Dataset for: Collection of image-based points from bands of Sentinel 1, Sentinel 2, SRTM and derived indexes for the classification of vicuña grasslands in 7 communities from Ayacucho region

Read full text
Authors: Javier Ochoa, Henry Juárez, Stef de Haan
Cited by: 0

Abstract

Point extraction based on Sentinel satellite imagery was used to collect georeferenced data for land use classification into nine categories. These categories included grasslands (pajonal), puna grasslands, shrublands, wetlands, forest plantations, water bodies, snow, cropland, or no vegetation (e.g., bare soil, rocks, and infrastructure) for seven communities in the Ayacucho region: Uruiza, San Cristóbal, Lucanas, Huancasancos, Cabana, Aucara, and Andamarca. As a first step, the presence of predominant and homogeneous land use categories was identified through field studies in the communities, prioritizing high-Andean soils and vegetation cover in vicuña management areas by creating a composite of Sentinel images for the dry season. In a second step, points were delineated within homogeneous land use areas using Google Earth Engine. Geographic coordinates were recorded and extracted the Sentinel-2 spectral variables, S1 radar variables, SRTM data, and indices derived from composite image for each point. This process captured the main features variables of 2245 image-based points in 22 vicuña areas across 7 communities in the Ayacucho region.

Fields of Study & Subjects

Remote Sensing in Agriculture Remote Sensing and Land Use Remote-Sensing Image Classification

Publication details

Journal / source
International Potato Center
Publication date
1 January 2026
Document type
Dataset