×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Remote sensing for resources development and environmental management
Author
Damen, M. C. J.

Figure 8. Rose diagram of dike lineaments in J.
Salala area (a. SIR-B; b. LFC).
LINEAMENT ANALYSIS OVER THE TEST AREAS
The dike lineament data over the four test areas
have been taken together for a general lineament
analysis in 10 degree directional classes. The LFC
photographs allow interpretation of a larger number
of dikes, with the exception of test area 3, de
spite an approximately similar ground resolution
for the LFC photo and radar images. The cumulative
dike lineament length for radar is 778 km for a
frequency of 610. For the LFC photos, there are 810
km and 672 respectively (table 2).
Table 2. Cumulative length and frequency and
length/frequency ratio of dike lineaments for radar
and LFC image interpretation.
Freq.
LFC
Length
in km
LFC
Length
:freq.
LFC
Freq.
SIR
Length
in km
SIR
Length
:freq.
SIR
Area
1
136
150.1
1.10
110
125.1
1.14
Area
2
314
336.1
1.07
260
314.2
1.21
Area
3
173
251.4
1 .45
201
283.8
1 .41
Area
4
49
72.8
1.49
39
55.3
1.42
Total
672
810.5
1 .21
610
778.1
1 .28
The larger sensitivity range in greytones of the
LFC may be the reason for this difference.
The following survey parameters may influence the
differences between lineament interpretations. For
side-looking radar, it is well-known that lineament
interpretability is strongly directional-dependent
(MacDonald et al. 1969, Koopmans, 1983). The look
direction of the radar is always underrepresented,
whereas the azimuth direction (along-flight direc
tion) is often overrepresented. For the SIR-B data
take, these directions are 51° and 321°, respecti
vely.
For the Large Format Camera photos, the sun azi
muth direction may influence lineament detection,
especially when the sun elevation is low. The di
rection parallel to the sun azimuth direction may
Figure 9. Histograms for dike lineaments:
a. all dike lineaments from SIR-B
b. dike lineaments from SIR and not present on LFC
c. all dike lineaments from LFC
d. dike lineaments from LFC not present on SIR-B.
be underrepresented and perpendicular to it overre
presented. For the LFC photo over Sudan, the sun
azimuth was direction 219° under a sun elevation of
49° .
The directions of illumination in both cases
(radar and LFC) are not very different: 50 E from N
(midmark directional class) for SIR-B and 40 E from
N for LFC. The incidence angle of illumination is
52° ± 3 for the radar against 41° for sunlight on
the LFC photographs.
The larger incidence angle for the radar energy
will cause a better enhancement of lineaments on
the radar images. On the other hand, volume scat
tering and ground penetration will be less than
with small incidence angles.
To check the influence of these parameters four
groups of distributional classes were made:
I. All dike lineaments interpreted from SIR-B
images.
II. Dike lineaments interpreted from SIR-B image
but not present on LFC photos.
III. All dike lineaments interpreted from LFC
photos.
IV. Dike lineaments interpreted from LFC photo
not present on SIR-B images.
The histograms for these four groups (fig. 9)
indicate for SIR-B two peak concentrations: a broad
one between 290° and 340° and a narrow peak around
80°. For the LFC photos, the peak directions are
290°, 330-350° and 60°-80° (midmark classes).
Table 3. R«
directional
]
I
II 25,
III 96.
IV 124.
df : 17 ( c
* Ho: Disti
rande
To detei
between the
is used (t;
As can b<
(both fron
variance oi
We can a
ficant dif
that the d
same is tr’
interpretai
and IV, ant
the null
samples ha\
Consideri
directional
be seen (tc
320° and m
principal
and III (;
eaments), 1
observed f:
considerabl
according 1
LFC photo-i
tional clai
the observ,
than the i
hypothesis.
Apparenti
perpendicul
the overrey
to the sun
of no influ
represented
tive influ
expected, b
have little
Comparing
not present
tribution t
the directi
to the rad
50° (radar
eament has
visible on
image not i
Consideri
radar in c
radar linea
negative in
eament det<
340°, an o -
on the rad;
radar azimu
may be a r
sand cover.
CONCLUSIONS
In the Irar
cognition i
radar and f
Nearly doul