Reappraisal of DEMs, Radar and optical datasets in lineaments extraction with emphasis on the spatial context

Abstract

Lineaments mapping is an authentic issue for deciphering the tectonic setting, geological history, mineral prospecting, and other several applications. Consequently, the study objective is to examine various remote sensing datasets at wide various spatial resolutions (10, 12.5, 15, 20, 30 m), to recommend the best in lineaments elicitation, for usage in the geological scientific community. Toward this aim, nine various remote sensing datasets including optical sensors (Landsat OLI, ASTER, Earth Observing-1 Advanced Land Imager, Sentinel 2A), radar data (Sentinel 1), and digital elevation models (ALOS PALSAR, SRTM, NASA, ASTER V3), are tackled. In this scope, we created an entirely automatic lineament derivation environment through the integration of edge detection and line-linking algorithms. Results show that the used optical sensors are less efficient than DEMs having the same spatial resolution. Sentinel 1 radar data is more competent than optical data sources. ALOS PALSAR DEM (12.5m) is more eligible than any other utilized data type even sentinel 1 data (10 m). Wholly, DEMs built from radar data (e.g., PALSAR DEM) proved their leverage in lineament extraction to a limit that can deviate from the well-known relationship between the number of extracted lineaments and pixel size.

Full text

Remote Sensing Applications: Society and Environment 24 (2021) 100617
Available online 22 August 2021
2352-9385/© 2021 Published by Elsevier B.V.
Reappraisal of DEMs, Radar and optical datasets in lineaments extraction
with emphasis on the spatial context
Ali Shebl
a
,
b
,
*
, ´
Arp´
ad Cs´
amer
a
a
Department of Mineralogy and Geology, University of Debrecen, Egyetem t´
er 1, 4032, Debrecen, Hungary
b
Department of Geology, Tanta University, Tanta, 31527, Egypt
ARTICLE INFO
Keywords:
Lineament extraction
ALOS PALSAR DEM
Sentinel 1
DEMs
Geological structures
Landsat OLI
ASTER
Sentinel 2A
ABSTRACT
Lineaments mapping is an authentic issue for deciphering the tectonic setting, geological history, mineral pro-
specting, and other several applications. Consequently, the study objective is to examine various remote sensing
datasets at wide various spatial resolutions (10, 12.5, 15, 20, 30 m), to recommend the best in lineaments
elicitation, for usage in the geological scientic community. Toward this aim, nine various remote sensing
datasets including optical sensors (Landsat OLI, ASTER, Earth Observing-1 Advanced Land Imager, Sentinel 2A),
radar data (Sentinel 1), and digital elevation models (ALOS PALSAR, SRTM, NASA, ASTER V3), are tackled. In
this scope, we created an entirely automatic lineament derivation environment through the integration of edge
detection and line-linking algorithms. Results show that the used optical sensors are less efcient than DEMs
having the same spatial resolution. Sentinel 1 radar data is more competent than optical data sources. ALOS
PALSAR DEM (12.5m) is more eligible than any other utilized data type even sentinel 1 data (10 m). Wholly,
DEMs built from radar data (e.g., PALSAR DEM) proved their leverage in lineament extraction to a limit that can
deviate from the well-known relationship between the number of extracted lineaments and pixel size.
1. Introduction
Lineaments are considered essential features for most surcial
studies and represent the key for solving various issues in several dis-
ciplines (Pirasteh et al., 2013). Geologically, linear tectonic features
which may be faults, dykes, and shear zones are considered yardsticks in
several surface and subsurface studies, such as, ore deposits and mineral
prospecting ( Abd El-Wahed et al., 2021; Manuel et al., 2017; Pour et al.,
2016; Pour and Hashim, 2015a; Shebl et al., 2021), hydrogeology (Adiri
et al., 2017; Akinluyi et al., 2018; Bhuiyan, 2015; Dasho et al., 2017;
Hashim et al., 2013; Koike and Ichikawa, 2006; Qari, 2011; Takorabt
et al., 2018), petroleum exploration (Marghany and Hashim, 2010),
engineering constructions (Adhab and Hassan, 2014; Rahnama and
Gloaguen, 2014; Sukumar et al., 2014), landslides (Alizadeh et al., 2018;
Ashournejad et al., 2019), placement of conceivable geothermal reser-
voirs (Saepuloh et al., 2018), and groundwater exploration (Bruning
et al., 2011; Ibrahim and Mutua, 2014; Magaia et al., 2018). The iden-
tication of lineaments in an automatic way is more efcient and much
faster than the manual (visual) process, which is inuenced by subjec-
tive parameters like quality analysis and experience (Muhammad and
Awdal, 2012). Recently, the availability of a various and wide range of
remote sensing datasets and their potency to supply steady obvious data
across large areas compared to ground-based assessments, offer an
easier approach compared to the manual methods for lineaments
extraction. Consequently, the latest two decades have witnessed
considerable advances in scientic research toward assessing linear
features, using several techniques and in various applications (Hashim
et al., 2013; Meshkani et al., 2013; Nath et al, 2017, 2019; Pour et al.,
2016; Pour and Hashim, 2014, 2015b; Singh and Garg, 2013). The
automatic methods have resulted in a more efcient lineament extrac-
tion process (Masoud and Koike, 2006, 2011; Tripathi et al., 2000). A
lineament extraction process comprises two main steps, namely edge
detection, and line-linking or line extraction, utilizing digital data like
satellite images, determining algorithms, and certain software like the
frequently used LINE module of PCI Geomatica.
