Rocky, shallow soils and deep, highly leached and 
moderately leached red clay with a humic topsoil 
prevail in the two zones. Neither the aerial photo 
graphs nor the satellite imagery were of great use 
for the survey of this part of the region. As in 
Zone 14 the vegetation / land use determines the 
reflectance characteristics. Therefore, only some 
major (land unit) boundaries could be mapped from 
the satellite imagery. Cloud cover (parts of Zone 
3) and extensive burnt areas caused additional in 
terpretation problems. 
The aerial photographs were even less useful than 
the satellite imagery. Due to the poor quality of 
the photos and the uniform vegetation / land use 
little discrimination was possible on the basis of 
graytone and texture. Relief-related boundaries, 
which could be delineated on the photos, were most 
ly too detailed for the scale of the survey. 
As a consequence of these constraints the mapping 
had to be very stronly based on the field observa 
tions. Most boundaries were delineated on the basis 
of an interpolation of the field observations sup>- 
ported by the interpretation of the contour lines 
of the toposheets. 
7. SUMMARY OF THE FACTORS INFLUENCING THE UTILITY 
OF DIFFERENT TYPES OF REMOTE SENSING DATA 
As described in the previous chapter, various 
factors influence the usefulness of different types 
of remote sensing data. Three major groups of fac 
tors can be distinguished: 
- soil-related factors, 
- area-related factors, 
- factors related to the characteristics of the 
survey and the type of remote sensing data. 
Usually there is a strong interdependence betweer 
the three factor groups; therefore the following 
classification is somewhat artificial. 
7.1 Soil-related factors 
The soil colour was the most important soil-re 
lated factor. In dry areas where sufficient bare 
soil is exposed, the colour differences were very 
well reflected by the IHS image. The colour differ 
ences were less pronounced on the FCC, and they are 
of little use for the interpretation of black & 
white aerial photographs. 
Field checks showed, however, that not all colour 
differences on the satellite imagery correlate with 
soil boundaries. In parts of Land Unit 16 I, for 
example, a veneer of light sand covers a dark loam. 
These few millimetres of sand on the soil surface 
cause a colour change from green to yellow on the 
IHS image. Vice versa, sane areas with identical 
colour reflectance characteristics turned out to 
have very different soils. Examples for both cases 
are demonstrated in Table 3 and Figures 1 and 2. 
Soil drainage was the second important soil-re 
lated factor. Since drainage differences usually 
coincide with a strong change in vegetation, the 
resulting reflectance characteristics of e.g. a 
poorly or imperfectly drained area are normally 
very different from those of the surrounding land. 
Thus the drainage differences are reflected by a 
clear change in colour or graytone. In contrast to 
soil-colour differences variations in drainage can 
usually be well mapped on all three typos of remote 
sensing data and in all agro-ecological zones. 
7.2 Area-related factors 
Climate, relief and vegetation / land use are the 
most important area-related factors. Satellite ima 
gery proved to be most important in the drier 
agro-ecological zones where, due to the limited ve 
getation cover, much bare soil surface is exposed. 
In these areas excellent use could be made of the 
spectral resolution qualities of the satellite ima 
gery (especially of the IHS images). On the other 
hand, areas with more rainfall usually conincide 
with a denser vegetation cover which often hides 
soil boundaries. 
The black & white aerial photographs rendered the 
greatest use in an area with complex relief and 
preferably relief-determined soil boundaries, while 
they were much less valuable in flat areas. 
Land-use boundaries were often found to have 
little correlation with soil boundaries, some 
land-use types (e.g. afforestations) sometimes com 
pletely prevent the detection of any soil boundary. 
The habit of land burning is another important 
factor. These areas are distinctly reflected on the 
satellite imagery. The burnt areas cause napping 
problems for two reasons. First, their reflectance 
characteristics are very similar to those of sane 
areas with dark cracking clay. Also some rocky 
areas sometimes cause similar reflectance characte 
ristics. Second, the dark reflectance of the burnt 
vegetation prevents the detection of soil-colour 
differences. 
7.3 Factors related to the requirements of the 
survey and the typie of remote sensing data 
Only three main aspects are to be listed under 
this heading: 
- spatial resolution, 
- spectral resolution, 
- stereoscopic qualities. 
For a soil survey at the scale of 1:100,000 re 
mote sensing material of at least the same scale is 
required to get a sufficient spatial resolution. 
Therefore the available 1:250,000 MSS enlargements 
were not always able to supply the necessary detail 
for this soil survey. However, although some small 
er units could not be mapped from the satellite 
imagery, generally the image resolution was still 
adequate for the achievable mapping detail under 
the given survey conditions. The spatial resolution 
of the 1:50,000 aerial photographs proved to be 
ideal for the mapping scale. 
The advantages of the satellite imagery concern 
ing the spiectral resolution and the superiority of 
the aerial photographs in stereoscopic qualities 
are well known and need no further explanation. 
8. COLOUR-INTERPRETATION KEYS FOR SATELLITE IMAGERY 
OF SELECTED AREAS 
In a small-scale soil survey it is usually at 
tempted to correlate the satellite image characte 
ristics colour, texture and brightness with the 
field findings. Hereby colour-interpretation keys 
which relate the image colours to the different 
soils play an inportant role. 
To demonstrate the limitations of such keys, the 
IHS image colour keys for two selected land units 
are presented in Table 3 and Figures 1 and 2. 
The results demonstrate that even within the same 
(relatively small) agro-ecological land unit 
- areas with very different soils may give iden 
tical colours on the satellite images and, vice 
versa, 
- the same soil may result in different image co 
lours. 
The soil keys also show that it is impossible to 
establish a key which is valid for different agro- 
ecological units. 
Additionally, the two figures make plain the dif 
ferences between the IHS- and the FCC imagery. Both 
figures excellently display the very good colour 
discrimination of the IHS image. High albedo values 
considerably reduce the colour contrast of the FCC 
(compare especially the upper right corner of Fig 
ure 1). The area marked with "a" in Figure 1 oor-