Content uploaded by Rachid Lhissou
Author content
All content in this area was uploaded by Rachid Lhissou on Oct 26, 2017
Content may be subject to copyright.
Comparison of Landsat-8, ASTER and Sentinel 1 satellite
remote sensing data in automatic lineaments extraction: A case study
of Sidi Flah-Bouskour inlier, Moroccan Anti Atlas
Zakaria Adiri
a,⇑
, Abderrazak El Harti
a
, Amine Jellouli
a
, Rachid Lhissou
a
, Lhou Maacha
b
,
Mohamed Azmi
b
, Mohamed Zouhair
b
, El Mostafa Bachaoui
a
a
Team of Remote Sensing and GIS Applied to Geosciences and Environment, Faculty of Sciences and Techniques, PO BOX 523, Beni Mellal, Morocco
b
MANAGEM Group, Twin Center, Tower A, Zerktouni Boulevards and Abdelkarim Khattabi, PO BOX 5199, Casablanca, Morocco
Received 14 December 2016; received in revised form 26 August 2017; accepted 7 September 2017
Available online 18 September 2017
Abstract
Certainly, lineament mapping occupies an important place in several studies, including geology, hydrogeology and topography etc.
With the help of remote sensing techniques, lineaments can be better identified due to strong advances in used data and methods. This
allowed exceeding the usual classical procedures and achieving more precise results. The aim of this work is the comparison of ASTER,
Landsat-8 and Sentinel 1 data sensors in automatic lineament extraction. In addition to image data, the followed approach includes the
use of the pre-existing geological map, the Digital Elevation Model (DEM) as well as the ground truth. Through a fully automatic
approach consisting of a combination of edge detection algorithm and line-linking algorithm, we have found the optimal parameters
for automatic lineament extraction in the study area. Thereafter, the comparison and the validation of the obtained results showed that
the Sentinel 1 data are more efficient in restitution of lineaments. This indicates the performance of the radar data compared to those
optical in this kind of study.
Ó2017 COSPAR. Published by Elsevier Ltd. All rights reserved.
Keywords: Automatic lineament extraction; Remote sensing; Landsat-8; ASTER; Sentinel 1; DEM
1. Introduction
Certainly, lineaments mapping plays a major role in
geological studies, especially in mining and petroleum
exploration (Marghany and Hashim, 2010). In other
words, a detailed geological study imperatively means the
knowledge of the present structural information, princi-
pally the lineaments (Pour and Hashim, 2015a). The latter
may correspond to natural objects, including a structural
alignment (Faure, 2001), geomorphologic consequences
(Corgne et al., 2010), structural weaknesses (Masoud and
Koike, 2006), faults (Hashim et al. 2013), valleys (Lacina,
1996;Hung et al., 2005), drainage areas, fractures or lines
separating the different formations (Hobbs, 1912). Linea-
ments may also represent the boundaries between the dif-
ferent lithological units or vegetation covers (Abarca,
2006; Marghany and Hashim, 2010; Saadi et al., 2011),
or artificial objects (road, bridge...). In addition, the
importance of lineaments is also manifested by their local-
ization often close to several mineralogical deposits
(Meshkani et al., 2013; Pour et al., 2016; Pour and
Hashim, 2014a,b), which qualifies them as an indirect indi-
cator of mining potential.
Today, the strong progress in the remote sensing disci-
pline allows exploiting a variety of sources and methods
http://dx.doi.org/10.1016/j.asr.2017.09.006
0273-1177/Ó2017 COSPAR. Published by Elsevier Ltd. All rights reserved.
⇑
Corresponding author.
E-mail address: zakariaadiri@gmail.com (Z. Adiri).
www.elsevier.com/locate/asr
Available online at www.sciencedirect.com
ScienceDirect
Advances in Space Research 60 (2017) 2355–2367
in the characterization of lineaments. Consequently, the
automatic approaches have become more requested than
those manuals which are difficult, time-consuming and
highly dependent on the quality of the analysis (Masoud
and Koike, 2006). Therefore, remote sensing data have
been widely used in lineaments mapping (Dubois, 1999;
Solomon and Ghebreab, 2006; Masoud and Koike, 2006;
Ranganai and Ebinger, 2008; Li, 2010; Corgne et al.,
2010; Saadi et al., 2011; Hashim et al., 2013; Meshkani
et al., 2013; Pour and Hashim, 2014b, 2015b,c; Pour
et al., 2016).
Launched in 2013, Landsat 8 is the new generation of
the series of Landsat satellites. It carries two sensors
namely OLI (Operational Land Imager) characterized by
9 spectral bands: 4 in the Visible (VIS) (0.43–0.67 mm), 1
band of the Near Infrared (NIR) (0.85–0.88 mm), 2 bands
of the Shortwave Infrared (SWIR) (1.57–2.29 mm) and 1
band of cirrus (1.36–1.38 mm) (spatial resolution of 30 m).
In addition, it has a Panchromatic band (0.50–0.68 mm)
(spatial resolution of 15 m). The second sensor is TIRS
(Thermal Infrared Sensor) with 2 bands of 100 m in spatial
resolution, operating between 10.60–11.19 mm and 11.50–
12.51 mm respectively. Landsat 8 data are characterized
by a high radiometric resolution (16 bits), and the scenes
cover 185 180 km, available at free of charges (Roy
et al., 2014).
