Detection of citrus huanglongbing using Sentinel-2 satellite imagery in Huimanguillo, Tabasco

Abstract

Objective: develop a fast, practical and economical methodology for the determination of plants with HLB symptoms, by exploring the potential use of Sentinel-2 satellite images.

Design/methodology/approach: Sentinel-2 images were used to perform a supervised classification to discriminate healthy Persian lime trees and trees with HLB molecularly verified by qPCR, as well as other land uses in the citrus growing area of Huimanguillo, Tabasco.

Results: The results showed that in the green (560nm), red (665 nm) and near infrared (705 nm) band regions the spectral response of trees with HLB is higher than that of healthy trees. An area of 26.96 and 12 981 hectares of Persian lime affected and free of HLB was found, respectively, with an accuracy of 0.84 % and a Kappa coefficient of 0.83. Field verification corroborated three plantations in which plants with HLB were previously detected and are currently in production.

Limitations on study/implications: The study was carried out in the period of less cloud cover, as this is a limiting factor for the analysis of the image since it can lead to inadequate tracking of the affected area.

Findings/conclusions: The results showed that using Sentinel-2 satellite images, it is possible to differentiate Persian lemon plants affected by HLB from healthy plants.