The main objective of this study is to compare and examine the po-
tentialities of the most currently used data types of the three main cat-
egories; optical, radar, and DEMs considering various spatial resolutions
(10, 12.5, 15, 20, 30 m) to build a suitable recommendation to be famed
for automatic lineaments extraction process. To achieve this aim, an
enormous amount of data was obtained, preprocessed, and processed, in
a GIS environment. More than 100 lineament maps, from different
* Corresponding author. Department of Mineralogy and Geology, University of Debrecen, Egyetem t´
er 1, 4032, Debrecen, Hungary.
E-mail address: ali.shebl@science.tanta.edu.eg (A. Shebl).
Contents lists available at ScienceDirect
Remote Sensing Applications: Society and Environment
journal homepage: www.elsevier.com/locate/rsase
https://doi.org/10.1016/j.rsase.2021.100617
Received 19 May 2021; Received in revised form 19 August 2021; Accepted 20 August 2021

Remote Sensing Applications: Society and Environment 24 (2021) 100617
2
sources, were established and compared to be efciently judged, under
unifying most of the parameters and making the only variable is the data
type. The data used in this study include multispectral data (Landsat 8
OLI, ASTER, EO1 ALI, Sentinel 2A), radar data (Sentinel 1), and DEMS
(ALOS PALSAR, SRTM, NASA, ASTER V3).
2. Data and methods
2.1. Study area
Um Salatit – Mueilha area is located at the extreme southern part of
the Central Eastern Desert (CED), Egypt. It extends between latitudes
24◦49” to 25◦18” N and longitudes 33◦50” to 34◦05” E covering an
area of about 1400 km
2
. The investigated area was selected for this
Fig. 1. Geological map of the study area showing its lithological units, after (Zoheir et al., 2019a).
Table 1
Characteristics of the utilized optical datasets.
Landsat 8 ASTER Sentinel 2 EO1 ALI
B.n C.W. (
μ
m) S.R (m) B.n C.W. (
μ
m) S.R (m) B.n C.W. (
μ
m) S.R (m) B.n C.W. (
μ
m) S.R (m)
1 0.442 30 1 0.560 15 1 0.443 60 Pan. 0.585 10
2 0.483 30 2 0.660 15 2 0.490 10 1 0.443 30
3 0.561 30 3N 0.820 15 3 0.560 10 2 0.482 30
4 0.654 30 3B 0.820 15 4 0.665 10 3 0.565 30
5 0.864 30 4 1.650 30 5 0.704 20 4 0.660 30
6 1.609 30 5 2.165 30 6 0.740 20 5 0.790 30
7 2.203 30 6 2.205 30 7 0.782 20 6 0.867 30
8 0.598 15 7 2.260 30 8 0.842 10 7 1.250 30
9 1.373 30 8 2.330 30 8a 0.865 20 8 1.650 30
10 10.90 100 9 2.395 30 9 0.945 60 9 2.215 30
11 12.00 100 10 1.375 60
11 1.610 20
12 2.190 20
Band number (B.n), Central wavelength (C.W), Spatial Resolution (S.R) and Pan for panchromatic.
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
3
work, as it is famous for old mining activities, dykes, faults, hydro-
thermal alteration zones, which manifest the role of lineaments in
controlling mineralization. Furthermore, it has a recently published
geological map, which is useful in comparing and verifying the results.
Geologically, the considered area is covered mainly by a widely
distributed stretch of Neoproterozoic ophiolitic m´
elange constituted
mainly of allochthonous ophiolitic fragments blended in a sheared ma-
trix, as well as, other different mappable units (Zoheir et al., 2019a).
Ophiolitic m´
elange is considered the most extensive unit of the area. The
other mappable units include metavolcanics, metagabbro-diorites, and
granitic rocks as shown in Fig. 1.
2.2. Data characteristics and preprocessing
A Landsat-8 (L8) scene, with an ID of LC81740432019298LGN00,
and acquired on October 25, 2019 with a path 147 and row 43, covers
the whole study area. Also, the investigated area is entirely covered by
Advanced Spaceborne Thermal Emission and Reection Radiometer
(ASTER) scene named AST_L1A_00303062007083043, and acquired on
March 6, 2007. Earth Observing-1 Advanced Land Imager (ALI) data
used in this study (EO1A1740422003070110PZ) was acquired in 2003
and imaged by ALI sensor. L8, ASTER, and ALI are attained through the
U.S. Geological Survey Earth Resources Observation and Science Center
(EROS) (https://earthexplorer.usgs.gov/) and their characteristics are
described in Table 1. Sentinel 2 (S2) data Utilized in the current study is
obtained through the European Space Agency (ESA). A cloud-free, S2A
MSI scene (S2A_MSIL1C_20200505T081611_N0209_R121_T36RWN
_20200505T095132) was acquired on May 5, 2020, and its character-
istics are summarized in Table 1. The utilized multispectral optical data
of Landsat-8 level 1T, ASTER (AST_L1A), and ALI are geometrically
corrected according to UTM, WGS 84 zone 36 N. Consequently, running
Fast Line-of-Sight Atmospheric Analysis of Spectral Hypercubes
(FLAASH) atmospheric correction, resizing the data to the borders of the
study area. Sentinel-2 Dataset bands were geo-referenced to the zone 36
North UTM projection using the WGS-84 datum then, radiometrically-
corrected using sen2cor processor in Sentinel Application Platform
(SNAP). Time of acquisition is the main controller of the shadows within
optical images. Thus, Unifying the time of imaging was desired during
the data selection phase to remove any preferability among the datasets
as differences in the time result in different shades which directly affect
the detected lineaments. Time of scene acquisition was 07:59:31,
08:03:44, 08:13:29, and 8:16:11 for EO1 ALI, ASTER, Landsat OLI, and
Sentinel 2. We also tried to preserve sun angles within the same range as
much as we can (e.g., 50.38, 52.7, 48.94 for EO1 ALI, ASTER, Landsat
OLI respectively). On the other hand, the current research data considers
timings of seasonal events, which may affect the quality of optical im-
ages and the evaluation feasibility by selecting all the optical data in
months of steady climatic conditions. After revising the timings of major
ash ood events within the study area and its environs, it was found
that most of these events are evident between November and February.