ASTER (Advanced Spaceborne Thermal Emission and
Reflection Radiometer) is a multispectral sensor launched
in December 1999 on board the Terra platform. It records
the electromagnetic radiation reflected and emitted in 14
bands: 3 bands in the Visible Near Infrared (VNIR)
(encoded on 8 bits) with a spatial resolution of 15 m, 6
SWIR bands (8 bits) with spatial resolution of 30 m, and
5 Thermal Infrared (TIR) bands (12 bits) with a spatial res-
olution of 90 m. This sensor has an advantage of stere-
oscopy allowing the possibility of extracting Digital
Elevation Models (DEM) called ASTER GDEM (Global
Digital Elevation Model). Besides this spectral richness,
ASTER data are now available at free of charges, under
several levels of processing and with scenes of 60 60 km
in width (Abrams et al., 2015). Given these characteristics,
the ASTER sensor has been widely used in geological stud-
ies (Rowan and Mars, 2003; Moghtaderi et al., 2007; Gad
and Kusky, 2007; Di Tommaso and Rubinstein, 2007; Pour
and Hashim, 2012, 2014a,b; Pournamdary et al., 2014a,b).
On the other hand, the Radar Sentinel 1 represents the
new generation of the ESA (European Space Agency)
radars, which the main purpose is the monitoring of land
and oceans while ensuring the continuity of SAR (Syn-
thetic Aperture Radar data) (C-band) data (ESA, 2014).
Recently launched in April 2014, Sentinel 1 includes a
SAR instrument operating in band C (3.75–7.5 cm) (ESA,
2014). It allows the acquisition of data in several modes,
mainly that of Interferometric Wide Swath (IW) (main
operational mode on the Earth’s surface) (Sentinel-1 User
Handbook, 2013; ESA, 2014). In this mode, the scenes
cover 250 km in width, with very good spatial (5 20 m)
and radiometric (1 dB (Decibels)) resolutions, and an angle
of incidence between 29.1°and 46°(ESA, 2014). Each
scene can contain images with the polarizations HH
+ HV, VV + VH, HH or VV (H: Horizontal and V: Verti-
cal), distributed by the ESA at free of charges (Sentinel 1
Team, 2013; Amitrano et al., 2014; ESA, 2014). Conse-
quently, these characteristics allow qualifying Sentinel 1
as very suitable radar for several applications, particularly
geology (ESA, 2014).
In addition to satellite imagery, this study was also
based on the DEM ‘‘ASTER GDEM”. The latter is the
most available, the most updated, the most complete in
the world, at free of charges and with a high spatial resolu-
tion (30 m) (Abrams et al., 2015).
The objective of this work is to compare, for the first
time in this kind of studies, the potentialities of the
ASTER, the new OLI multispectral sensors and the new
Sentinel 1 radar sensor in automatic lineaments extraction
task. We note that the Sidi Flah-Bouskour inlier is known
by strong mining potential as well as an absence of vegetal
cover. These characteristics are very appropriate for remote
sensing applications.
2. Geology of the study area
On the geological map, the Sidi Flah-Bouskour inlier,
which encompasses the study area of this work, is located
in the large massif of Jbel Saghro, at the oriental Moroccan
Anti Atlas (Maacha et al., 2011). It consists mainly of vol-
canic and volcano-sedimentary formations belonging to
the middle and upper Neoproterozoic (Chaker, 1997;
Johann, 2005; Pique
´et al., 2006). The upper complex for-
mations are illustrated by ignimbrites, rhyolitic dykes as
well as pink granite inter-stratified with granodiorites
(Maacha et al., 2011; Rezeau et al., 2014).
Moreover, the study area is characterized by the pres-
ence of other formations including rhyolites, rhyolitic ign-
imbrites, tuffs, breccias, conglomerates belonging to the
Oum Idermil complex, in addition to the alkaline granite
of Isk n’Allah and the microgranites of the Tarhia of
Dra complex (Maacha et al., 2011; Walsh et al., 2012;
Rezeau et al., 2014). The Precambrian formations end with
sandstones and conglomerates of Tikrit (serie of lie of Vin)
and bab n’Ali which constitute the cover. Concerning the
Quaternary, it is represented by floodplain terraces, scree
as well as pebbles deposits (Fig. 1). The available and the
recent geological map used in this study is that of uˆJbel
Saghro-Dade
`sy
´(1/200,000), published by the Moroccan
Ministry of Industry, Trade, Energy and Mines. Besides
that, this area is characterized by the absence of vegetation
cover and remote sensing studies, which explain the objec-
tive and the importance of the present work (Fig. 2).
From a structural viewpoint, the study area is inter-
sected by a rhyolitic dykes system with a major direction
NE-SW. This latter also presents the major direction of
faults (Maacha et al., 2011). In addition, most of the faults
in the massif of Jbel Saghro follow this direction (Walsh
2356 Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367
et al., 2012). As long as there is no lineament map of the
study area, the appeared lineaments have been digitized
from a Google Earth image. This information was reported
in Fig. 1 as lineament (in blue color).
