Keywords

1 West Africa Land Use Land Cover

Product

LULC general

Dates

1975, 2000, 2013

Formats

Raster

Pixel size

2 km

Thematic resolution

30 classes:

2 (a), 5 (ag), 12 (v), 3 (m), 3 (na)Footnote 1

Compatible legends

None

Extent

West Africa and Cape Verde

Updating

Not expected

Change detection

Yes

Overall accuracy

Not specified

Website of reference

Website Language English

https://eros.usgs.gov/westafrica/

Download site

https://eros.usgs.gov/westafrica/data-downloads

https://www.sciencebase.gov/catalog/item/5deffc05e4b02caea0f4f3fc

Availability

Format(s)

Open Access

.tiff

Technical documentation

CILSS (2016)

Other references of interest

Cotillon (2017), Cotillon and Mathis (2017)

Project

West Africa Land Use Dynamics was a project led by the AGRHYMET Regional Centre in collaboration with the Sahel Institute (INSAH), the USGS Earth Resources Observation and Science (EROS) and the US Agency for International Development (USAID). 17 different countries took part: Benin, Burkina Faso, Cape Verde, Chad, Ivory Coast, Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Mauritania, Niger, Nigeria, Senegal, Sierra Leone and Togo.

As a result of the project, a LUC map series was created to monitor natural and environmental trends in the West Africa region. The dataset is part of a wider effort to create an atlas about landscape and environmental changes in West Africa.

Production method

West Africa Land Use Land Cover was obtained through photointerpretation of Landsat imagery with the Rapid Land Cover Mapper (RLCM) tool at a spatial resolution of 2 km. Gambia was photointerpreted at a spatial resolution of 2 km and Cape Verde at 500 m. Photointerpretation guidelines were developed specifically for the task.

Product description

Users can download a separate edition of the West Africa Land Use Land Cover dataset for each year of reference. In each case, the download includes the raster file with the LUC map as well as a file to symbolize the raster in GIS. An Excel file with the legend can also be downloaded from the website together with the detailed metadata files for each LUC map.

Downloads

West Africa Land Use Land Cover (2013)

– Raster file with LUC map

– File to symbolize the raster in GIS (.clr)

Legend and codification

Code

Label

Code

Label

0

No data

15

Gallery forest and riparian forest

1

Forest

16

Shrub and tree savanna

2

Savanna

21

Degraded forest

3

Wetland – floodplain

22

Bowé

4

Steppe

23

Thicket

5

Oasis

24

Agriculture in bottomlands and flood recessional

6

Plantation

25

Woodland

7

Mangrove

27

Cropland and fallow with oil palms

8

Agriculture

28

Swamp forest

9

Water bodies

29

Sahelian short grass savanna

10

Sandy area

31

Herbaceous savanna

11

Rocky land

32

Shrubland

12

Bare soil

78

Open mine

13

Settlements

98

Cloud shadow

14

Irrigated agriculture

99

Cloud

2 SERVIR-ESA—SERVIR Eastern and Southern Africa

Product

LULC general

Dates

1990, 2000, 2010 (Malawi)

1990, 2000, 2010, 2015 (Rwanda)

2000, 2010 (Botswana, Namibia, Tanzania, Zambia)

2000, 2014 (Lesotho, Uganda)

2003, 2008 (Ethiopia)

Formats

Raster

Pixel size

30 m

Thematic resolution

7 classes:

1 (a), 1 (ag), 2 (v), 0 (m), 1 (na)

Compatible legends

IPCC

Extent

Eastern and Southern Africa

Updating

No updating confirmed

Change detection

Yes, but potential uncertainties have not been specified

Overall accuracy

Expected to be >63%

Website of reference

Website Language English, Spanish, French

https://www.servirglobal.net/ServiceCatalogue/details/5bd052d451ebdcae79683375

Download site

http://geoportal.rcmrd.org/layers/?limit=100&offset=0

Availability

Format(s)

Open Access

.tiff

Technical documentation

Oduor et al. (2016)

Other references of interest

Al-Hamdan et al. (2017), Searby et al. (2019)

Project

SERVIR is an initiative led by the National Aeronautics and Space Administration (NASA) and the United States Agency for International Development (USAID) that aims to help developing countries to produce geospatial information suitable for climate risks and land use management. SERVIR operates in West Africa, Eastern and Southern Africa, Hindu Kush Himalaya, the Lower Mekong, South America and Mesoamerica.

