Mapping avocado in Michoacán with Sentinel-2 images and a mixed methodology

Avocado is a fashionable superfood, and a very profitable crop. Mexico is the first world producer of avocado and Michoacán the first national producer. The objective was to map the avocado crops in Michoacán with Sentinel-2 images, as well as establishing the main coverages. The study area was divided into three zones (Tancítaro, Pátzcuaro and Zitácuaro) and for each one five recent Sentinel-2 images (2019-2020) were downloaded from Copernicus web site. A preprocessing was carried out to concentrate the information of each area in a 12-band image. Avocado and forest plots were identified in . An unsupervised k-means classification algorithm of 50 classes was applied and classes corresponding to avocado, forest, secondary vegetation, tropical dry deciduous forest, agriculture, soil, urban area or water covers were identified. A success rate of 89.58 % was obtained in the avocado and 92.12 % in the forest. In total, 149 729 ha of avocado were identified. The map obtained was compared with Series VI of INEGI, in which the orchards are classified as rainfed agriculture; the forests coincide in more than 70%, while the soil gathers very seasonal covers, and there is confusion in the discrimination of other types of vegetation.