3. Methodology
3.1. Preprocessing
The OLI image used in this work corresponds to 04 July
2014, at the L1T (corrected terrain) level, with a Universal
Transverse Mercator (UTM) projection and a World
Geodetic System WGS 84 datum. Bands 2 and 9 are not
used in this study because the first is intended for retrieving
atmospheric aerosol properties, and the second for cirrus
cloud detection (Adiri et al., 2016). The image was con-
verted to the radiance and then corrected atmospherically
using the Dark Object Subtraction (DOS) method
(Chavez, 1988; Zhang et al., 2007; Chrysoulakis et al.,
2010; Adiri et al., 2016).
Regarding to the ASTER image, it was acquired on
September 18, 2004, at the L1B (radiance to the sensor)
level, with the same projection and the same world geodetic
system as the OLI image. First, the ‘‘Lee”filter was used to
minimize the noise effect. Afterward, all the bands are cor-
rected from atmospheric effects using the FLAASH module
Fig. 1. (a) Location of the study area at the national scale, (b) Color Composite (CC) RGB (4/3/2) extracted from the Landsat 8 OLI image and (c)
geological map of the study area. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article.)
Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367 2357
(Fast Line-of-sight Atmospheric Analysis of Spectral
Hypercubes) (Bedini, 2011; Pour and Hashim, 2012;
Adiri et al., 2016).
As long as the spatial resolution affects the ability to
identify lineaments (Alonso-Contes, 2011), the panchro-
matic band (15 m) of the OLI sensor was used to improve
the spatial resolution of the two images OLI and ASTER.
Thus, the spectral bands are resampled to 15 m using the
Pan Sharpening ‘‘Gram-Schmidt”method widely used in
remote sensing studies (Laben and Brower, 2000; Amer
et al., 2012; Maurer, 2013).
For the Sentinel 1 image, it is acquired on November 4,
2015, at L1 GRD (Ground Range Distance) level, accord-
ing to IW mode with the polarizations (VH) and (VV).
These are Digital Number (DN) data that require conver-
sion to backscatter values reflected by the surface
(Corgne et al., 2010; Amer et al., 2012). Furthermore, the
‘‘Speckle”is well known in SAR images with a ‘‘salt and
pepper”effect. This makes the interpretation of the image
content a very difficult task. To resolve this problem, the
‘‘Lee”filter was used (Corgne et al., 2010; Pour and
Hashim, 2014a,b). The last step was the correction of the
geometrical distortions attributed to the topographic vari-
ations in the scene as well as the inclination of the radar.
These distortions manifest in the shadowing (slopes charac-
terized by a very steep angle), foreshortening, and layover
problems (problems of the nearest elements which reflect
pulses before those more distant). Consequently, geometric
corrections task is especially necessary so that the image
will be represented geometrically as close as possible to
the reality, and subsequently be ready for use.
Hence, the ‘‘Range Doppler Orthorectification”method
was applied using the Shuttle radar Topography Mission
(SRTM) DEM (90 m), which is recommended for radar
images (Help SNAP). After the preprocessing steps, the
Sentinel 1 image becomes at 10 m in spatial resolution.
3.2. Processing
3.2.1. Principal component analysis
Principal Component Analysis (PCA) is a statistical
method widely used in geological studies (Zhang et al.,
2007; Gabr et al., 2010; Pour and Hashim, 2011, 2012;
Amer et al., 2012; Adiri et al., 2016). It has the advantage
of compressing the information contained in initial bands
into new bands called Principal Components (PCs) (Gabr
et al., 2010; Adiri et al., 2016). Consequently, this transfor-
mation eliminates the redundancy of data, isolates noise,
and then enhances the targeted information in the image
(Amer et al., 2012).
Several studies have been based on PCA in the detection
of lineaments. According to Li (2010), the comparison of
five different enhancement techniques (average value of
all bands, PCA, Band Ratios (BR), histogram equalization
and High Pass filter) performed by Stephen and Mynar
(1986) showed that the PCA is more efficient in the identi-
fication of lineaments. In their turn, Paganelli et al., 2003
are used the PC
2
for mapping lineaments. Then, a standard
PCA transformation was performed for the spectral
regions of VNIR and SWIR of OLI and ASTER images.
3.2.2. Lineament extraction
In the present study, the automatic lineaments extrac-
tion was carried out on the basis of two fundamental calcu-
lations: first, the use of a filter for the detection of edges
(contours). The latter give information on areas of abrupt
Fig. 2. Methodology flowchart of the present study.
2358 Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367
changes in the values of neighboring pixels, often referring
to lineaments. The second step is the detection of lines
(Corgne et al., 2010; Li, 2010; Hashim et al., 2013). In other
words, the extraction processes was based on six main
stages:
Contours detection
–RADI (filter radius)(in pixels): The radius of the filter
that will be used in contours detection. Values
between 3 and 8 are recommended to avoid introduc-
ing noise;
–GTHR (Edge Gradient Threshold): The value of the
gradient to be taken as the threshold in contours
detection (between 0 and 255). Values between 10
and 70 are acceptable;
Lines detection
–LTHR (Curve Length Threshold)(in pixels): The min-
imum length of a curve to be taken as the lineament
(a value of 10 is suitable);
–FTHR (Line Fitting Threshold)(in pixels): The toler-
ance allowed in the curve fitting (results of the previ-
ous parameter) to form a polyline. Values between 2
and 5 are recommended;
–ATHR (Angular Difference Threshold)(in degree):
Defines the angle not to be exceeded between two
polylines to be linked. Values between 3 and 20 are
suitable;
–DTHR (Linking Distance Threshold)(in pixels): The
maximum distance between two polylines to be
linked. Values between 10 and 45 are acceptable.
