ere rar
I
|
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004
interest of the area is focused on lens-shaped pyrrhotite
stratabound massive deposits (iron caps), which appear on the
surface as rust-colored oxidized minerals caused by the
oxidation of iron sulphides.
4.000
Metres Legend
N ert Faults Basic intrusions
! ge [] cuatemary alluvions EB Acidic intrusions
=) EA Serie of Bou Gader a Pyrrhotite Stratabounds
£ [| schist of Sarhlef
n Maroc [777] Jebilet Occidentales
— K* Kettera G — J.Ahmar
H
A
B — S' Mahjoub — K*' Dalaa
C - K" Ahril I - K* Mirouga
D-GouresSefra |. J— K* Aicha
E — J. Hadit K — K* Hamra
F — Menaa el Kahla L — K? Bouzelfar
Figure 1 — Simplified geological map of Central Jebilet
(Huvelin, 1970)
2.2. Data
Two images were used in this study. First of all, the Landsat
TM image was acquired on June 26" 1987 and the Quickbird
image on February 26^ 2003. Both images were acquired over
the same area and at the same solar time, thus have similar
illumination characteristics. They also show low vegetation
distribution, which is good for geological and lithological
mapping.
2.3. Pre-processing
Various factors affect the signal measured at the sensor, such as
drift of the sensor radiometric calibration, atmospheric and
topographical effects. For accurate analysis, all of these
corrections are necessary for remote sensing imagery. The
images were corrected from atmospheric effects using the
updated Herman transfer radiative code (H5S) adapted by
Teillet and Santer (1991). H5S simulates the signal received at
the top of the atmosphere from a surface reflecting solar and
sky irradiance at sea level while considering terrain elevation
600
and the sensor altitude (Teillet and Santer 1991). Radiometric
corrections on TM were done with the TM post-calibration
dynamics ranges and the solar exoatmospheric spectral
irradiance values (Clark, 1986). The image distributors did
radiometric corrections on the Quickbird image. Topographical
effects were corrected on both images (orthorectification) using
a digital elevation model (DEM) derived from a topographic
map using a digitizing table and a geographic information
system (GIS) tool. A 30 x 30 meter grid was created for TM
and a 2.5 x 2.5 meter grid for Quickbird.
2.4 Spectral Angle Mapper algorithm (SAM)
The SAM is a classification method that permits rapid
mapping by calculating the spectral similarity between the
image spectrums to reference reflectance spectra (Yuhas ef al.,
1992: Kruse et al., 1995; Van der Meer er al., 1997; Crosta et
al, 1998; De Carvalho and Meneses, 2000; Schwarz and
Staenz, 2001; Hunter and Power, 2002). The reference spectra
can either be taken from laboratory or field measurements or
extracted directly from the image. SAM measures the spectral
similarity by calculating the angle between the two spectra,
treating them as vectors in n-dimensional space (Kruse et a/.,
1993; Van der Meer et al, 1997; Rowan and Mars., 2003).
Small angles between the two spectrums indicate high
similarity and high angles indicate low similarity.
This method is not affected by solar illumination factors,
because the angle between the two vectors is independent of the
vectors length (Crosta et al., 1998; Kruse et al., 1993). It takes
the arccosine of the dot product between the test spectrums "1"
to a reference spectrum "7" with the following equation (Yuhas
et al., 1992; Kruze et al., 1993; Van der Meer et al, 1997; De
Carvalho and Meneses, 2000):
Su (1)
i=l
nb S bh nb ?
Se
i=l
Where nb = the number of bands
t; = test spectrum
r; = reference spectrum
The main advantages of the SAM algorithm are that its an easy
and rapid method for mapping the spectral similarity of image
spectra to reference spectra. It is also a very powerful
classification method because it represses the influence of
shading effects to accentuate the target reflectance
characteristics (De Carvalho and Meneses, 2000). The main
disadvantage of this method is the spectral mixture problem.
The most erroneous assumption made with SAM is the
supposition that endmembers chosen to classify the image
represent the pure spectra of a reference material. This problem
generally occurs with medium spatial resolution images, such
as Landsat TM. The surface of the Earth is complex and
heterogeneous in many ways, thus having mixed pixels is
incontestable. The spectral confusion in pixels can lead to
underestimation and overestimation errors for a spectral class.
In general, the spectral mixture problem should decrease with
higher resolution images like Quickbird. But in some cases, it
can also increase the mixture problem because more local
Intern
variat
appar
(Gebt
2.5 F1
Image
select:
combi
transf
since
compx
correl
exami
spectr
shown
acidic
percer
spectr:
spectr
As ill
extrac
magm
deposi
endme
thus c
carbon
was ot
(figure
carbon
with s
results