In Eastern and Southern Africa in 2008, SERVIR started a project in partnership with the Kenya-based Regional Centre for Mapping of Resources for Development (RCMRD). Training, geospatial tools and geospatial datasets were developed as part of the project, including a dataset specifically aimed at LUC monitoring. Six countries were initially mapped (Botswana, Malawi, Namibia, Rwanda, Tanzania, and Zambia), with three more countries participating since 2014/15 (Ethiopia, Uganda, and Lesotho).

As a result of this project, a LUC map covering all 9 countries was developed. National LUC maps, with detailed national legends, were also provided as part of the project.

Production method

SERVIR-ESA was produced by aggregating LUC maps created at the national level according to the same 7-class classification scheme. For each country, a map with a legend adapted to the country’s specificities was also developed following the same general guidelines.

The maps were obtained through supervised classification of Landsat imagery through a Maximum likelihood classifier. Auxiliary spatial and non-spatial data were also used in the classification. Settlements were manually photointerpreted from Google Earth imagery.

Errors and uncertainties in the classification resulting from this process were corrected in a post-classification step, which included expert review.

Product description

The SERVIR-ESA LUC map is distributed at national level. For each country, users can download the harmonized map for all Eastern and Southern Africa (Scheme I) or the specific LUC map with a detailed legend for the selected country (Scheme II).

Downloads

Scheme I product/Scheme II product

– Raster file with coloured LUC map

Legend and codification

In this description, we only include the general 7-class legend adopted for all the LUC maps. However, a specific legend is available for each national map, which can be consulted online.

Scheme I legend

Code

Label

Code

Label

0

Non data

4

Wetland

1

Forestland

5

Settlement

2

Grassland

6

Other land

3

Cropland

  

Practical considerations

The maps for each country were usually produced at different dates, so making inter-country comparison difficult.

3 SADC Land Cover Database

Product

LULC general

Dates

1990 / 91 (Malawi)

1997 (Tanzania, Zimbabwe)

1999 (Mozambique, South Africa, Lesotho, and Swaziland)

Formats

Vector

Scale

1:250,000

Thematic resolution

13 classes:

1 (a), 2 (ag), 5 (v), 0 (m), 1 (na)

Compatible legends

None

Extent

Southern African Development Community (Lesotho, Malawi, Mozambique, South Africa, Swaziland, Tanzania, Zimbabwe)

Updating

No

Change detection

No (only one date)

Overall accuracy

Not specified

Website of reference

Website Language English

http://gsdi.geoportal.csir.co.za/projects

Download site

http://gsdi.geoportal.csir.co.za/projects

Availability

Format(s)

Open Access

.shp

Technical documentation

Other references of interest

Project

The SADC Land Cover Database is fruit of a project funded by the South African Department of Arts, Culture, Science and Technology (DACST) through the Regional Science and Technology Programme. It was coordinated by the Council for Scientific and Industrial Research (CSIR) in South Africa, with the participation of organizations from the different countries being mapped.

The objective of the project was to deliver a coherent Land Use Cover map covering the Southern African Development Community (SADC) region. The project builds on earlier LUC mapping work carried out at national and regional scales for each of the mapped countries.

The map covers those SADC countries that already had a LUC dataset available for their territory: Lesotho, Malawi, Mozambique, South Africa, Swaziland, Tanzania, Zimbabwe. The other countries in the region are not included in the map.

Production method

The SADC Land Cover Database was obtained by harmonizing and fusing the different national and regional LUC datasets. All the datasets were originally obtained by classification or photointerpretation of Landsat imagery, although the reference years vary from country to country.

The maps were combined by resampling to a spatial resolution of 1 km, before being reclassified according to the same classification system. This reduced the detail of the original maps, a deliberate action to avoid copyright and commercialisation issues.

Product description

The dataset is downloaded as a single compressed file (.zip), which includes the vector LUC map, a metadata file and a complete map (i.e. with colours, graphics, scale and legend) in jpg format that is ready to print out.

Downloads

SADC

– Vector file with LUC map (.shp)

– Edited map in a non-modifiable format (.jpg)

– Metadata file (.html)

Legend and codification

Label

Label

Label

Forest

Bare ground

Open water

Woodland

Plantation

Wetland

Bushland

Cultivation

Ice-cap/Snow

Low shrubland

Built-on

Not classified

Grassland

  

Practical considerations

A detailed description of the map categories is available in the dataset’s metadata. The map’s production method entails certain limitations and uncertainties, in that each country has been mapped by a different team, using different sources of imagery for different reference years. Inconsistencies may therefore arise when comparing information between countries.