(Hashim et al., 2013).
First, the ‘‘Canny”filter is executed. The latter is known
for its good results in contours detection (Corgne et al.,
2010; Marghany and Hashim, 2010). This task is completed
using the radius defined in the RADI parameter. Then, the
gradient fixed in the GTHR specifies the value of the pixels
to be taken as edges and those remaining as background.
The obtained result will be a binary map (contour and
not contour). According to the pixel value defined in the
LTHR parameter, this binary map will undergo a contour
number reduction in order to leave just those referring to
curves. In other words, a curve with a number of pixels
Fig. 3. Principal components (a) (PC
5
) and (b) (PC
6
) of OLI and ASTER sensors respectively, in addition to the (d) VH and (e) VV polarizations
(Sentinel 1).
Table 1
The parameters values applied for automatic lineaments extraction.
Parameters Applied values
RADI 5
GTHR 55
LTHR 10
FTHR 2
ATHR 20
DTHR 20
Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367 2359
smaller than that indicated in LTHR will not be taken into
consideration. The resulting curves generate polylines if
they respect the tolerance defined in the FTHR. Finally,
two polylines (the simplest case) will bind to form a linea-
ment if their two end points form an angle respecting the
value specified in ATHR and the distance in DTHR
Fig. 4. Superposition of lineaments obtained from (a) OLI and (b) ASTER images, in addition to the (c) VH and (d) VV polarizations (Sentinel 1) on the
geological map of the study area.
Fig. 5. Distribution histograms showing the number of lineaments according to the length for the (a) OLI, (b) ASTER, (c) VH and (d) VV images.
2360 Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367
Fig. 6. Superposition of the lineaments resulted from the (a) OLI, (b) ASTER images in addition to the (c) VH and (d) VV polarizations (Sentinel 1) on
the PCA CC (PC
2
/PC
3
/PC
4
) of the OLI image.
Fig. 7. Superposition of the lineaments resulted from the (a) OLI, (b) ASTER images, in addition to the (c) VH and (d) VV polarizations (Sentinel 1) on
the slope map.
Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367 2361
parameters (Hashim et al., 2013). This operation was car-
ried out using the PCI Geomatica software. The applica-
tion of these parameters on the data used was carried out
while evaluating several combinations of values with
respect to the literature, the various validation criteria as
well as the ground truth, before arriving at the optimum
values which gave a satisfactory result.
3.2.3. Slope
In lineament studies, the slope appears as a very impor-
tant parameter. It is an under product easily extractable
from DEM data. In the slope map, abrupt changes in val-
ues are often key indicators of the presence of linear struc-
tures (Li, 2010; Nkono et al., 2013).
3.2.4. Shading
In addition, the analysis of the shading map extracted
from the DEM data contributes in the characterization of
lineaments. In the aim of analyzing the shaded pixels with
regard to those nearby, an oriented illumination of the
study area is done. Thereafter, the boundaries between
the shaded and unshaded areas may indicate the presence
of lineaments (Masoud and Koike, 2006; Abarca, 2006;
Li, 2010; Saadi et al., 2011). After comparing different
lighting angle, the north-south orientation (0°) was chosen
because it better exposes shaded and unshaded areas.
4. Results and discussion
For the OLI and ASTER images, the analysis of the
principal components (PCs) resulting from the calculation
of the PCA showed that the PC
O5
and PC
A6
are those
which highlight the valley (floodplain terraces in the geo-
logical map) with bright pixels. This criterion (enhance-
ment of a linear object) justifies the choice of these PCs
for the extraction of lineaments (Li, 2010), besides the
two VH and VV images of the radar Sentinel 1 (Fig. 3).
After applying the parameter values given in Table 1, the
resulting lineaments are superimposed on the geological
map relative to the study area, in order to analyze their dis-
tribution with regard to the different existing lithological
units (Fig. 4). The analysis of the results indicated that
the lineaments extracted from the OLI and ASTER sensors
are more numerous and have a high concentration com-
pared to those resulting from the Sentinel 1 images. This
can be explained by the sensitivity of the optical images
to the soils occupation such as lithological units, shadow
areas, steep slopes and vegetation cover (Mansour and
Ait Brahim, 2005; Alonso-Contes, 2011). On the contrary,
radar data are not influenced by the nature of land use and
the associated lineaments are less abundant. For the OLI
sensor, the concentration of lineaments occurs mainly in
the lithological units of granite and diorite, the basic vol-
canic serie, as well as the rhyolites. The lineaments resulted
Fig. 8. Superposition of the lineaments resulted from the (a) OLI and (b) ASTER images, in addition to the (c) VH and (d) VV polarizations (Sentinel 1)
on the shading map.
2362 Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367
from the ASTER sensor are characterized by a random dis-
tribution that occupies the entire study area. On the other
hand, the lineaments obtained from the VH and VV images
have relatively similar in term of distributions, with a supe-
riority of the VH image in the number of extracted linea-
ments (Table 1).