Consequently, All the optical scenes are acquired beyond this time
range.
Shuttle Radar Topography Mission (SRTM) was own aboard the
space shuttle Endeavour on 11–22 February 2000. For this study, SRTM
DEM was obtained from the USGS Earth Explorer web-based data
(https://earthexplorer.usgs.gov) to extract lineaments within the study
area. NASADEM product (30m) is a state-of-the-art global digital
elevation model (DEM) derived from a combination of SRTM processing
improvements, elevation control, void-lling, and merging with data
unavailable at the time of the original SRTM production. NASA DEM is
distributed in 1◦by 1◦tile and consist of all land between 60◦N and 56◦
S latitudes. For this study, NASA DEM was obtained from the USGS web-
based data (https://lpdaac.usgs.gov/). Advanced Spaceborne Thermal
Emission and Reection Radiometer–Global Digital Elevation Model
(ASTER GDEM) has 1 arc-second (30 m) spatial resolution, and this
project is done by the Ministry of Trade, Economy and Industry (METI)
of Japan and the United States National Aeronautics and Space
Administration (NASA) to provide high-resolution DEM to the public.
In this study, four scenes (ASTGTMV003_N24E033, ASTGTMV
003_N24E034, ASTGTMV003_N25E033, and ASTGTMV003_N25E034)
were obtained, mosaicked to one raster, and resized to the area outline.
The Advanced Land Observing Satellite (ALOS) was launched on
January 24, 2005. ALOS has three remote-sensing instruments: Panc
hromatic Remote-sensing Instrument for Stereo Mapping (PRISM) for
digital elevation mapping, the Advanced Visible and Near Infrared
Radiometer type 2 (AVNIR-2) for precise land coverage observation, and
the Phased Array L-type band Synthetic Aperture Radar (PALSAR) for
day-and-night and all-weather land observation. From Alaska Satellite
Facility (https://asf.alaska.edu/), ALOS-PALSAR FBS (Fine Beam Single
polarization mode, HH) RT1 (Radiometric Terrain Corrected with Pixel
spacing is 12.5 m) DEM is used in this work to extract the linear features
of the study area. PALSAR sensor is an active microwave sensor (help to
avoid weather barrier conditions and day or night effect), L-band (1.27
GHz) synthetic aperture radar aid at achieving high-resolution DEM
products. All DEMS are geometrically corrected and prepared to obtain
16 shaded reliefs with different azimuths (0◦, 45◦, 90◦, 135◦, 180◦, 225◦,
270◦, and 315◦), and altitudes (at 30◦and 45◦) using spatial analyst tools
in ArcMap.
Sentinel-1 is the premier satellite of the Copernicus Programme
satellite constellation conducted by ESA. The rst satellite, Sentinel-1A,
launched on April 3, 2014, and Sentinel-1B on April 25, 2016. They
carry a C-band synthetic-aperture radar (SAR) instrument, which pro-
vides a collection of data in all-weather, day or night. In Sentinel 1,
scattered energy recorded as radar data is considered one of the most
powerful techniques in extracting lineaments. Radar data can be trans-
formed into bands depending on the signals transmitted to, and received
back from, land surfaces (e.g., Horizontal send and Horizontal receive:
HH, and Horizontal send and Vertical receive: HV), and the intensity, as
well as polarization, can provide insights into the scattering mechanisms
and, hence, the physical structure of scattering elements (e.g., linea-
ments). Furthermore, techniques, such as interferometric synthetic
aperture radar (InSAR), use differential phases of reected signals to
detect land surface changes and can be used to map various land
properties (Engdahl and Hyyppa, 2003; Hong and Wdowinski, 2014).
Higher look angles mostly manifest topographic attributes in at terrain
but exaggerate shadowing in areas of high reliefs. Contrarily, smaller
incidence angels severely distort higher reliefs (Richards and others,
2009). Thus, a mid-angle (33◦), standard Dual Polarization (SDV),
Ground Range Multi-Look Detected (GRD), High Resolution (HR)
product, with approximately square pixel spacing product, was found to
be suitable with the varied topography and different lineaments repre-
sentations within the study area (e.g., controlled drainage lines, gentle
and steep tectonic fractures, dykes with various orientations and dips,
rock boundaries with varied attitudes and elevations). Moreover, these
products almost deliver images with reduced speckle. Sentinel-1A of
Interferometric Wide (IW) mode and pixel-spacing 10 m ×10m, ob-
tained from ESA to fulll the aim of this study. The Granule ID:
S1A_IW_GRDH_1SDV_20200926T154707_20200926T154736_034532
_0404DC_2A0B, acquired on September 26, 2020. Sentinel-1 data was
preprocessed by applying several steps including the precise orbit of
acquisition, removing thermal and image border noise, radiometric
calibration, and range-doppler and terrain correction using SNAP
toolbox.
