A total of 85,152 stratified random samples were generated to validate the annual land cover and use maps of MapBiomas Perú. Samples were visually inspected using Landsat imagery, assigning land use and land cover class for each year from 1985 to 2021.

DOWNLOAD THE SET OF VALIDATION POINTS

This set of samples is available in the following csv, xlsx and shapefile files. Each sample presents coordinates, biome and information per year on cover class and land use in the 'CLASS_' column (code/ID of the legend used in the maps).

UNDERSTAND HOW THE SAMPLES WERE GENERATED

The samples were stratified at two levels: by superchart (uniting every four IBGE 1:250,000 maps) and by slope classes defined by EMBRAPA. Considering a minimum of 500 points per super card and corrected using the Bonferroni equation. Sized to guarantee a maximum margin of error of 5% and a confidence level of 95% in the smallest unit of geographic coverage, that is, in each supermap. For the national territory, the maximum expected error was 0.5%.

For visual interpretation, the TVI (Temporal Visual Inspection) platform developed by Lapig was used, in which each pixel was initially interpreted by 3 independent analysts. Pixels for which there was no agreement between the assigned classes were inspected again by a supervisor and assigned the pixel class.

To date, the most current validation was carried out for Collection 1, which considers the period 1985 to 2021.

For more information on the content description and sampling method, please visit the ATBD. If you have any questions, suggestions or other information, please contact contato@mapbiomas.org and participate in the MapBiomas Forum.