In terms of statistics, the comparison of the obtained
results shows that 541 and 766 lineaments were extracted
from the OLI and ASTER images respectively, whereas
the VH and VV Sentinel 1 images gave rise to 394 and
346 lineaments respectively. Fig. 5 illustrates the number
of the obtained lineaments as a function of the length (in
Meters), in addition to some statistics relating to this length
for each of the used data. Concerning those of Sentinel 1,
the values range from around 90 m to a maximum of
800 m. Regarding the OLI image, the values vary between
155 m and 2076 m, while they are between 0.2 m and
1524 m for the ASTER image. This difference can be
explained by the different natures of the images (optical
and radar). Furthermore, the most abundant lengths are
86 m (VH), 91 m (VV), 150 m (OLI) and close to 198 m
(ASTER). The comparison of these numbers indicates that
the lineaments identified by the radar data are character-
ized by smaller lengths with regard to those extracted from
the optical data (OLI and ASTER). This favors again the
effectiveness of the radar data which are independent of
the soils nature contrary to the optical images.
4.1. Accuracy assessment
4.1.1. Discontinuities
On the basis of the PCA image extracted from the OLI
sensor, the Color Composite (CC) RGB (PC
2
/PC
5
/PC
6
)
was used in order to analyze the relationship between the
resulted lineaments and the discontinuities (boundaries)
between the different lithological units (Solomon and
Ghebreab, 2006)(Fig. 6). In addition, second superposition
of these results on the slope and shading maps is carried
out for checking the correspondence with the abrupt
changes of slope and illumination areas respectively
(Masoud and Koike, 2006; Li, 2010)(Figs. 7 and 8). The
results of the analysis shows that most lineaments extracted
from the OLI image are perfectly located in the geological
boundaries (contacts) between the lithological units, con-
Fig. 9. Correlation between the density of lineaments and the location of the faults for the (a) OLI and (b) ASTER images, in addition to the (c) VH and
(d) VV polarizations (Sentinel 1).
Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367 2363
trary to those obtained from the ASTER and Sentinel 1
images which show no correlation with these areas. On
the other hand, the distribution of lineaments obtained
from the Sentinel 1 images mainly follows the areas of
abrupt changes in slope and shading, while those extracted
from the OLI sensor are also found in areas where there are
no changes in values (blue color). Concerning the linea-
ments obtained from the ASTER image, it appears clearly
that they do not respect these criteria. Consequently, these
results indicate the high sensitivity of radar data to geo-
morphology as opposed to optical data.
4.1.2. Density
In lineament studies, the density is a widely used param-
eter (Lachaine, 1999; Hung et al., 2005; Corgne et al., 2010;
Hashim et al., 2013). It informs about the concentration of
lineaments per unit area (Lachaine, 1999). In the present
work, the density was used to find the correlation between
the concentration of lineaments and the distribution of
existing faults in the study area (Fig. 9). The geological
map (Fig.9e) was used for better correlate the density dis-
tribution and the lithological units in the study area. For
the OLI sensor, the high density values are concentrated
mainly in the areas of the rhyolitic dykes, whereas the most
of the faults are located near the average density values. On
the contrary, the results obtained from the Sentinel 1
images showed a better correlation with the distribution
of the faults. Most of the latter, as well as the rhyolitic
dykes, correspond perfectly to the high density values,
especially with the VH polarization where there are more
high density values in these areas (in comparison with the
VV polarization). For the ASTER sensor, the high values
randomly occupy most of the study area. This comparison
proves that the results obtained from the radar Sentinel 1
data are the best correlated in relation to the faults. This
conclusion converges very well with the other criteria
described previously (slope and shading).
4.1.3. Orientation
In turn, the orientation allows identifying the most fre-
quent directions of a linear layer (in this case the linea-
ments). Therefore, they can be compared with directions
relating to existing faults in the study area (Lachaine,
1999; Meshkani et al., 2013; Hashim et al., 2013). This task
is completed using the rose diagram. With an angular spac-
ing of 30°and without using the length, Fig. 10 indicates
that the most dominant directions of lineaments extracted
from the OLI image are those of Southwest (SW 215°)
and South (S 180°)(Fig. 10a). Regarding the radar polar-
izations, that of VV shows the South (S 185°) and South-
west (SW 215°) as majority directions (Fig. 10d), while
that of VH (Fig. 10c) indicates the (SW 205°), (S 180°)
and (SW 230°) directions. The latter corresponds perfectly
to that of the faults. Concerning the directions indicated by
the ASTER image (Fig. 10b), they are characterized by
random directions unrelated to those of the faults or the
other results. This result was confirmed by the field survey.
Figs. 11 and 12 contain the location of some field pho-
tographs illustrating some examples of lineaments in the
study area.
Fig. 10. Orientations of lineaments obtained from the (a) OLI, (b) ASTER, (c) VH, and (d) VV images comparing to those of (e) the faults of the
study area.
2364 Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367
5. Conclusion
In this work, Landsat-8, ASTER in addition to Sentinel
1 satellite data were evaluated in automatic lineaments
extraction task. The comparison of the obtained results
while including the pre-existing geological map, lithological
units, elevation data (slope, shading), density, orientation
compared to the pre-existing faults as well as the field sur-
Fig. 11. Locations of field photographs on the geological map relating to the study area.
Fig. 12. Field photographs showing some examples of lineaments in the study area.
Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367 2365
vey showed that the Sentinel 1 give better results than the
optical sensors, especially with the VH polarization. The
lineaments obtained from Landsat-8 and ASTER data fol-
low mainly the boundaries between the lithological units,
whereas those extracted from Sentinel 1 corresponds very
well with the abrupt changes areas of slope and shading.
