858 -
will help in drainage delineation, whereas in areas with a tropical forest
cover the vegetational differences may be indicative for stream channels.
There is little difference between stereoscopic and single strip interpreta
tion in these areas (table 1).
4. RADAR IMAGERY FOR THEMATIC MAPPING
Stereo radar is important for thematic mapping where relative altitude differences
are to be measured or observed. Fig. 12a shows the near range image (x band) of
a sediment plane along Salt River, Arizona, USA. Not much radar shadow is present
due to the large depression angle in the near range. The light tones result from the
slopes dipping towards the radar and reflecting a lot of energy due to the high inci
dence angle.
Relative altitude differences between the pediment horizons are not visible. In the
far range a differentiation is visible, but correlation between the altitudes of the
different surfaces is only possible with stereo vision (fig. 12).
For geological interpretation of structures in sedimentary regions use is made of
strike ridges, slope assymmetry, dip-slope/face-slope relation, "v" shape in
valleys, etc. These characteristics are normally well developed on radar images,
when the structures are oriented more or less perpendicular to the scan direction.
With strike ridges parallel or obliquely oriented to the scan direction, interpretation
may turn out to be more difficult when viewing monoscopically. It is well known
that lineaments parallel to the scan direction are not or only faintly visible. This
is partly due to the absence of shadows in linear valleys or behind strike ridges.
In stereoscopic interpretation the strike ridge will be interpretable by its visible
height. A dip slope will be seen stereoscopically as a dip slope, and can be
measured. Many more examples can be given in which the parameter of "height
differentiation by stereoscopic vision" is the decisive parameter for a correct
interpretation.
5. CONCLUSION
Observation of altitude differences and correlation of height levels is of prime
importance for many interpretation purposes. Shadows on radar imagery, giving
the interpreter a relief impression, are in low relief areas often insufficient in the
near range for monoscopic interpretation. In high relief areas the shadows in the
far range may obliterate too much of the terrain information. Correlation of rela
tive altitudes of not immediately adjacent points is of doubtful fidelity on single
strip imagery.