603 
ie braided river 
lepositional pro- 
rse soil mate- 
1 be distinguished; 
ralleytype in the 
in the down slope 
i rate of erosio- 
iccumulation. 
rtype in the low 
iu. This valley- 
,on. 
íg in the main 
! developed after 
die pathway 
l rock. 
isert is charac- 
i, especially 
material is de- 
. The debris 
.ed in the adja- 
ey. 
dsat images of 
U's different 
boundaries, 
f these PMU's 
microphotomorphic 
rmation about the 
¡en is visible, 
.ess than 1%, 
:0%., a faint 
red or magenta 
that the micro 
be red colour 
reas they become 
tges. A photomor- 
r and fine tex- 
Valley on the 
t fringes are 
getation resul- 
colour compo- 
re strong and 
reas inside the 
near inf ra red 
a continuous 
magenta, which indicates a healthy vegetation. These 
observations could be made easely on the enlarged 
colour composites (scale 1:10C,000). 
Band 7 was used to delineate the areas in the de 
sert fringes, which are still resisting the deserti 
fication and these which have been covered by the 
sand sheet. Sample areas were located upon aerial 
photographs and studied in detail. The steroscopic 
view revealed the encroachement of the eolian depo 
sits along the Nile Valley and covering already some 
fields (fig. 7) . 
Fig. 7 - An aerial photograph of the inter 
ference zone between the Nile volley 
and the western desert (scale 
1 : 40,000). 
It was also possible to detect and map the damaged 
irrigation network by sand deflation, the remnants 
of the damaged villages were also visible. 
This study is concerned with the assessment of 
desertification in the areas bordering the Nile 
Valley, although different photomorphic units could 
be distinguished, in the Nile Valley they were not 
taken into account. 
5. FIELD OBSERVATION AND LABORATORY-ANALYSIS 
An intensive purposive field observation was perfor 
med. Linear traverses scheme , edited by Justice 
and Townshened (1981) has been followed. It resulted 
in a detailed description of the terrain, necessary 
for the understanding of the previous image inter 
pretation. 70 soils samples were collected occuring 
in the different physiographic units. Particle size 
distribution was determined for all the samples. 
Frequency histograms of particle size distribution 
show in most of the sand sheet and interdune samples 
(fig 8A) a wide range including an amount of fine 
material. The frequency distribution of sand dune 
samples (fig 8B) revails a narrow range of grading 
without a marked amount of fine material. The inter 
ference zone samples show a higher amount of the 
coarse fraction at the surface of the cultivated area 
(fig.8C) than in the sub-surface. The contrary was 
found in the samples of the non-cultivated "deserti 
fied" areas(fia,8D). That might indicate the conti 
nuous contribution of eolian sediments in this zone. 
10 
-1 r7ÍT~rTÍ~jTn |~i~r-f~)- 
0.0 2.0 4.0 6.0 8.0 
mean grain diameter $ 
tig. 8 - Histograms of the particle si/e distribution of different physiographic , 
A. An interdune ( ) and a sand sheet sample ( ) 
8. Sand dune sample 
C. Non-cultivated area of the interference zone . 
surface samp I 
D. Cultivated area of the interference zone subsurface 
E. Sample of the regular cultivated area in the Nile Valley. 
The histogram of the particle size distribution of 
the regular cultivated Nile Valley (fig. 8E) shows 
a wider range of grading and the absence of marked 
modal frequency in most of the samples. Some samples, 
in the western borders, have a naarow range of grading 
which could be attributed to an encroachment of desert 
sediments inside the valley. A relative increase of 
the coarse constituents was found in the sites adja 
cent to the Eastern desert cliffs. That might reflect 
the influence of the Eastern desert. 
6. MAPPING OF SOIL CONDITIONS 
All the results from the landsat and aerial photo 
interpretation, field observation and laboratory ana 
lysis have been fed back in the photomorphic unit maps. 
A map of soil conditions (fig. 9)was derived on the 
base of a system modified from El-Shazly et al (1978) . 
The study area has been divided, according to the 
potential land use, to arable and non arable. The a- 
rable area is further subdivided, according to the 
priority in agricultural development into the follo 
wing classes : Grade I ; soils of the river Nile flood 
plain, Grade II ; soils of mixed Nile alluvium and 
eolian sediments, Inside this group a sub-divisious is 
possible in cultivated land and desertified land. The 
size of these units is so small that they could not 
be mapped on such a small scale. The non arable areas 
are divided into following grades, Grade III ; soils 
of sand sheet belts, Grade IV ; soils of sand dune 
belts and Grade V the desert pavements. 
Grade VI ; the slightly eroded plateau, Grade VII ; 
highly sloping eroded plateau, Grade VIII ; wady bot 
tom and debris accumulation zone with coarse soil ma 
terial. 
7. CONCLUSIONS 
The results of remote sensing application in the 
assessment of desertification could be considerably 
improved by using a multiple view approach ; multi 
stage, multiscale and multispectral. 
7.1 Multiscale sensing 
The multiscale approach indicates the analysis of 
satellite data in conjunction with the aerial photo 
graphs. The synoptic view, provided by the satellite