In addition, the Sentinel 1 lineaments converge very well
with the localization and the orientation of the existing
faults in the study area. This performance can be explained
by the sensitivity of the Radar data to geomorphology,
whereas the optical data are influenced by the shading
areas as well as by the soils occupation. Moreover, the lin-
eaments obtained from the optical data (OLI and ASTER)
reflect the local structure of the study area, while the radar
data were allowed to map regional lineaments with a good
correlation with the rhyolitic dykes as well as with the
faults. Furthermore, the obtained maps and results have
been also validated and approved by a team of experts affil-
iated to MANAGEM group, responsible of exploration
and prospecting in this study area.
Consequently, the methodology used in the present
work has shown great efficiency in automatic lineaments
extraction and can be applied in other similar regions.
The use of the HH and HV polarizations in future works
will improve, certainly, the obtained results in this study.
Acknowledgements
The authors thank the MANAGEM group for their col-
laboration as well as their help during the field survey. We
thank also the Faculty of Sciences and Techniques of Beni
Mellal and all members of the Team of Remote Sensing
and GIS Applied for Geosciences and the Environment.
In addition, we acknowledge the CNRST (National Center
for Scientific Technical Research) for his financial support.
Finally, we grateful to the U.S. Geological Survey (USGS)
for providing Landsat 8 OLI, Terra ASTER, and ASTER
GDEM data, and the European Space Agency (ESA) for
the Sentinel 1 data. The authors thank the reviewers for
their helpful comments and recommendations.
References
Abarca, M.A.A., 2006. Lineament Extraction from Digital Terrain
Models, Case Study San Antonio del sur area, south-eastern Cuba.
In: Thesis submitted to the international institute for geo-information
science and earth observation Enschede. The Nettherlands.
Abrams, M., Tsu, H., Hulley, G., Iwao, K., Pieri, D., Cudahy, T., Kargel,
J., 2015. The Advanced Spaceborne Thermal Emission and Reflection
Radiometer (ASTER) after fifteen years: review of global products.
Int. J. Appl. Earth Obs. Geoinf. 38, 292–301.
Adiri, Z., El Harti, A., Jellouli, A., Maacha, L., Bachaoui, E., 2016.
Lithological mapping using Landsat 8 OLI and Terra ASTER
multispectral data in the Bas Dra
ˆa inlier, Moroccan Anti Atlas. J.
Appl. Remote Sens. 10(1), 016005 14.
Alonso-Contes, C.A., 2011. Lineament Mapping for Groundwater
Exploration Using Remotely Sensed Imagery in a Karst Terrain: Rio
Tanama and Rio De Arecibo Basins in the Northern Karst of Puerto
Rico. Michigan Technological University (Master’s Thesis).
Amer, R., Kusky, T., El Mezayen, A., 2012. Remote sensing detection of
gold related alteration zones in Um Rus area, Central Eastern Desert
of Egypt. Adv. Space Res. 49, 121–134.
Amitrano, D.Di, Martino, G., Iodice, A., Mitidieri, F., Papa, N.M.,
Riccio, D., Ruello, G., 2014. Sentinel-1 for monitoring reservoirs: a
performance analysis. Remote Sens. 6, 10676–10693.
Bedini, E., 2011. Mineral mapping in the Kap Simpson complex, central
East Greenland, using HyMap and ASTER remote sensing data. Adv.
Space Res. 47, 60–73.
Chaker, M., 1997. Ge
´ochimie et me
´talloge
´nie de la mine d’or de Tiouit.
anti-atlas oriental. sud du maroc. In: The
`se pre
´sente
´ea
`l’universite
´du
que
´bec a
`chicoutimi.
Chavez Jr., P.S., 1988. An improved dark-object subtraction technique for
atmospheric scattering correction of multispectral data. Remote Sens.
Environ. 24, 459–479.
Chrysoulakis, N., Abrams, M., Feidas, H., Arai, K., 2010. Comparison of
atmospheric correction methods using ASTER data for the area of
Crete, Greece. Int. J. Remote Sens. 31, 6347–6385.
Corgne, S., Magagi, R., Yergeau, M., Sylla, D., 2010. An integrated
approach to hydro-geological lineament mapping of a semi-arid region
of West Africa using Radarsat-1 and GIS. Remote Sens. Environ. 114,
1863–1875.
Di Tommaso, I., Rubinstein, N., 2007. Hydrothermal alteration mapping
using ASTER data in theInfiernillo porphyry deposit, Argentina. Ore
Geol. Rev. 32, 275–290.
Dubois, J., 1999. Identification Des Line
´aments Dans Les Images
Satellitaires Par Ajustement Et Suivi Des Segments. Project applica-
tion submitted to school top technology university of Quebec, Canada.
ESA, 2014. http://www.esa.int/.
Faure, S., 2001. Analyse des line
´aments ge
´ophysiques en relation avec les
mine
´ralisations en or et me
´taux de base de l’Abitibi. Rapport, Projet
CONSOREM 2000-03A.
Gabr, S., Ghulam, A., Kusky, T., 2010. Detecting areas of high-potential
gold mineralization using ASTER data. Ore Geol. Rev. 38, 59–69.
Gad, S., Kusky, T., 2007. ASTER spectral ratioing for lithological
mapping in the Arabian-Nubian shield, the Neoproterozoic Wadi Kid
area, Sinai, Egypt. Gondwana Res. 11, 326–335.
