2.Legend

The map legend of the land cover and land use initiative corresponds to all the names and descriptions of the classes to be represented in the maps. Each country or region defines its legend according to specific needs and characteristics of its territory. The legend definition is a bottom-up approach, where each team decides the legend that makes sense for their territory.

This chapter describes the best practices concerning the main characteristics of the legends of land use and land cover mappings, how the legend evolves over the collections, and the documents that describe different aspects of the legend in the products and methods shared in the websites and platforms. It is highly recommended to review the supplementary material, with detailed step-by-step instructions and examples, whenever starting the construction of a new legend and/or revising an existing one.

The legend file for all initiatives in 2025 can be found in this link.

This section provides guidelines on the structure and harmonization of the legend considering other territories, the definition of classes and assignment of colors and codes, and the legend compatibility with other global or national classifications systems.

• The legend must be hierarchical and  harmonized with other countries, biomes or regions (hereafter referred to as territories);  

• The higher hierarchical levels must correspond to the grouping of the lower level classes (for example, if you only detail "soybeans" at level 3 of "agriculture" (level 2), it is necessary to create a level 3 class of "other agriculture" to group them as "agriculture" level 2);

• It is a requirement that all the countries and regions keep the same classes in level 1. From level 2, the class hierarchy could differ, but keeping the coherence (ex. Palm Oil could be a class level 4 integrated to Annual Crop or class level 3 integrated to Agriculture);

• Each class definition by country or region should harmonize with the other neighboring country or region  without losing each class's local meaning. The process of harmonization allows the discussion of the border areas between territories and identifies discrepancies that need to be solved. It can also  be applied by the users when using more than one territory through the unique ID class code (even if the class name between territories differs);

• It is a requirement to make sure that each class description is compatible between territories.  

• Each class mapped has a numeric unique code to identify it on the maps and all documents linked to it. The assigned code is valid for that class in all territories of the Network even if they have different names;

• If the class is not mapped anymore, the unique code is not assigned to a new class later on;

• Before assigning a new class, the teams are invited to describe it and check if it is not included in an already existing class. A new class should be communicated to the MapBiomas Network Secretariat, and a code for a new class will be assigned. 

• Each class mapped has a unique color in an hexadecimal code to identify it on the maps and all documents linked to it;

• The assigned color is valid for that class in all territories of the Network even if they have different names and it is also a requirement of the legend;

• The color for each class is  assigned by the MapBiomas’ Network Secretariat.

• Each class mapped should be compatible with global and national systems definitions such as the global ones (IPCC, FAO and IUCN Ecosystem Typology), or national definitions such as IBGE and National Inventory of Greenhouse Gas Emissions in Brazil. One class in MapBiomas could include more than one class of the other systems and vice-versa.

• The spectral and spatial separability and temporal dynamics of different classes are initial aspects to consider when designing the legend  for a specific territory and a specific period to produce a collection of maps;

• The legend definition is a bottom-up approach. Each territory describes the classes to be mapped based on local information, scientific-technical and regulatory aspects and references, and/or other legend definition system comparisons;

• The visual information suggested  to better typify  each class could be Landsat images RGB:654 and/or RGB:564 combination, Google Earth and or Planet images and any other complementary information such as temporal dynamics information, photos of the landscapes and/or schematic representation or criteria of the class definition;

• For some classes, it is also useful to have information related to space, spectrum and time that could help to characterize the class (e.g. flooded or dry seasons, digital elevation models, etc.);

This section provides guidelines on the creation of new classes. 

• The number of mapped classes may increase over the collections’ development (Figure 1);

Figure 1:  Example of progressive legend characteristic. The sequence of classes mapped in Brazil from Collection 1 (7 classes) to Collection 7 (26 classes grouped in 4 levels).

• Whenever a new class is created, it must be mapped across the entire historical series and it should be referred to as Beta until it shows enough consistency and covers the entire territory.

• It is important to consider technical limitations to distinguish classes and revise and re-dialogue before the final version of the legend is included in a new collection;

• This is a cycling process that will be updated before each new collection;

• This process could also include consultation with external stakeholders.

• Each country must discuss the inclusion of new classes that are relevant to the local context ;

• When proposing the inclusion of a new class, it is necessary to evaluate the Legend document consolidated in all countries to check if it is possible to use a class previously added by another country; 

• To propose the inclusion of a new class, it is necessary to present a justification at meetings with the network secretariat showing that it is different from all previously created classes; 

• The creation of a new ID is centralized with a single person in the network to ensure consistency across all countries and regions. 

This section describes the documents that should be made available to the users with different information concerning the legend. 

• Legend Code is a document (pdf; see an example here) that includes the class level, class name, class ID number and the hexadecimal color code.  It is available in the MapBiomas website on the Download page in the Maps Collections and Legend Code pages.

• Legend Description is a document (pdf; see an example here) that includes each class level, class name, class description and class equivalence with other global and national references (e.g. the FAO ,  IUCN Ecosystem Typology, and IPCC ones). It is available  in the MapBiomas Platform/Land Cover Box and website on the Download page in the Maps Collections and Legend Code pages. This document is also included as an Annex of the Algorithm Theoretical Basis Document (ATBD) .

• Class Characterization is a document (pdf/ppt) that includes local information, technological and regulatory aspects, temporal dynamics information, medium and high resolution satellite, photos of the landscape and or schematic representation of the class (Figure 2). It may be available in MapBiomas Platform together with the legend. This document can also be included in the Annex of the ATBD and on the information icon next to each class in the Data Platform. 

Figure 2. Example of the information used in the description and characterization of Forest Formation class.

• Consolidated Network Legend Table is a document (.xlsx) that compiles all legends from all territories and includes each class level, class name, class ID number and observations or comments. It is available in English, Spanish, Portuguese and Bahasa Indonesia. In this document we identify consistency between the contiguity of two territories and discrepancies that need to be solved. This is an internal document available in this link.

• Evolution of the Legend is a document (.xlsx) that compiles all legends from all collections from each territory and includes each class level, class name, class ID number and observations or comments. It is available also in the Annex of the ATBD (pdf; see an example in the Annex IV of Brazilian ATBD of Collection 9).

• All the reference documents -the global and national ones used in the legend description and definition process should be available in the Legend Description document and ATBD.

This section describes the data and variables produced and presented in other initiatives.

Other initiatives have their own classifications, which also evolve over time.

• The fire initiative produces the annual map of burn scars or the monthly monitoring of burn scars. The data are presented as:

• LULC (Land Use and Land Cover) classes within the burn scar,

• month of fire occurrence,

• burning frequency categories,

• years since the last burn,

•  burned-area size categories.

• The water initiative produces the annual map of the water surface. The data are presented as:

• % frequency of water presence,

• water/non-water transitions,

• trends in water gain or loss,

• types of water bodies (natural, reservoirs, hydroelectric plants, or mining).

• The soil initiative produces a map estimating soil organic carbon. The data are presented as:

• 10 tons of Carbon per hectare categories.

• The MapBiomas Alerta initiative is a system for validating and refining alerts of native vegetation loss using high-resolution satellite imagery.