tion in agriculture in the study of the relationships between red and near-infrared (NIR) reflectance and crop Traditional remote sensing methods for yield estimation rely on yield and development. During the development of the broadband vegetation indices, such as the Normalized Difference Vegetation Index, NDVI. Despite demonstrated relationships beLandsat sensors, MSS and TM, spectral channels were tween such traditional indices and yield, NDVI saturates at larger selected to maximize the collection of agricultural and leaf area index (LAI) values, and it is affected by soil background. We other vegetation indicators while minimizing the sensor present results obtained with several new narrow-band hyperspectral payload and data download. Common methods to obindices calculated from the Airborne Visible and Near Infrared tain spatial and temporal crop status based on these sen-(AVNIR) hyperspectral sensor flown over a cotton (Gossypium hirsusors rely on calculating vegetation indices such as the tratum L.) field in California (USA) collected over an entire growing ditional NDVI. The NDVI is built on normalized red and season at 1-m spatial resolution. Within-field variability of yield moni-NIR spectral bands, which are affected by both pigment tor spatial data collected during harvest was correlated with hyperabsorption (red) and the scattering by the medium spectral indices related to crop growth and canopy structure, chloro-(NIR), a function of the arrangement of elements of phyll concentration, and water content. The time-series of indices calculated from the imagery were assessed to understand within-field the canopy (structure). Therefore, NDVI is sensitive to yield variability in cotton at different growth stages. A K means clusvegetation greenness and canopy scattering, causing its tering method was used to perform field segmentation on hyperrelationship with crop growth (Yuzhu, 1990; Benedetti spectral indices in classes of low, medium, and high yield, and confuand Rossini, 1993; Plant et al., 2000). For example, Plant sion matrices were used to calculate the kappa () coefficient and et al. (2000) used false color infrared aerial photography overall accuracy. Structural indices related to LAI [Renormalized to calculate NDVI, studying its relationship to cotton Difference Vegetation Index (RDVI), Modified Triangular Vegetayield. Defoliation, boll opening, and regrowth control tion Index (MTVI), and Optimized Soil-Adjusted Vegetation Index in cotton were evaluated with NDVI calculated from (OSAVI)] obtained the best relationships with yield and field segmencolor infrared digital images (Yang et al., 2003), sugtation at early growth stages. Hyperspectral indices related to crop gesting potential applications of remote sensing data for physiological status [Modified Chlorophyll Absorption Index (MCARI) and Transformed Chlorophyll Absorption Index (TCARI)] were sucotton defoliation strategies. In a different study, growth perior at later growth stages, close to harvest. From confusion matrices conditions and ...