2.3. Processing (Lineament extraction)
2.3.1. Automatic extraction
The major preference of automatic technique is its vantage of line-
aments detection that could not be perceived and observed by naked
eyes (Adhab and Hassan, 2014) as well as, saving time and effort
compared to visual extraction. Lineament detection utilizing computer
algorithms relies basically on two concepts, the rst is edge detection
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
4
and the second is line extraction. The most widely used, LINE module of
the PCI Geomatica software is employed for this study. To accomplish
this task, several parameters are required to determine the edges and,
curves taken into consideration to ensure a reasonable level of lineament
extraction. These parameters are RADI (lter radius), GTHR (Edge
Gradient Threshold), LTHR (Curve Length Threshold), FTHR (Line
Fitting Threshold), ATHR (Angular Difference Threshold), and DTHR
(Linking Distance Threshold). Table 2, denes the importance of each
parameter in the process of extraction. Depending on the previous
studies, literature, default values, and experimenting with more than 20
permutations of the previously mentioned parameters over different
data types, the applied values in this study are determined and tabulated
in Table 2. These assigned parameters are kept constant in all the stages
and during all phases of automatic extraction of lines to ensure equitable
comparison of all data. For RADI, this study implemented Canny lter
due to its ability to detect contours as proved in previous studies (Corgne
et al., 2010; Marghany and Hashim, 2010). GTHR uses the RADI product
image to obtain a binary image depending on the discrimination of edge
values from the remaining background using a certain value. The latter
binary product is converted to curves depending on LTHR value, only
curves with a higher value than LTHR are considered and extracted as
lines. The lines coinciding with the FTHR value are, in turn, converted to
polylines. The last step is polyline transformation to lineaments, this is
only evident depending on the angular value (ATHR) and distance
(DTHR) between endpoints of a line. The input data for the automatic
extraction process are more than 100 images from different sources and
by different spatial resolutions, including 41 images from optical data,
64 images from DEMs, and 2 images from Sentinel 1 radar data. The PCI
line module is executed over all these inputs separately and by utilizing
the same previously mentioned parameters to get more than 100 line-
ament maps.
2.3.2. Manual extraction
This visual method involves three main steps: image processing, vi-
sual interpretation, and manual digitization of the lineaments (Ibrahim
and Mutua, 2014). To ensure an efcient manual delineation of linea-
ments, image enhancement techniques (e.g., optimum index factor) are
calculated for the data, to enhance all types of lineaments e.g., geolog-
ical structures (faults, joints, dykes), geomorphological (cliffs, terraces,
topographic alignments, and controlled stream segments), tonal contrast
(various rock compositions), and even anthropogenic activities and ef-
fects (roads). Also, the PCA data transformation technique was
Table 2
The assigned values and tasks for the LINE module parameters.
Mission Value
Edge
detection
RADI To detect contours efciently 10
GTHR To determine the pixels value considered as
edges
50
Curve
extraction
LTHR To omit shorter curves having number of
pixels lower than LTHR value
30
FTHR To generate polylines tting FTHR value 3
ATHR Linking polylines with end point angle
coinciding with ATHR
15
DTHR Joining polylines whose endpoints respect
DTHR value
20
Fig. 2. Manual digitization of lineaments from 15m resolution data: (a) RGB 752 pan-sharpened HSV L8 and, (b) PC1 of L8 to produce, (c) Manually-
derived lineaments.
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
5
established to be compared with the other composites. Pan sharpened
15m L8 RGB 752 composite, PC1 of L8, and ASTER VNIR bands were
used collectively to visually interpret and manually delineate lineaments
for the study area as shown in Fig. 2.
3. Results
Ample ndings were obtained (e.g. The best VNIR and SWIR bands
for each type of optical data, and the overall best data type for each
resolution category) as shown in Tables 3 and 4, and Fig. 3. For single-
band analysis of L8, the superiority of VNIR b8 is evident over all other
bands by extracting 2951 lines and this undoubtedly is interpreted by its
higher spatial resolution (15m) compared to the other bands (30m). For
the other bands (1–7), b4 proved its efciency by getting 789 lines
compared to the others. However, PC1 of the seven pan-sharpened
stacked bands, with the higher resolution b8 gives the maximum num-
ber of lines (3108) for L8. This is attributed to gathering the advantages
of ner resolution and a higher amount of transformed data in the uti-
lized PC1 (15m). Hence, data transformation and using the rst prin-
cipal component is recommended for lineament extraction from Landsat
OLI data. For ASTER data, single VNIR band 3 gives 2861 lines
compared to PC1 (15m) obtained 2756 lines and this could be inter-
preted by higher data variance in VNIR band 3. This could be explained
by the data exuberance in VNIR bands, and this is conrmed by the
previous studies (e.g. (El-Magd et al., 2015; Hung et al., 2005; Shebl
et al., 2021)) which always recommend VNIR b3 of ASTER as one of the
most efcient sources to extract lineaments. Concerning EO-1 ALI and
by a number exceeding all the results of ASTER and L8, the panchro-
matic band (10m) of this sensor extracts 5795 lines manifesting the
spatial resolution effect in this process. Sentinel 2 fullled the maximum
ultimate number for all-optical sensors by its VNIR b2 (7081 lines).