Hashim, M., Ahmad, S., Md Johari, M.A., Pour, A.B., 2013. Automatic
lineament extraction in a heavily vegetated region using Landsat
Enhanced Thematic Mapper (ETM+) imagery. Adv. Space Res. 51,
874–890.
Help Snap (Sentinel 1 toolbox).
Hobbs, W.H., 1912. Earth Features and Their Meaning, an Introduction
to Geology for the Student and General Reader. The Macmillan, New
York.
Hung, L.Q., Batelaan, O., De Smedt F., 2005. Lineament Extraction and
Analysis, Comparison of LANDSAT ETM and ASTER Imagery.
Case study: Suoimuoi Tropical Karst Catchment, Vietnam. Remote
Sensing for Environmental Monitoring, GIS Applications, and Geol-
ogy V 5983, 59830T.
Johann, T., 2005. Processus de formation et relations spatio-temporelles
des mine
´ralisations a
`or et argent en contexte volcanique Pre
´cambrien
(Jbel Saghro, Anti-Atlas, Maroc). Implications sur les relations
de
´formation-magmatismevolcanisme-hydrothermalisme. In: The
`se
pre
´sente
´e dans l’universite
´d’Orle
´ans.
Laben, C.A., Brower, B.V., 2000. Process for Enhancing the Spatial
Resolution of Multispectral Imagery using Pan-Sharpening. Patent
Number: 6,011,875, Date of Patent: Jan. 4.
Lachaine, G., 1999. Structures ge
´ologiques et line
´aments, Beauce
(Que
´bec}: apport de la te
´le
´de
´tection. In: Memory Presented in Faculty
of Humanities. University of Sherbrooke Que
´bec, Canada.
Lacina, C., 1996. Interpre
´tation structurale des line
´aments par traitement
d’image satellitaire: cas des sousprovinces d’abitibi et d’opatica
(Que
´bec). In: Memory Presented in Faculty of Humanities. University
of Sherbrooke Que
´bec, Canada.
Li, N., 2010. Textural and rule-based lithological classification of remote
sensing Data, and geological mapping in Southwestern Prieska Sub-
basin, Transvaal Supergroup, South Africa. In: Thesis presented in at
2366 Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367
the Faculty of Earth Sciences the Ludwig Maximilian University of
Munich, Germany.
Maacha, L., Ouadjou, A., Azmi, M., Zouhair, M., Saquaque, A.,
Alansari, A., Soulaimani, A.La., 2011. Mine de Cuivre et argent de
Bouskour (J. Saghro, Anti-Atlas oriental). In: New Geological and
Mining Guidebooks of Morocco. Notes et memoires du service
ge
´ologique N 564, vol. 9. Ministry of Energy, Mines, Water and
Environment, Morocco, pp. 59–64.
Mansour, M., Ait Brahim, L., 2005. Apport de la Te
´le
´de
´tection radar et
du MNT a l’analyse de la fracturation et la dynamique des versants
dans la Re
´gion de Bab-taza, Rif, Maroc. Teledetection, 5(1-2-3), 95–
103.
Marghany, M., Hashim, M., 2010. Lineament mapping using multispec-
tral remote sensing satellite data. Int. J. Phys. Sci. 5, 1501–1507.
Masoud, A., Koike, K., 2006. Tectonic architecture through Landsat-7
ETM+/SRTM DEM-derived lineaments and relationship to the
hydrogeologic setting in Siwa region, NW Egypt. J. Afr. Earth Sci.
45, 467–477.
Maurer, T., 2013. How to pan-sharpen images using the gram-schmidt
pan-sharpen method – a recipe. In: International Archives of the
Photogrammetry, Remote Sensing and Spatial Information Sciences,
Volume XL-1/W1, ISPRS Hannover Workshop 2013, 21 – 24 May
2013, Hannover, Germany.
Meshkani, S.A., Mehrabi, B., Yaghubpur, A., Sadeghi, M., 2013.
Recognition of the regional lineaments of Iran: using geospatial data
and their implications for exploration of metallic ore deposits. Ore
Geol. Rev. 55, 48–63.
Moghtaderi, A., Moore, F., Mohammadzadeh, A., 2007. The application
of advanced space-borne thermal emission and reflection (ASTER)
radiometer data in the detection of alteration in the Chadormalu
paleocrater, Bafq region, Central Iran. J. Asian Earth Sci. 30, 238–252.
Nkono, C., Fe
´me
´nias, A., Lesne, A., Mercier, J.C., Demaiffe, D., 2013.
Fractal analysis of lineaments in equatorial africa: insights on
lithospheric structure. Open J. Geol. 3, 157–168.
Paganelli, F., Grunsky, E., Richards, J., Pryde, R., 2003. Use of radarsat-1
principal component imagery for structural mapping: a case study in
the buffalo head hills area, northern 11 central Alberta, Canada. Can.
J. Remote Sens. 29 (1), 111–140.
Pique
´, A., Soulaimani, A., Laville, E., Amrhar, M., Bouabdelli, M.,
Hoepffner, C., Chalouan, A., 2006. Ge
´ologie du Maroc.
Pour, A.B., Hashim, M., 2011. Identification of hydrothermal alteration
minerals for exploring of porphyry copper deposit using ASTER data,
SE Iran. J. Asian Earth Sci. 42, 1309–1323.
