Performance of different Sentinel-2A vegetation indices to estimate soybean yield in precision farming

Precision agriculture (PA) aims to identify crop  and soil variability to improve management and optimize the use of inputs. Yield maps become a  relevant tool for PA planning. Vegetation indices (VI) from remote sensing allow monitoring the spatio-temporal variation of crops in the growing  season....

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Ovando, Gustavo, de la Casa, Antonio, Díaz, Guillermo, Díaz, Pablo, Bressanini, Luciano, Miranda, Cristian
Formato: Artículo revista
Lenguaje:Español
Publicado: Facultad de Ciencias Agropecuarias 2021
Materias:
Acceso en línea:https://revistas.unc.edu.ar/index.php/agris/article/view/25148
Aporte de:
Descripción
Sumario:Precision agriculture (PA) aims to identify crop  and soil variability to improve management and optimize the use of inputs. Yield maps become a  relevant tool for PA planning. Vegetation indices (VI) from remote sensing allow monitoring the spatio-temporal variation of crops in the growing  season. The objective of this work was to evaluate the performance of different Sentinel-2A VIs to estimate soybean (Glycine max (L.) Merril) yield within the framework of PA. The study was carried out using a soybean yield map during the  2017/2018 season from a plot located to the south of Córdoba city, Argentina. Every two bands were taken from a Sentinel-2A image from 4/Feb/2018  and VI were calculated using differences, ratios  and normalized differences. The absolute value of  the Pearson linear correlation coefficient (|r|) between the different IVs and soybean yield was computed. The highest value of |r| (0.726) corresponded to the difference between bands 8  (NIR) and 12 (SWIR), allowing to reproduce with sufficient precision the spatial variability of yields in the plot.