DEMs are still efcient sources for comprehensive terrain, and
morphometric analysis (Ashmawy et al., 2018) besides lineaments
elicitation through hill shade analysis, and its supremacy over all the
optical multispectral data, having the same resolution, is boosted in this
study. In other words, the smallest number of extracted lines from DEMs
exceeds the highest one extracted from multispectral data, provided
that, both sources have the same pixel value Fig. 3a. Consequently,
DEMs are strongly recommended over multispectral data in lineament
extraction. Among SRTM, NASA, and ASTER DEMs, the latter evidenced
its superiority over the formers by extracting 1231 lines compared to
1052 for SRTM DEM and 1148 for NASA DEM, as described in Table 4, at
the same azimuth (0◦or 180◦). Which in turn, indicates the predomi-
nance of EW, NW-SE, NE-SW DEM extracted lineaments, coinciding with
the structural setting of the study area (Zoheir et al., 2019a). We also
considered the differences in timings of DEMs acquisition, which may
result in any movements by selecting a case study of low geodynamics in
the last 20 years compared to the coastal and northern part of the Red
Sea (around the Gulf of Suez and the Gulf of Aqaba) (Abd El Nabi, 2012;
Mohamed et al., 2019; Sawires et al., 2015). Moreover, we performed 16
lineament extraction processes for each DEMs over the main eight azi-
muths and by different angles, and ASTER DEM supremacy was
conrmed over SRTM and NASA DEMs in all directions as shown in
Table 4. ALOS PALSAR is the fourth DEM implemented for this study
with its ner resolution of 12.5 m, outweighs all the data types and
numbers in its ability to extract lineaments by giving the topmost line-
aments number (9453) (Fig. 3a and b) at the 0◦azimuths and 30◦alti-
tudes. This number obviously conrms the robustness of DEMs in
lineament extraction especially if they are built from radar data. Thus,
ALOS PALSAR DEM is strongly recommended for usage in geological
applications. Besides that, this number introduced an odd issue to the
well-established inversely proportional relationship between the pixel
value and the extracted number of lineaments as it transcends the
number of lines extracted from all 10 m optical data with its 12.5m
resolution Fig. 3b. Over and above, it also surpasses sentinel 1 (VH and
VV, 8229 and 9043 respectively, which also have a cell size of approx-
imately 10m. The best results of lineament maps from different cate-
gories are displayed in Fig. 4 a-h. It is should be emphasized that these
results are surely valid for orogenic belts and rugged terrains of arid
environments. In these regions, abundant mineral deposits could be
found and mostly controlled with the tectonic lineaments, thus the more
Table 3
Bands of optical datasets, and their extracted number of lineaments in each
spatial resolution category.
Pixel size Spectral Range Optical Datasets band Lineaments number
30m VNIR Landsat 8 1 706
2 686
3 768
4 789
5 753
EO-1 ALI 1P 535
1 578
2 613
3 599
4 629
4P 627
SWIR Landsat 8 6 784
7 725
ASTER 4 576
5 539
6 582
7 423
8 453
9 459
EO-1 ALI 5P 643
5 751
7 603
20m VNIR Sentinel 2A 5 1887
6 1817
7 1844
8a 1770
SWIR Sentinel 2A 11 1574
12 1446
15m VNIR Landsat 8 8 2951
ASTER 1 2826
2 2852
3 2861
10m VNIR EO-1 ALI pan 5795
Sentinel 2A 2 7081
3 7010
4 6935
8 6754
Table 4
The extracted number of lineaments for utilized DEMs, at different azimuths and
altitudes.
Altitude (◦) Azimuth (◦) 30 m pixel value 12.5 pixel value
SRTM NASA ASTER ALOS PALSAR
30 0 1052 1131 1231 9453
45 913 1045 1189 9046
90 868 1096 1054 8764
135 955 1047 1053 9042
180 1094 1186 1165 9303
225 924 1034 1191 8886
270 909 1037 1052 8946
315 945 1048 1061 8960
45 0 1012 1148 1186 9438
45 983 968 1075 8913
90 905 971 932 8493
135 915 1018 997 8933
180 1047 1062 1210 9131
225 910 1029 1084 8782
270 868 934 1015 8887
315 975 1062 1035 9043
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
6
efcient the data source implemented in extracting linear features the
better fathoming of promising tectonic trends of several economic de-
posits. Other conditions (e.g. humid, vegetated, and densely populated
regions) may affect the extracted number of lineaments. It should also be
emphasized that the extracted number of lineaments considerably varies
from one rock type to another depending on inherent lithological
properties (e.g., claystone generally exhibits higher lineament density
compared to limestones), tectonic setting, geomechanical situation, etc.
Thus, the study measures the total number of lineaments extracted from
various rock units within the same area, to ensure a wise evaluation (the
registered numbers are extracted from 9 lithological units not only ho-
mogenous rock units.
3.1. Lineament lengths analysis
At 30 m spatial resolutions, the best data could be ordered in the next
sequence started with the ablest in extracting lineaments as follows,
ASTER DEM, NASA DEM, SRTM DEM, VNIR b4 L8, SWIR b6 L8, VNIR b4
EO1, and SWIR b6 ASTER (Fig. 3a). For all of these data, the highest
percentage of lineaments numbers have lengths between 400 and 600 m
and are arranged as follows 26.64%, 26.13%, 25.80%, 24.96%, 23.72%,
23.68%, and 22.16% respectively, as shown in Fig. 5 a–d. At 20 m pixel
size, the highest percentage of lineaments numbers are of lengths
ranging between 200 and 400 m and are registered for VNIR S2 b5
(Fig. 5 e) and SWIR S2 b11 as, 41.38% and 38.43%, respectively. At 15
m spatial resolution data (PC1 L8, panchromatic L8 b8, ASTER VNIR b3,
and PC1 ASTER) the dominant line lengths are between 200 and 300 m,
and by percentages of 24.38%, 22.83%, 23.83%, and 23.7% respectively
(Fig. 5 f and g). At 12.5 m spatial resolution, ALOS PALSAR DEM fol-
lowed the rule by giving prevailing line lengths as 100–200m and
200–300m, with a percentage of 31.31% and 23.23% respectively
(Fig. 5 h). Decreasing the line lengths with lowering pixel size is still the
controlling affair when 10 m pixel resolution data (Panchromatic EO1,
VNIR sentinel 2 band 2, VH Sentinel 1, and VV sentinel 1) showed the
predominant lineament lengths as 100–200 m by the following per-
centages 34.6%, 37.00%, 39.89%, and 41.19% respectively (Fig. 5 i–l).
3.2. Accuracy assessment
3.2.1. Density
Density and lineaments analyses seem to be inseparable, as the
former estimates the concentration of the latter per unit area (Lachaine,
Fig. 3. (a) The best data types are ordered according to the ability of lineaments extracting, in each spatial resolution category (b) Ordering of the best data types,
showing directly proportional relationship of extracted number of lineaments and pixel resolution except an odd behavior of ALOS PALSAR DEM reecting its
notability over all the used data at the study area.
