While this lack of vegetation cover facilitates the use of
remotely sensed data, the fact that about two thirds of the
country are made up of rather flat-lying sedimentary rocks,
thus permitting the extrapolation of field data to larger areas
by satellite image interpretation, is more important in this
context. The basement rocks in the Eastern Desert along the
Red Sea, on the other hand, required a considerably higher
percentage of field work due to the more irregular way the
lithologic units of the basement complex are distributed.
Normally the production of a small-scale map is the final step
in a long and tedious process working in a "bottom-up" man-
ner by compilation of a great number of already existing
large-scale maps. Thus the final small-scale map is a general-
ized version of many mosaic-like pieces of detailed maps.
In the present project, however, detailed large-scale maps
were not available for most of the country. Consequently, a
"top-down" approach had to be used. This meant that basi-
cally small-scale data had to be used throughout the map-
ping, and that the extent of field and laboratory work was
guided by question of just how much detail could be shown
at the final map scale (List, Meißner & Póhlmann, 1989).
In order to meet the objectives lined out above, a two-way
approach was used. One line of work was the collection of
geologic information from image interpretation and field
work, for which optimized color composites without geomet-
ric corrections were utilized, and the preparation of the final
map. The second line was the preparation of a semi-controll-
ed image map in 80 sheets at a scale of 1: 250 000. This
product was to serve as "work sheets" on an intermediate
level in want of any available maps with sufficient detail.
Landsat MSS Published | ,| Field
Digital Data Data Data
DL Image Processing
Optimized B/W
Color Images Images
Geological
Visual Manual Field
Interpretetion Mosaicking Checking
Preliminary Work
Interpret. - Map |< Sheets
1:250 000 1:250 000 Control Point
Bi J Determination
: igital
Re-Interpretation
P Mosaicking Geol. Field
| Checking
Final Geological Digital
Overlay «— Mosaic
1:250 000 1 : 250 000
- Scale Reduction e.
Color Printing
Final Geological Map 1:500 000
Fig. 2: Simplified flow diagram for the preparation of the
1 : 500 000 scale Geological Map of Egypt.
326
2.2 Image processing and interpretation
Remotely sensed data provide information on objects of the
earth's surface, based on sensor characteristics and their
spectral reflectance properties. Lithologic differentiation by
means of Landsat MSS data is not very satisfactory due to
their limited spectral resolution. Therefore, digital classifica-
tion rarely gives satisfactory results when applied to geology.
Visual interpretation, while rather tedious and difficult, is still
the best tool in the hands of an experienced interpreter. It
can take into account not only spectral but also textural
properties of image segments and, what is even more impor-
tant, can make full use of the "expert knowledge" in the
interpreter's mind.
A prerequisite for successful visual interpretation is, of
course, imagery that is geared to the interpreter's needs.
Such imagery can be derived from digital data by means of
image processing. In the present case it was aimed at ob-
taining "optimized" color imagery from early Landsat MSS
data by noise removal, histogram stretching and mild edge-
enhancement filtering. Interpretation itself was performed on
1 : 250 000 color transparencies using light tables and clear
overlays for annotations. The film transparencies were pre-
pared by photographic enlargement from color composite
masters, plotted on an Optronics Colorwrite film plotter.
2.3 Preparing the map base
As mentioned, one of the objectives of the final map was
that its topographic accuracy should conform to international
map standards. That meant that digital mosaics would have
to be created, requiring an appropriate number of control
points identifiable in the imagery. Since such mosaics could
only be prepared toward the of the project, an intermediate
level in the preparation of the final map was introduced.
To this end, a map at the scale of 1 : 250 000, comprising
80 "work sheets", was prepared from black-and-white MSS
images. Uncorrected imagery was used for generating a
semi-controlled image mosaic with an overlay of existing
topographic information from published maps and field sur-
veys, e.g. for new roads or settlements. These work sheets
also served as a base for field work and compilation of inter-
pretation data.
2.4 Field and laboratory work
Due to the remoteness of most of the area, field work had to
be planned and executed in an expedition-like style. A major
consideration was to conduct the field trips in the most
effective way possible. Small field parties typically consisting
of a team of 2 to 4 four geologists and a mechanic were set
up, using two or three four-wheel drive vehicles and a sturdy
field truck carrying fuel and water. Duration of field trips was
3 to 6 weeks. Since no base camps were made and only
flying camps were used at night, large areas could be cov-
ered during these trips and a considerable amount of infor-
mation gathered in a short time.
Rock and stratigraphic samples were collected for subse-
quent laboratory analysis. All in all, the results of over 200
man months of field work were input into the map. Part of
the field work was carried out by post-graduate students of
the Free University and the Technical University of Berlin
under a research project (Sfb 69) funded by the German Re-
search Foundation.
For orientation in the field, color enlargements of the Landsat
MSS data at 1 : 250 000 were used, together with the work
sheet:
tation
prior t
and,
subse
makir
essar
Durint
geom
field.
Trans
night
positi
the re
point
suffic
2.5 |
The c
aratic
overl:
As a
digita
scale
durini
were
tions
indivi
basis
of 9 |
1 mil
radio
tiona
RMS
map.
The ¢
inter]
film 1
the v
light
like li
were
com[
chec
tory
able
The
mos:
1:2
grap
highl
line 1
From
Sinc
coloi
three
grea
foun
map
avail
Cair
2.6
The
cour
