All MapBiomas Amazonia products are constantly evolving. So far, the effort is composed of four collections:

  • Collection 1 - Maps of the Amazon with a simplified legend for the years 2000 to 2017, based on empirical decision tree classification (published in March 2019).
  • Collection 2 - Maps of the Amazon for the years 1985-2018, using the Random Forest classifier (published in March 2020).
  • Collection 3 - Maps of the Amazon for the years 1985-2020, using the Random Forest classifier and including 8 cross-cutting themes (published in September 2021).
  • Collection 4 - Maps of the Amazon for the years 1985-2021, using the Random Forest classifier and including 9 cross-cutting themes (published in December 2022).

  • Collection 1 - Maps of water surface of the Amazon's countries for the period 2000 to 2022 (published in Setember 2023).


      • The maps in the MapBiomas Amazonia collections are an evolving product. When selecting data make sure you are always using the latest version available.
      • It is recommended to use MapBiomas Amazonia maps up to a scale of 1:100,000. Although it is possible to view them at larger scales (1:50,000), we do not recommend using them at this scale.


      • Land cover and land use maps: Maps in raster format (30x30 m pixel). The legend of these maps, as well as the covered period, evolve from one collection to the next.
      • Mosaics of cloud-free satellite images, representative for each year of the historical series, with a maximum spatial resolution of 30 meters. The mosaics are formed by the composition of cloud-free pixels in each set of images for a certain time period. The periods of the year in which the images are selected vary according to the region of each country. The mosaics contain 141 layers of information formed by the original bands of Landsat images as well as indices, spectral bands and fractions derived from them. The mosaics can be accessed directly in Google Earth Engine for processing or download.
      • Web platform for public consultation with images, maps, and statistics of the MapBiomas Amazon Collection.


        • Collection of image classification and processing scripts.
        • Methodological notes (in ATBD format) explaining the whole working method including the particularities of each country and cross-cutting themes.