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
7
1999), thus, clearly displays their distribution. In the current study, the
results are veried through the following steps.
1 Overlaying manually digitized conrmed faults over density maps
gives reasonable tting for all the data sources (Fig. 6 a).
2 Correlating density maps with two detailed (0.5 km scale) geological
maps of Dungash and Samut areas (Zoheir et al., 2019b; Zoheir and
Weihed, 2014). An outstanding matching is noticed, where the maps
match well with higher to medium density regions proportional to
the number of lines and their distribution on the geological maps as
shown in Fig. 6 b–f.
3 Correlating the number of lines for higher density areas with visually
interpreted lines to be able to answer the following question. Are
those extracted lines (reected as a high-density area) reasonably
distributed, and reect real linear features or not? To answer this
question, we selected Gabal Mueilha area, which is composed mainly
of post-orogenic alkali feldspar granite. So, ALI 752 RGB composite is
prepared as it offers the best display for the linear features of Gabal
Mueilha. After that, lines from different sources are dropped over
Gabal Mueilha, and the best matching was noticed between the lines
and real faults, fractures, and joints of Gabal Mueilha as shown in
Fig. 6 g, h.
Generally, density analysis results have revealed that the southern
part of the study area is denser than the northern part (Fig. 6 a). This
could be interpreted by the presence of various lithologies with different
elevations in the southern part (gabbros, granites, metavolcanics, ser-
pentinites, and m´
elange) compared to the northern part of the study
Fig. 4. Grey scale images of: (a) Shaded relief map of ASTER GDEM and, (b) ASTER b3, (c) PC1 of L8, (d) Panchromatic ALI, (e) S2 b2, (f) S1 VH, (g) S1 VV and, (h)
Shaded relief map (0
◦–30◦) of ALOS PALSAR DEM, overlaid by lineaments derived from each of them.
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
8
area, which is mainly occupied by extensively weathered granites and
dykes. Consequently, more lithological boundaries, straightness of val-
leys, various slopes, and shades due to the variable elevations, as well as,
the manifested structural features that are still preserved, are reected
as lineaments compared to that vanished due to extensive weathering in
the northern part. However, ner resolution radar DEMs (ALOS PAL-
SAR) and Sentinel 1 data can reasonably extract some of these foliations
and dykes in the northern granitic bodies reected as higher density
areas, which again reect the suitability of radar data in lineament
extraction over various conditions. The central part of the study area is
obviously divided into 2 main distinct zones: the northern central
(higher density) and the southern central (lower density). The northern
central high-density region is represented by Gabal Um Salatit and Gabal
Um Salim areas.
They possess high elevations, different slopes, aspects, and hill
shades which are directly reected as lineaments. The best evidence for
this interpretation is the manifestation of these areas in all the DEMs
irrespective of their type or even their resolution and by a density value
exceeding that derived from optical data. In contrast, and nearly for all
the density maps, the at area in the southern central part around the
latitude of 25◦4" N, consists mainly of wadi deposits and mining work
tailings appeared as lower density areas.
3.2.2. Orientations and trend analysis
This analysis easily identies the predominant directions of linea-
ments by counting the frequent directions and exemplifying them via
rose diagram petals. An accurate comparison between the result of this
analysis (predominant bearings) with the prevailing lineaments trends
of previous studies boost the study and is considered as a verication
way for the study results (Hashim et al., 2013; Lachaine, 1999; Meshkani
et al., 2013). Toward this target, the percentage of the total population
(lineaments) are represented using petal radii in the angular spacing of
Fig. 5. Distribution histograms showing the lengths of lineaments from the different data types: (a) ASTER DEM, (b)NASA DEM, (c)SRTM DEM and, (d)VNIR b4 L8,
manifesting 400–600 m lineament length as the dominant at 30m resolution. (e) 20m S2 b5, giving 200–400m as a dominant lineament length. At 15 m: (f) PC1 L8,
and (g) ASTER VNIR b3, show the preference to 200–300m lineament lengths. (h) 12.5m ALOS PALSAR DEM manifest 100–200m lineament length as the dominant
at this pixel size. (i) Panchromatic ALI, (j) VNIR sentinel 2 band 2, (k) VH Sentinel 1 and, (l) VV sentinel 1, giving 100–200 m as a dominant lineament length.
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
9
15◦half style rose diagrams with numbered half-circles indicating the
frequencies. The results from optical L8, ASTER, ALI, and S2 b5 indicate
that the predominant trends are NE-SW and WNW-ESE as shown in
Fig. 7 a–c. However, b2 of S2, with its higher number of lineaments and
shorter lengths gives the predominance to NNE-SSW direction with
preserving NE-SW and WNW-ESE trend as the second common trend,
(Fig. 7 d). Consequently, it is noticed that when the lineaments get
shorter (by increasing spatial resolution), the vertical trending becomes
controlling with preserving the common NE-SW and WNW-ESE. This
case is conrmed by S1 VH and VV polarizations that give the preference
to NNE-SSW, NNW-SSE, and NW-SE rather than NE-SW and WNW-ESE
(Fig. 7 e–f). For DEMs, the highest numbers of lineaments are ob-
tained by azimuths (0◦and 180◦) and by performing trend analysis for
this product, it is reasonably noticed the absence of sub-vertical trend
but manifesting the other 3 main trends as shown in Fig. 7 g.
Overall and from the whole trend analysis, it is deduced that
preferred lineaments trends could be grouped into 4 main categories as
follow, (1) Sub-vertical trends (NNE-SSW to NNW-SSE), (2) Sub-
horizontal trends (ENE-WSW to WNW-ESE), (3) NE-SW and, (4) NW-
SE, as shown in Fig. 7. These results are embedded into a profound
and careful geologic evaluation with the detailed previous studies. The
extracted trends perfectly coincide with recently published results
concerning the structural setting of the study area. This study grouped
the deformation phases and their relevant structures into 3 main phases
D1, D2, and D3 (Zoheir et al., 2019a). D1 is NNW–SSE shortening, which
is responsible for NE- SW trending lines, D2 is NNE-SSW Shortening
results in NW–SE trends. Moreover, D1 and D2 collectively participated
in the generation of sub-horizontal trends (ENE-WSW to WNW-ESE).
