Analysis of NDVI Benefits From Biomass Estimates

Researchers have been using spectral sensing in croplands for over two decades.  The last decade has seen a shift from satellite and aerial sensing platforms to near-canopy ground-based platforms.  Close proximity to crop targets allows simultaneous collection of additional information that is not possible at greater distances.  Spatial resolution is also greatly increased.  Even at high spatial resolution, spectral information from row crop canopies can be significantly confounded by crop biomass.  A multi-year production-scale manipulative study conducted in Milan, TN included three cotton seeding rates and four rates of nitrogen side-dress.  Seeding rates were included to investigate potential interference of plant spacing and biomass on the ability to detect nitrogen status with NDVI measurements.  Reflectance values used to calculate NDVI were made with a commercially available active light sensor mounted on a ground-based platform.  Plant height was measured simultaneously with NDVI using ultrasonic sensors.  Plant height was significantly correlated with manual height measurements and was strongly related to NDVI.  The coefficient of variation for NDVI measurements also differed by seeding rate treatments and, in other crops, has been shown to improve in-season yield estimates.  Plant height measurements were included as a covariate variable during ANOVA analysis of NDVI.  Results of these analysis and the benefits of supplementing NDVI with biomass information will be discussed.