Spectral Measures from Sentinel-2 Imagery vs Ground-Based Data from Rapidscan© Sensor: Performances on Winter Wheat

Precision agriculture can be supported by different instruments and sensors to monitor crops and adjust agronomic practices. Remote sensing and derived vegetation index are one of the main techniques that allows to derive related-vegetation information. In this work the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Red-Edge index (NDRE) derived by active handheld Rapidscan© (RS) and passive Sentinel-2 (S2) sensors were compared focusing on the wheat crop. To deal with different sensor wavebands centers, different S2 wavebands were considered and two different NDVI and four different NDRE derived by S2 data were computed. The comparison between RS and S2 was performed during three phenological stages of wheat: first node, flowering and milk. In each period, RS-derived indices were modelled to estimate the S2 ones. Results show that the best conversion models found was linear. In addition, a high correlation and R2 (>0.7) coefficient was found, except during flowering stage. Results confirm the opportunity to scale data and related agronomic information from ground sensor to satellite improving decision support system in agriculture.