Whereas, lines possessing sub-vertical trends (NNE-SSW to NNW-SSE)
resulted from the D3 phase named as E–W Oblique Convergence.
4. Discussion
Due to the great importance of lineaments in several applications,
convenient and serviceable automated extraction methods, several re-
searchers studied lineament extraction from optical and radar satellite
Fig. 6. (a) Density map of lineaments extracted from ALOS PALSAR DEM, overlain by manually digitized faults from the geological map, (b) RGB 752 L8 composite
as a location map for the three selected test areas (green box is Dungash, the yellow box is Samut and red box is Mueilha), (c) Dungash geologic map (Zoheir and
Weihed, 2014), and (d) Its ALOS PALSAR DEM density map, (e) Samut geologic map (Zoheir et al., 2019b), and (f) Its ALOS PALSAR DEM density map, (g)
Pan-sharpened 10m RGB 752 ALI image of Gabal Mueilha superimposed by automatically extracted lineaments from ALOS PALSAR DEM, and (h) its density map.
(For interpretation of the references to color in this gure legend, the reader is referred to the Web version of this article.)
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
10
data (Bruning et al., 2011; Koçal, 2004; Mostafa and Bishta, 2005), and
DEMs (Abdullah et al., 2010; Al-Obeidat et al., 2016). However, the
implemented studies in comparing lineaments from various data sources
are still countable. Authors of (Hung et al., 2005)compared the extracted
lineaments from Landsat ETM and ASTER images and stated that the
VNIR band 3 of ASTER and the fused band 4 of the Landsat ETM gave the
best results. Furthermore, the authors (Op. cit.) stated that the accuracy
of extracted lineaments relies mainly on the spatial resolution of the
imagery, a higher resolution imagery results in a higher quality of
lineament map. This is not always the case in remote sensing studies and
this concept needs to be better outlined and limited in remote sensing
studies by adding the next sentence “for the same data type”. For
example, in this study, ALOS PALSAR DEM (12.5m) provided a better
lineament extraction result than the ner S1 & S2 (10m bands).
Furthermore, in this study, VNIR ASTER imagery considerably shows
higher accuracy in lineament extraction than even its PCs, which is
congruous with earlier observations of (El-Magd et al., 2015; Hung et al.,
2005; Shebl et al., 2021). Authors of (Adiri et al., 2017) compared
ASTER, Landsat 8, and Sentinel 1 images in their efciency in lineament
extraction and claried great efciency in automatic lineaments
extraction of the latter, compared to the formers, which is conrmed by
observations of this study. They (Adiri et al., 2017) accentuated the role
of Sentinel 1 VH polarization, contrary to this study, which shows the
better performance of VV than VH polarization. One of the recent studies
(Javhar et al., 2019) compared Sentinel 2, Landsat 8, and Sentinel 1
images and in a harmony with the current results, emphasized the great
fulllment of radar data over optical data in lineament extraction.
DEM superiority over optical data is interpreted by its independence
from sensor characteristics (Sahoo et al., 2018), and radar superiority is
referred to the absence of weather conditions effect (Adiri et al., 2017).
S1 (C-band) is less efcient when compared to ALOS PALSAR (L-band)
data especially in this arid environment where L-band radar waves
capability of penetration become higher for thin sedimentary covers
revealing more subsurface information. The spatial resolution effect of
several data types needs to be considered and data of the same pixel size
should be compared together. In this study, the comparison incorpo-
rated pixel size to be wise enough to compare the capability of data as it
is unfair, for example, to compare ASTER and Landsat 8 (15m or 30m) to
Fig. 7. Rose diagrams showing the orientations of lineaments extracted from: (a) ASTER b6; (b) S2 b5; (c) ASTER b3; (d) S2 b2; (e) S1 VH; (f) S1 VV and; (g) ALOS
PALSAR DEMs, at azimuth (0
◦) and altitude (30◦). (h) manifestation of the four common trends over Gabal Mueilha: Sub-vertical trends (NNE-SSW to NNW-SSE) is
represented by green line; Sub-horizontal trends (ENE-WSW to WNW-ESE) is highlighted by yellow lines; NE-SW trend is represented by red color and cyan color for
denoting NW-SE orientation. Moreover, and in a sinistral sense of displacement, the yellow lines are shifted in a way manifesting the existence of NNE-SSW fault
(green line). (For interpretation of the references to color in this gure legend, the reader is referred to the Web version of this article.)
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
11
sentinel VV and VH (10m). It is worth mentioning that manual linea-
ment extraction or at least visual check of the automatically extracted
lineaments is still indispensable whatever the implemented data source
or the applied technique. Of course, automatic lineament extraction
could save time, effort, avoid estimation errors, be independent of image
qualities or user experiences, increases objectivity (based on a computer
algorithm) rather than subjectivity (Ramli et al., 2010) accompanied
with manual methods. However, manual methods could easily set apart
tectonic and anthropogenic linear features (Javhar et al., 2019) and
assign full chronological sequences between the derived lineaments (e.
g., detection of a relatively older lineament pattern overlain by a
younger pattern) which is too difcult in automatic technique. The latter
is a valuable additional tool for the structural analysis, however, it has to
be combined into a thorough geologic evaluation.
