriminatory performance of a combined chlorophyll and soil-adjusted vegetation index was compared to the results of the selected vegetation indices to estimate the effects of soil background and LAI on those indices.