From the Landsat mosaics, the classifications that result in land cover and land use maps are carried out for each year. Within the logic proposed by MapBiomas the maps will be updated every time there is an improvement in the classification algorithms. The classification methodology is dynamic and procedural, with the purpose of improving the classification of each typology.
Important: when composing a single mosaic or calculating statistics on maps it should be noted that:
- For area calculation, it is necessary to use an appropriate metric projection.
- Until Collection 4.0, the biome limits were adopted by the original scale project published by the IBGE from 1: 5,000,000 to 1: 1,000,000 from the territorial boundaries maps of Brazil 1:250,000 (IBGE) and phytophysiognomies 1: 1,000,000, also of IBGE. Since Collection 4.1. the new limits of biomes on 1:250.000 scale produced by IBGE (2020) were adopted. The boundary shapefile is available for download along with reference maps.
- The RGB color palette of each class is available in the file, and also in ArcMap and QGIS.
- All data is in GeoTiff format and has LZW compression. To get the reference of the codes with the caption classes go to: https://mapbiomas.org/en/codigos-de-legenda?cama_set_language=en
The most current MapBiomas Collection can always be accessed in five ways:
1. browsing online through the MapBiomas dashboard: https://plataforma.brasil.mapbiomas.org/
2. access the data asset in Google Earth Engine: ID: projects/mapbiomas-workspace/public/collection6/mapbiomas_collection60_integration_v1
3. direct access to QGis through the plugin: (We are also already preparing one for ArcGis).
4. Direct download of Collection 6 with Coverage and Land Use maps for all of Brazil in GeoTiff format (one map for each year)
For other years, just change the final year of the link. Available from 1985 to 2020.
5. download by state, biome, municipality and any other desired geometry using Toolkits prepared in Google Earth Engine (GEE). Access the tutorial video to learn how to use the toolkit. Instructions are also available on Github. The following data are currently available:
Transition Maps Legend
The transition maps have codes calculated by joining the first and second year codes, according to the formula below:
(year1 x 100) + year2
303 = forest that remained forest
315 = forest that has been converted to pasture
1515 = pasture that remained pasture
1524 = pasture that has been converted to urban infrastructure
Mosaic Quality Caption
Obs. All data are downloaded with a .CSV table (comma separated text) with the calculated areas. This script is an example of how to perform the Area Calculation in the GHG.
It allows to download maps and statistics of deforestation and regeneration. The value of each pixel is calculated using the following formula:
- 115 = anthropic in the pasture class
- 203 = primary vegetation in the forest class
- 212 = primary vegetation in the class of countryside formations
- 303 = secondary vegetation in the forest class
- 404 = suppression of primary vegetation of the savanna class
- 503 = recovery for secondary vegetation for forest class
- 604 = suppression of secondary savanna vegetation
- 703 = transition noise (all classes above 700 should be disregarded)
the identification of codes for the legend can be seen sheet "legenda" in the statistics table.
- Fire Scars per year
- Fire Scars per month
- Fire Frequency
Explanation of the legend of meaning of pixel value here.
- MAPBIOMAS WATER - Collection 1 - access through the specific toolkit
- PASTURE QUALITY - Collection 1 - access through the specific toolkit
a. Moderately Degraded Pasture
b. Severely Degraded Pasture
c. Pasture with no Signs of Degradation
- IRRIGATION - Collection 2 - access through the specific toolkit
a. Pivot Irrigation
b. Irrigated rice
c. Other irrigation system