height 777 km, elevation angle 23 degrees mid-track), a DTM
has been used to simulate illumination effects in direction
intervals of 22.5 degrees and to examine the results (Domik
et al. 1985). For this limited arrangement, it oan be
stated that any two flights with an intersection of some 20
degrees should be enough to extract all major linear geo-
morphological features. However, in areas with only subtle
changes in relief and/or vaguely expressed lineaments, which
strike (sub-) parallel to the illumination direction, two
simulated flight paths with only 20 degrees difference might
be insufficient. More definite statements on that topio
would need more intensive investigations with different
types of terrain.
Although radiometric analysis aiming at a classifi-
cation of surface cover types is not understood to be a task
of radargrammetry proper, simulation of thematic contents in
SAR images can certainly lead to a better understanding of
radar image data. Thus, it is obvious that in the long term
radar image simulation will not be the ultimate objective
but represents an appropriate tool useful in several
aspects.
Synthetic radar data can assist in the interpretation
of radar images of complex terrain. Combining real and
simulated images allows a separation of thematic information
and relief effects. However, generation of most realistic
SAR images by including the speckle effect would require
highly detailed thematic information, backscatter data banks
and DTMs, which usually are not available. Therefore radar
image simulation will always be an imperfect substitute.
4,2 Benefits of Radar Stereoscopy
Future satellite missions will yield more and more
overlapping strips of SAR images. Like in photogrammetry,
on one hand, radar stereo-pairs are useful to improve
thematic image interpretation; on the other hand, they
enable the restitution of three-dimensional models of the
terrain by stereo-evaluation.
For the human interpreter who wants to exploit radar
stereo-images visually, stereo-viewability is of basic
relevance. However, the active illumination of the terrain
from different sensor positions causes differences in the
geometry as well as in the grey value distribution of the
stereo-partners. Hence, there exists some kind of
contradictory interaction between an optimum stereopsis and
a good geometric configuration of the stereo-images (Leberl
et alı 1985),
For the geomorphological evaluation of radar images,
which is also one of the major keys for lithological and
even more for structural geological mapping, visual stereo-
interpretation plays a most important role. The vertical
exaggeration factor, which usually is between 1.5 and 5, may
328
ouo»ou
pute
a CS at M a t VU P9 AN F0 En Tlf er SU