Comparison Between the Vegetation Indices Obtained from Sentinel-2 and Planet: A Case Study over a Rice Farm in Northern Italy

Precision Agriculture has shown how to improve the production chain by reducing fertilizers and consequently reducing pollution and costs, sometimes with improvements in yield. This research is carried out in Italy, in collaboration with the agronomic consultancy agency “Dr.Agr. P.A. Barbieri and Dr.Agr.G.L. Rognoni” (Casteggio - Pavia, Lombardia) over “Riserva San Massimo”, a farm near “Gropello Cairoli”, made up of rice crops and a forest (about 15 km2 in total). Agronomists have been using Sentinel-2 images in rice crops for many years and have found the need to use images with higher spatial and temporal resolution: this is the research question. Sentinel-2 has a temporal resolution of 5 days and a spatial resolution of 10 m (at best): some results have been found using Planet images. Sentinel-2, free, is considered a reference in this field; Planet, for a fee, has a higher spatial resolution (3 m) and a longer revisit time (almost daily). The research was conducted through 3 main phases: availability and usability of the images, self-consistency of the Planet images acquired quite simultaneously and coherence between the Sentinel-2 images and those of Planet, captured on the same day. The NDVI and NDRE indices have been studied for their importance in fertilization. It has been shown that Sentinel-2 is not sufficient for all required applications and that Planet could represent an opportunity under some conditions. The planet-derived indices have good self-consistency and the comparison between the Sentinel-2 and Planet indices showed seasonal miscalibration: lower for NDVI, higher for NDRE.