411 
3.5 Specific procedures utilized for generating the 3.5.4 Development of the specific terrain unit 
soil/vegetation/land form base maps descriptions 
The following sections will detail the procedures 
utilized to generate the S/V/LF base maps and the 
subsequent determination of forest surface area. 
These procedural descriptions are made in the 
context of the eventual use of the maps, and forest 
surface area figures, in generalized planning 
activities. 
3.5.1 Initial stratification using LANDSAT images 
The zone indicated in Figure 1 was initially broken 
down into 26 natural regions, each having a 
characteristic mix of geology, climate, soils, land 
use pattern, and vegetation community conditions. 
Lines separating each natural region were first 
placed as a result of interpretation of a black 
and white mosaic of LANDSAT images at a scale of 
1/2,500,000. The placement was then refined using 
the CIR format LANDSAT images at the scale of 
1/200,000. 
3.5.2 Stratification within the natural regions 
Within each natural region the area is further 
divided into sub-zones homogeneous from the standpoint 
of the reflectance characteristics of the LANDSAT 
CIR images. These units, called "cartographic 
units" (CU's), are generally not less than 25 square 
kilometers in contiguous surface area, and are 
characterized by: 1) a single distinct terrain 
condition, or 2) a homogenous mix of several. These 
CU's, with their associated minimum size, were a 
functional necessity due to the scale of mapping and 
were appropriate to the use of the information at 
the planning level. The information will be used to 
identify areas of interest on a national basis, and 
the minimum mapping area selected is well below the 
size requirements for this type of decision. 
3.5.3 Ground data collection 
The objective of the collection of ground data was 
to define the correlation between ground conditions 
and the various reflectance types observed in the 
two remotely sensed image bases available for use as 
the extrapolation implement: the LANDSAT images and 
the IGN aerial photography. Assuming some 
correspondence between ground conditions and the 
LANDSAT signatures, ground data collection sites 
were identified in as many of the different LANDSAT 
signature types as feasible. Each site was selected 
based on it's representativity of a major, or 
important, reflectance type, and, due to normal 
project constraints related to field work time 
availability, it's accessibility by road. 
At each selected site the data collected included: 
1) a detailed list of vegetation species and an 
occular estimate of their relative cover percentage, 
2) a description of the physical characteristics of 
the soil profile, 3) a description of the land form, 
including gross form and details of the terrain 
configuration, 4) a description of the present and 
past land use, and 5) an assessment of the effects 
and/or impacts of these uses. At some of the sites 
vegetation species cover percentages and vegetation 
height data were systematically collected using a 
series of linear transects. In order to permanently 
record the location of each site the IGN aerial 
photos were pierced and annotated with the site 
number at it's interpreted position. The site 
was also located on the IGN topographic map sheets 
and LANDSAT images of the area. 
The next step is to evaluate the ground data and 
form groups having similar conditions of land form, 
soils characteristics and vegetation species 
compositions, which relate to their respective 
forestry resource value. The landform, soils and 
vegetation data from a total of more than 1,200 
field sites have been analysed and 52 groups, called 
"terrain units", or TU's, identified. The specific 
procedure used in the definition of these TU's is as 
follows. Land form and terrain configuration 
information, taken from the field sheets and 
appearance on the aerial photos, was used to make 
initial groupings. The groupings were further 
refined using vegetation species composition, soils 
characteristics and conditions, and individual 
on-site descriptions. After the grouping was refined 
to the point where there was no major confusions 
regarding placement of an individual site, the TU 
descriptions were developed. These descriptions 
included: 1) a verbal description of the land form 
supported by schematic diagrams, 2) a description of 
the physical characteristics of a "typical" soil 
profile, 3) a listing of the primary, or 
characterizing dominant species of vegetation, with 
their respective cover percentages, 4) a list of the 
field sites used to develop the description, and 5) 
a vegetation species matrix showing the percentage 
categories of each individual species observed at 
each field site. An example of one of these 
descriptions, for the series 841, is presented in 
Figure 2. 
TERRAIN UNIT: 841/III 
TERRAIN TYPE: Flat, or slightly undulating, top of a later 
itie plateau. 
SOILS: The soils are typically shallow to a laterized hard- 
pan. The surface layers are sandy silts changing to a 
skeletal structure with a silt dominated matrix near 
the restrictive layer. Below the laterized layer, which 
can be in varying thicknesses depending on specific 
location, is generally an old horizon with moderate 
clay content. Associated with undulations of the later- 
ite layer, there can be depressions with deeper soils. 
VEGETATION TYPE: The vegetation type is in the form of a 
"fourree" or "brousse tigree", ie. there are small 
areas of very dense vegetation associated with small 
depressions and large areas in between with very sparse 
vegetation. The species diversity is high in the 
depression vegetation community and low in the inter- 
depressional zone. The area occupied by the depression- 
associated vegetation community is generally less than 
25%. The species that characterize the composition are: 
SPECIES 
COVER PERCENTAGES 
Acacia ataxancantha 
1-5% 
Boscia anguistifolia 
1% 
B. sengalensis 
1% 
Combretum glutinosum 
1-5% 
C. micranthura 
10-25% 
C. nigricans 
5-10% 
Comiphora africana 
1% 
Guiera senegalensis 
10-25% 
Grewia flaviscens 
1% 
FIELD SITES: D24-44 
5-08 
5-36 
5-03 
5-02 
5-30 
6-50 
5-47 
5-48 
6-74 
3-38 
3-01 
3-35 
3-36 
5-62 
5-51A 
4-43 
5-49 
5-57 
5-72 
5-73 
5-71 
4-52 
9-02 
2-10 
2-02 
2-09 
1-10 
2-49 
3-21 
5-05 
1-29 
Figure 2. The initial page of the description for 
the terrain unit 841/III. This represents a 
"typical" site and is a synthesis of ground data 
collected at the sites listed on the sheet.