Mapping of the normalized difference vegetation index in cotton field

Abstract

The remote sensing data obtained at field level can provide detailed information about the variability of biophysical parameters related to yield over large areas, and present potential for monitoring these parameters throughout the crop development cycle. This study aimed to map the spatial variability of the Normalized Difference Vegetation Index (NDVI) and its components in two commercial cotton (Gossipium hirsutum L.) fields, by using an active optical sensor at the ground level. Data were collected with the aid of an optical sensor installed in a self-propelled agricultural sprayer. A GPS receiver was connected to the sensor, in order to obtain the coordinates of the sampling points. The readings were performed in rows spaced 21.0 m apart by the sensor installed on a vehicle, during the spraying operation, and data analyzed by using the classical statistics and geostatistics. Spatial distribution maps of the variables were generated by kriging interpolation. It was observed a higher spatial variability of NDVI and spectral reflectance of vegetation in the region of near infrared (NIR) (880 nm) and visible infrared (590 nm) in the crop with higher physiological stress, due to the brown bug [Scaptocoris castanea (Hem.: Cydnidae)] attack, when compared to the healthy one.

Keywords

• Gossypium hirsutum L.
• geoestatistics
• remote sensing
• spatial variability
• active optical sensor