4 AFRICOVER

Product

LULC general

Dates

1994 / 01

(the reference year varies according to the country)

Formats

Vector

Scale

1:200,000

Thematic resolution

8 classes:

2 (a), 2 (ag), 2 (v), 0 (m), 0 (na)

Compatible legends

FAO-LCCS

Extent

Africa

Updating

No

Change detection

No (only one date)

Overall accuracy

Expected to be > 80%

Website of reference

Website Language English

http://www.fao.org/geospatial/projects/detail/en/c/1035404/

Download site

http://www.fao.org/geospatial/projects/detail/en/c/1035404/

Availability

Format(s)

Open Access

.shp

Technical documentation

Di Georgio and Jansen (1996), FAO (1997)

Other references of interest

Di Gregorio (2009), Kalensky (1998), Latham et al (2002)

Project

AFRICOVER was a project led and coordinated by the Food and Agriculture Organization (FAO) of the United Nations, which aimed to create georeferenced data for the African continent. The FAO helped the different countries and regions to develop their reference maps, establishing the standards for the final product. Twelve countries participated in the project (Burundi, Democratic Republic of Congo, Egypt, Eritrea, Kenya, Rwanda, Somalia, Sudan, Tanzania, Uganda, Libya and Malawi), which therefore required extensive coordination of many national and regional teams across Africa.

A keystone of the project was the production of LUC maps for Africa. In addition to LUC maps, other georeferenced data were created for a range of themes: hydrology, geomorphology, demography…

Production method

The production of AFRICOVER was decentralised at a national and regional level. Although the FAO defined the guidelines and standards for the product, national and regional teams from each country were responsible for its execution. This meant that although a set of common characteristics regarding the production of AFRICOVER had been established for all the countries involved, certain specificities could also arise.

AFRICOVER LUC maps were mainly obtained through photointerpretation of satellite imagery, of which Landsat was the main source. The photointerpretation scale was 1:200,000. When drawing LUC polygons, the FAO LCSS classification scheme was followed. The FAO provided national and regional teams with specific software and training to carry out LUC mapping according to this approach.

Product description

AFRICOVER LUC maps are distributed at a national level. A compressed file can be downloaded for each country. This includes the vector LUC map and a legend description to help users interpret it.

Downloads

Land cover folder

– Vector file with LUC map (.shp)

– PDFs describing the classification legend

– Excel file with the classification legend

Legend and codification

Label

Label

Cultivated Terrestrial Areas and Managed Lands

Artificial Surfaces and Associated Areas

Natural and Seminatural Terrestrial Vegetation

Bare Areas

Cultivated Aquatic or Regularly Flooded Areas

Artificial Waterbodies

Natural and Seminatural Aquatic Vegetation

Inland Waterbodies

Practical considerations

AFRICOVER LUC maps have been created following the FAO LCSS classification scheme. This means that each LUC polygon is described through a specific code that identifies the general cover of the polygon and characterizes it through a series of labels. Users may find this system difficult to understand, as it does not follow a common hierarchical classification legend in which each polygon is defined by a single category.

5 CCI LAND COVER – S2 PROTOTYPE

Product

LULC general

Dates

2016

Formats

Raster

Pixel size

20 m

Thematic resolution

10 classes:

1 (a), 1(ag), 5 (v), 0 (m), 0 (na)

Compatible legends

FAO-LCCS

Extent

Africa

Updating

Expected, but no specific date has been set

Change detection

No (only one date)

Overall accuracy

Expected to be >65%

Website of reference

Website Language English

http://2016africalandcover20m.esrin.esa.int/

Download site

http://2016africalandcover20m.esrin.esa.int/download.php

Availability

Format(s)

Open Access after registration

.tiff

Technical documentation

Lasiv et al. (2017)

Other references of interest

Project

The CCI LAND COVER – S2 PROTOTYPE map is part of the Land Cover – Climate Change Initiative led by the European Space Agency (ESA). The purpose of this initiative is to deliver Land Cover products that meet the requirements of the climate change research community.

The map was created as a prototype to collect feedback from users for future improvements. At the time it was released, it was the highest spatial resolution LUC map covering the whole African continent and one of the few products providing consistent LUC coverage for all of Africa.

