The following sections describe the procedure used in a 2003 FAO study to estimate erosion-sedimentation for Ghana, Kenya and Mali.
The erosion-sedimentation modelling with LAPSUS used five different input maps (ASCII files): DEM, rainfall, infiltration, soil depth and soil erodibility (K factor).
DEM
The most important input was the DEM, used to derive the slope gradients. This DEM needed to be free of sinks and flats (neighbouring grid cells with exactly the same altitude). A new sinkless DEM has been created with ArcInfo (using the ‘fill’ command). With a special program (arc_flat.cpp), developed by Schoorl, this new DEM has eliminated flats by combining the sinkless DEM and a flow-direction map. This new flatless and sinkless DEM served as an input for the LAPSUS model.
RAINFALL
For rainfall, the LAPSUS modelling used the IIASA rainfall map with a resolution of 0.5 degrees (available for the whole of Africa). Within the model, an evaporation factor was set at 0.4 (which means that 60 percent of all rainfall evaporates). This factor might be different for each country. The IIASA map was not used for Kenya because it was too coarse and contained errors, e.g. the effect of Mount Kenya was not visible. Therefore, the study used a climate map by Corbett and White (1998). The IIASA rainfall map was adequate for Ghana and Mali because these countries have a less pronounced topography and a more regular rainfall pattern.
INFILTRATION
The amount of infiltration should be known in order to enable a good estimation of runoff. Infiltration is related to land cover and land use. The infiltration capacity determines the amount of rainfall that infiltrates and the part that runs off. Therefore, the land cover map with the International Geosphere-Biosphere Programme classification was reclassified into eight classes with the following infiltration percentages:
|
· Forest (1-5) |
100 percent |
|
· Shrublands (6-8) |
70 percent |
|
· Grasslands (9, 10) |
85 percent |
|
· Permanent wetlands (11) |
100 percent |
|
· Croplands (12, 14) |
60 percent |
|
· Urban and built-up (13) |
60 percent |
|
· Barren (16) |
20 percent |
|
· Water (17) |
100 percent |
Soil depth
Soil depth determines the available water storage capacity and is, therefore, important to estimating the amount of runoff. Each soil unit was classified according to soil-depth classes. The following classes were distinguished and used to create a soil-depth map:
|
1. > 1 m |
1.2 m |
|
2. 0.4-1 m |
0.7 m |
|
3. < 0.4 m |
0.25 m |
|
0. Water |
2 m |
K FACTOR
The amount of runoff can be calculated using the above input parameters. The erodibility of the soil (detachment factor) should be known in order to estimate the amount of erosion. Each soil unit was classified according to soil erodibility classes. These classes were based on work by FAO (2002) and expert knowledge. The soil-erodibility classes were reclassified to K factors (below) and exported to a grid map.
|
0. Water |
1 × 10-6 |
|
1. Very low |
1 ×10-5 |
|
2. Low |
2 × 10-5 |
|
3. Medium |
3 × 10-5 |
|
4. High |
5 × 10-5 |
|
5. Very high |
1 × 10-4 |
SIMULATION
The run-time of the LAPSUS model simulation was set at one year for this study. The model produced two output files: an ASCII file with runoff; and an ASCII file with the profile change in metres (net erosion-sedimentation). These files can be imported in ArcView and displayed. Further information about the LAPSUS model is available in Schoorl et al. (2002) and Schoorl and Veldkamp (2001).
