Using the USGS/IGBP Global Land Cover Characteristics dataset (Loveland et al. 2000) as a base for analysis, the land cover of the Limpopo River basin is relatively hetereogenous, with four main classes dominating the landscape:
Cultivated agriculture
Savanna
Grassland
Shrubland
This information agrees with several other regional and global estimates of land cover, including the Global Land Cover Facility data (2005) and World Resources Institute (2003).
The distribution of the main landcover classes across the Limpopo River basin are shown in the map, table and chart below.
PRIMARY LANDCOVER CLASSES OF THE LIMPOPO RIVER BASIN. SOURCE: USGS/IGBP 2010
Summary of main generalised land cover for the Limpopo River basin.
Landcover type
Area (Ha)
Percentage of total
Bareground
112,154
<1
Cultivated land/grassland
21,678,104
53
Grassland
10,306,742
25
Savanna
8,067,132
20
Shrubland
192,269
<1
Forest
481,043
1
Urban
48,014
<1
Waterbodies
228,714
<1
Total
41,114,172
Source: Hatfield/ USGS 2010.
RELATIVE PROPORTION OF LAND COVER IN THE LIMPOPO RIVER BASIN.
Land Cover Data Availability
Although there are national land cover datasets available for all of the Limpopo River basin states and in some cases several iterations over a period of time; each country uses a different method to create the data and a different legend or classification scheme. There is some similarity between the assessments completed for Mozambique and South Africa, as their classification schema are both based upon the FAO AFRICOVER system; however, Botswana and Zimbabwe use completely different classification systems.
There are a variety of global datasets available, such as the USGS/IGBP dataset used above. However, to meet the requirements water resources management planning at a basin scale, it is important that a higher resolution data set is developed. There are several approaches to solving this problem. The first is to complete a new assessment of land cover for the Limpopo River basin, based upon new satellite imagery and field work. The second option would be to develop a consistent land cover map from existing land cover data and supporting information the basin. While the second option may result in a lower confidence result that has the potential to date quickly, it would be a relatively cost effective exercise to undertake.
TYPICAL LAND COVER IN THE LIMPOPO RIVER BASIN. SOURCE: SCHAEFER 2010