Production method

The map was obtained after classification of Sentinel-2A imagery for the reference year 2016. Two different classifications were carried out, through random forest and machine learning classifiers. The final map is a combination of the two classifications. Auxiliary datasets were used to map the “open water” (extracted from the Global Surface Water product) and “urban areas” (extracted from Global Human Settlement Layer and the Global Urban Footprint) categories.

Product description

CCI LAND COVER – S2 PROTOTYPE is distributed as a single compressed file, including the raster with LUC information and a style layer to symbolize the map in GIS software. The legend is described in two auxiliary files, in Excel and pdf.

Downloads

S2 PROTOTYPE LC at 20 m AFRICA 2016

– Raster file with LUC map (“ESACCI-LC-L4-LC10-Map-20 m-P1Y-2016-v1.0”)

– Layer style file for GIS software (.qml)

– Excel sheet with the map legend

– PDF describing the map legend

Legend and codification

Code

Label

Code

Label

1

Tree cover areas

6

Lichens and mosses

2

Shrubs cover areas

7

Bare areas

3

Grassland

8

Built up areas

4

Cropland (rainfed or irrigated)

9

Snow and/or Ice

5

Vegetation aquatic or regularly flooded

10

Open water

Practical considerations

The map is distributed as a single, very heavy file (6 Gb). Users with limited computer and internet capacities may find it difficult to download and work with this product. Nonetheless, a preview tool is available online for any user wishing to consult the map.

6 Congo Basin Vegetation Types

Product

LULC general

Dates

2000 / 07

Formats

Raster

Pixel size

300 m

Thematic resolution

20 classes:

1 (a), 2 (ag), 14 (v), 1 (m), 0 (na)

Compatible legends

FAO LCCS

Extent

Congo Basin region

Updating

Not expected

Change detection

No (only one date)

Overall accuracy

Expected to be > 71%

Website of reference

Website Language English

http://maps.elie.ucl.ac.be/geoportail/

Download site

http://maps.elie.ucl.ac.be/geoportail/

Availability

Format(s)

Open access

.tiff

Technical documentation

Verhegghen et al. (2012)

Other references of interest

Project

The Congo Basin Vegetation Types map was produced by a team of experts from the Université Catholique de Louvain, the Joint Research Centre (JRC) of the European Commission and the Observatory for the Forests of Central Africa (OFAC).

The map was produced in an attempt to aid forest and vegetation monitoring in Central Africa. It provided a spatially coherent dataset for all the Congo Basin region with improved spatial discrimination with respect to previous datasets of similar nature.

Production method

The Congo Basin Vegetation Types was obtained by unsupervised classification of imagery composites created from the images provided by the MERIS and VEGETATION sensors.

To account for the regional disparities of the mapped area and its different cloud coverage, the Congo Basin was split into four different zones: North, South, Western Centre and Eastern Centre. Seasonal imagery composites were created for each specific season in the northern and southern regions. In addition, an annual composite was generated for the whole mapped area.

A different classification exercise was performed for each mapped zone based on a cluster k-means algorithm. The resulting clusters were labelled on the basis of the information provided by reference maps when LUC information on these sources covered at least 50% of the identified cluster. The rest of the clusters were manually labelled on the basis of visual interpretation and expert knowledge.

Product description

A compressed file (.zip) containing the raster layer with the LUC data can be downloaded, together with other auxiliary information to interpret and symbolize the map content.

Downloads

Congo Basin Vegetation Types map

– Raster file with LUC map (.tif)

– Layer style files for ArcGIS (.lyr)

– Excel file with the map legend (.xls)

– Text file with the metadata for the product (.txt)

Legend and codification

Code

Label

Code

Label

1

Dense moist forest

11

Grassland

2

Submontane forest

12

Aquatic grassland

3

Mountain forest

13

Swamp grassland

4

Edaphic forest

14

Sparse vegetation

5

Mangrove

15

Mosaic cultivated areas/ vegetation

6

Forest/savanna mosaic

16

Agriculture

7

Rural complex (forest area)

17

Irrigated agriculture

8

Closed to open deciduous woodland

18

Bare areas

9

Savanna woodland/tree savanna

19

Artificial surfaces and associated areas

10

Shrubland

20

Water bodies

Practical considerations

Users can consult the LUC map online on the Université Catholique de Louvain website (http://maps.elie.ucl.ac.be/geoportail/).