Skip to content

MapBiomas (Land Cover and Use)

Tabular data from the MapBiomas Project — area (ha) by land cover and use class, biome and state, with an annual historical series since 1985.

mapbiomas.cobertura()

Area by land cover and use class x biome x state x year.

import agrobr

df = await agrobr.mapbiomas.cobertura(bioma="Cerrado", ano=2020, estado="GO")

Parameters

Parameter Type Required Description
bioma str No Biome: "Amazonia", "Cerrado", "Caatinga", "Mata Atlantica", "Pampa", "Pantanal". If None, all
estado str No Filter by state (e.g. "MT", "SP") or state name
ano int No Year (1985-2024). If None, all years
classe_id int No MapBiomas class code (e.g. 15 for Pasture)
nivel str No "estado" (default) or "municipio". Municipal downloads ~660 MB
municipio str No Partial filter by municipality name (case-insensitive). Requires nivel="municipio"
colecao int No Accepts only the current collection (10) or None; other values raise ValueError
as_polars bool No Return as polars.DataFrame
return_meta bool No If True, returns (DataFrame, MetaInfo)

Returned Columns

Column Type Description
bioma str Biome name
estado str State code (e.g. "MT")
municipio str Municipality name (only when nivel="municipio")
classe_id int MapBiomas class code
classe str Class name (e.g. "Pastagem", "Formacao Florestal")
nivel_0 str Category: "Natural", "Antropico", "Natural/Antropico", "Indefinido"
ano int Reference year
area_ha float Area in hectares

MapBiomas Classes (main)

Code Class Level 0
3 Formacao Florestal Natural
4 Formacao Savanica Natural
12 Formacao Campestre Natural
15 Pastagem Antropico
18 Agricultura Antropico
39 Soja Antropico
20 Cana Antropico
40 Arroz Antropico
9 Silvicultura Antropico
21 Mosaico de Usos Antropico
24 Area Urbanizada Antropico
33 Rio, Lago e Oceano Natural

mapbiomas.transicao()

Area of transitions between land use classes by biome x state x period.

import agrobr

df = await agrobr.mapbiomas.transicao(bioma="Cerrado", periodo="2019-2020")

Parameters

Parameter Type Required Description
bioma str No Filter by biome. If None, all
estado str No Filter by state or state name
periodo str No Period (e.g. "2019-2020", "1985-2024")
classe_de_id int No Source class code
classe_para_id int No Target class code
colecao int No Accepts only the current collection (10) or None; other values raise ValueError
as_polars bool No Return as polars.DataFrame
return_meta bool No If True, returns (DataFrame, MetaInfo)

Returned Columns

Column Type Description
bioma str Biome name
estado str State code
classe_de_id int Source class code
classe_de str Source class name
classe_para_id int Target class code
classe_para str Target class name
periodo str Period in "YYYY-YYYY" format
area_ha float Area in hectares

Available Periods

  • Consecutive annual: 1985-1986, 1986-1987, ..., 2023-2024
  • Five-year: 1985-1990, 1990-1995, ..., 2020-2024
  • Ten-year: 1985-2000, 2000-2024, 1990-2000, 2000-2010, 2010-2020
  • Total: 1985-2024

Synchronous Usage

from agrobr import sync

df = sync.mapbiomas.cobertura(bioma="Cerrado", ano=2020)
df_trans = sync.mapbiomas.transicao(bioma="Amazonia", periodo="2019-2020")

Examples

Deforestation in the Cerrado (loss of native vegetation)

import agrobr

# Transition from Formacao Florestal (3) to Pastagem (15) in the Cerrado
df = await agrobr.mapbiomas.transicao(
    bioma="Cerrado",
    classe_de_id=3,
    classe_para_id=15,
    periodo="2019-2020",
)
print(f"Area convertida: {df['area_ha'].sum():,.0f} ha")

Municipal coverage (Belem, PA)

import agrobr

# Downloads ~660 MB on the first call — filter biome/state/municipality to reduce
df = await agrobr.mapbiomas.cobertura(
    nivel="municipio", estado="PA", municipio="Belém", ano=2020
)
print(df[["municipio", "classe", "area_ha"]].head())

Soybean evolution in Brazil

import agrobr

df = await agrobr.mapbiomas.cobertura(classe_id=39)  # Soja
pivot = df.groupby("ano")["area_ha"].sum()
print(pivot)

Data Source

  • Project: MapBiomas — Annual Mapping of Land Cover and Use in Brazil
  • Collection: 10 (August 2025)
  • Historical series: 1985-2024
  • Resolution: 30m (Landsat)
  • Provider: Multi-institutional collaborative network
  • Data: brasil.mapbiomas.org/estatisticas
  • License: Public data — free to use with attribution to the MapBiomas Project