Pour, A.B., Hashim, M., 2012. Identifying areas of high economic-
potential copper mineralization using ASTER data in the Urumieh-
Dokhtar Volcanic Belt, Iran. Adv. Sp. Res. 49, 753–769.
Pour, A.B., Hashim, M., 2014a. ASTER, ALI and Hyperion sensors data
for lithological mapping and ore minerals exploration. SpringerPlus 3
(130), 2014a.
Pour, B.A., Hashim, M., 2014b. Structural geology mapping using
PALSAR data in the Bau gold mining district, Sarawak, Malaysia.
Adv. Sp. Res. 54 (4), 644–654.
Pour, A.B., Hashim, M., 2015a. Geolgical structure mapping of the
bentong-raub suture zone, peninsular malaysia using palsar remote
sensing data. ISPRS Annals of the Photogrammetry, Remote Sens.
Spat. Inf. Sci. II-2/W2, 89–92, http://dx.doi.org/10.5194/isprsannals-
II-2-W2-89-2015.
Pour, B.A., Hashim, M., 2015b. Structural mapping using PALSAR data
in the central Gold belt, Peninsular Malaysia. Ore Geol. Rev. 64, 13–
22.
Pour, B.A., Hashim, M., 2015c. Integrating PALSAR and ASTER data
for mineral deposits exploration in tropical environments: a case study
from Central Belt, Peninsular Malaysia. Int. J. Image Data Fusion 6
(2), 170–188.
Pour, B.A., Hashim, M., Makoundi, C., Zaw, K., 2016. Structural
mapping of the Bentong-Raub suture zone using PALSAR remote
sensing data, Peninsular Malaysia: implications for sediment-hosted/
orogenic gold mineral systems exploration. Resour. Geol. 66 (4), 368–
385.
Pournamdary, M., Hashim, M., Pour, B.A., 2014a. Application of
ASTER and Landsat TM data for geological mapping of Esfandagheh
ophiolite complex, Southern Iran. Resour. Geol. 64 (3), 233–246.
Pournamdary, M., Hashim, M., Pour, B.A., 2014b. Spectral transforma-
tion of ASTER and Landsat TM bands for lithological mapping of
Soghan ophiolite complex, south Iran. Adv. Space Res. 54 (4), 694–
709.
Ranganai, R.T., Ebinger, C.J., 2008. Aeromagnetic and Landsat TM
structural interpretation for identifying regional groundwater explo-
ration targets, south-central Zimbabwe Craton. J. Appl. Geophys. 65,
73–83.
Rezeau, H., Chelle-Michou, C., Calder, M., 2014. Geology and Metal-
logeny of the Atlas Mountains, Morocco Excursion guide.
Rowan, L.C., Mars, J.C., 2003. Lithologic mapping in the Mountain Pass,
California area using Advanced Spaceborne Thermal Emission and
Reflection Radiometer (ASTER) data. Remote Sens. Environ. 84,
350–366.
Roy, D.P., Wulder, M.A., Loveland, T.R., Woodcock, C.E., Allen, R.G.,
Anderson, M.C., Helder, D., Irons, J.R., Johnson, D.M., Kennedy,
R., Scambos, T.A., Schaaf, C.B., Schott, J.R., Sheng, Y., Vermote, E.
F., Belwardo, A.S., Bindschadler, R., Cohenq, W.B., Gao, F., Hipple,
J.D., Hostert, P., Huntington, J., Justice, C.O., Kilic, A., Kovalskyy,
V., Lee, Z.P., Lymburner, L., Masek, J.G., McCorkel, J., Shuai, Y.,
Trezza, R., Vogelmannc, J., Wynne, R.H., Zhud, Z., 2014. Landsat-8:
science and product vision for terrestrial global change research.
Remote Sens. Environ. 145, 154–172.
Saadi, N.M., Abdel Zaher, M., El-Baz, F., Watanabe, K., 2011. Integrated
remote sensing data utilization for investigating structural and tectonic
history of the Ghadames Basin, Libya. Int. J. Appl. Earth Obs. Geoinf.
13, 778–791.
Sentinel-1 User Handbook, 2013. Agence Spatiale Europe
´enne (ESA),
Re
´fe
´rence GMES-S1OP-EOPG-TN-13-0001.
Solomon, S., Ghebreab, W., 2006. Lineament characterization and their
tectonic significance using Landsat TM data and field studies in the
central highlands of Eritrea. J. Afr. Earth Sci. 46, 371–378.
Stephen, J.W., Mynar, F., 1986. Landsat digital enhancements for
lineament detection. Environ. Geol. Water Sci. 8 (3), 123–128.
Walsh, G.J., Benzianeb, F., Aleinikoff, J.N., Harrison, R.W., Yazidi, A.,
Burton, W.C., Quick, J.E., Saadane, A., 2012. Neoproterozoic tectonic
evolution of the Jebel Saghro and Bou Azzer—El Graara inliers,
eastern and central anti-atlas, Morocco. Precambrian Res. 216–219,
23–62.
Zhang, X., Pazner, M., Duke, N., 2007. Lithologic and mineral informa-
tion extraction for gold exploration using ASTER data in the south
Chocolate Mountains (California). ISPRS J. Photogrammetr. Remote
Sens. 62, 271–282.
Z. Adiri et al. / Advances in Space Research 60 (2017) 2355–2367 2367
