UAV-Based Hyperspectral Sensing for Yield Prediction in Winter Barley

In this study we evaluated the potential of the hyperspectral sensor “Cubert UHD 185 Firefly” for yield prediction in 76 plots of a field trial with different varieties of winter barley at Königslutter (Lower-Saxony, Germany) in 2017. An UAV was used as carrier platform for the sensor. In 2017 we used 63 channels in a wavelength range of 450 to 700 nm. Predicted yield using PLSR and reference yield closely agreed with R
²
=0.78. We also calculated the NDVIRGB and evaluated its suitability for yield prediction in the same field trial. NDVIRGB and reference yield were less well related to each other with R
²
=0.46. The results show that using additional information from hyperspectral datasets allowed for a better yield prediction compared to RGB data alone. In 2018 a field trial with 76 plots of winter barley at Poppenburg (Lower-Saxony, Germany) was assessed on June 6, using the complete wavelength range of the sensor from 450 to 950 nm. In 2018, predicted yield using PLSR and reference yield agreed with R
²
= 0,81.