5. Conclusions
In this study, Landsat OLI, ASTER, EO 1 ALI, Sentinel 2A, ALOS
PALSAR, SRTM, NASA, ASTER V3 DEMs, and Sentinel 1 were evaluated
for their competence in lineament extraction. Their results are ordered
based on the pixel size value aiming to recommend this scheme to the
geological scientic community. Our main ndings are.
•At 30 m spatial resolution, the preferred optical data types recom-
mended for usage, for each data type, are ASTER (b6 SWIR), ALI (b5
SWIR, b4 VNIR), and L8 (b6 SWIR, b4 VNIR) as the best data source
for this pixel size. At the same resolution, DEMs are arranged ac-
cording to their efciency in extracting lineaments as SRTM, NASA,
and ASTER v3 as the best.
•At 20 m, S2 b5 VNIR exceeds S2 b11 SWIR.
•At 15 m pixel value, in increasing order, the best data are PC1
ASTER, ASTER b3, L8 b8, and L8 PC1 as the best source for extracting
lineaments.
•At 12.5 m pixel value, ALOS PALSAR gives an outstanding result that
exceeds all the implemented data types whatever their pixel size
value.
•At 10 m cell size, radar VH & VV polarizations overcome optical data
of panchromatic band of EO1 ALI and S2 b2.
The novelty in this study could be concluded in comparing nine
different data types, categorized according to their spatial resolution, in
lineaments extraction. The study pointed out that DEMs and radar data
exceeds optical data sources in their capability of lineament extraction.
DEM preponderance may be due to independence from sensor charac-
teristics. Radar superiority could be explained by the sensitivity of the
radar data to geomorphology without affecting by weather conditions,
whereas the optical data are inuenced by the shading areas as well as,
by the occupation of the soil. Also, radar data could extract lineaments
from completely weathered or approximately vanished rock units.
Consequently, combining DEM with Radar data (e.g., PALSAR DEM)
transcend sentinel 1 (that is always recommended by researchers) by a
rate that could exceed the known relationship between pixel size and the
number of extracted lineaments. Finally, the study recommended the
best data types (depending mainly on the extracted number of linea-
ments and pixel size), commonly used in geological applications over a
wide range of pixel values (10, 12.5.15, 20, 30m) to meet the re-
quirements of the majority of geological studies and to help the users
give the most potential results depending on implementing the most
efcient data type.
Funding
This research received no external funding. This research is sup-
ported by University of Debrecen. Ali Shebl is funded by Stipendium
Hungaricum scholarship under the joint executive program between
Hungary and Egypt.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
The authors declare no conict of interest.
Acknowledgments
The authors are thankful to U.S. Geological Survey, Alaska satellite
facility and European Space Agency (ESA), for providing the data.
Thanks to Prof. Mahmoud Ashmawy, Prof. Mohamed Abd El-wahed and
Prof. Samir Kamh, for their kind support. The authors also greatly
appreciate the referee’s valuable and profound comments.
References
Abd El Nabi, S.H., 2012. Curie point depth beneath the Barramiya–Red Sea coast area
estimated from spectral analysis of aeromagnetic data. J. Asian Earth Sci. 43,
254–266. https://doi.org/10.1016/J.JSEAES.2011.09.015.
Abd El-Wahed, M.A., Zoheir, B., Pour, A.B., Kamh, S., 2021. Shear-related gold ores in
the wadi hodein shear belt, south eastern desert of Egypt: analysis of remote sensing,
eld and structural data. Minerals 11, 474. https://doi.org/10.3390/min11050474.
Abdullah, A., Akhir, J.M., Abdullah, I., 2010. Automatic mapping of lineaments using
shaded relief images derived from digital elevation model (DEMs) in the
Maran–Sungi Lembing area, Malaysia. Electron. J. Geotech. Eng. 15, 949–958.
Adhab, S.S., Hassan, M.A., 2014. Lineament automatic extraction analysis for Galal
Badra river basin using Landsat 8 satellite image. Iraqi J. Phys. 12, 44–55.
Adiri, Z., El Harti, A., Jellouli, A., Lhissou, R., Maacha, L., Azmi, M., Zouhair, M.,
Bachaoui, E.M., 2017. 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. Adv. Sp. Res. 60, 2355–2367.
Akinluyi, F.O., Olorunfemi, M.O., Bayowa, O.G., 2018. Investigation of the inuence of
lineaments, lineament intersections and geology on groundwater yield in the
basement complex terrain of Ondo State, Southwestern Nigeria. Appl. water Sci. 8,
49.
Al-Obeidat, F., Feltrin, L., Marir, F., 2016. Cloud-based lineament extraction of
topographic lineaments from NASA shuttle radar topography mission data. Procedia
Comput. Sci. 83, 1250–1255.
Alizadeh, M., Ngah, I., Hashim, M., Pradhan, B., Pour, A.B., 2018. A hybrid analytic
network process and articial neural network (ANP-ANN) model for urban
earthquake vulnerability assessment. Rem. Sens. 10, 975.
Ashmawy, M.H., Abd El-Wahed, M.A., Kamh, S.Z., Shebl, A., 2018. Scientic Journal of
Basic and Applied Sciences Comparative study of the drainage basin morphometry
extracted from topographic maps and SRTM DEMs: an example from Ghadir
watershed, Eastern Desert, Egypt. Sci JBAS 39, 52–64.
Ashournejad, Q., Hosseini, A., Pradhan, B., Hosseini, S.J., 2019. Hazard zoning for
spatial planning using GIS-based landslide susceptibility assessment: a new hybrid
integrated data-driven and knowledge-based model. Arab. J. Geosci. 12, 126.
Bhuiyan, C., 2015. Hydrological characterisation of geological lineaments: a case study
from the Aravalli terrain, India. Hydrogeol. J. 23, 673–686.
Bruning, J.N., Gierke, J.S., Maclean, A.L., 2011. An approach to lineament analysis for
groundwater exploration in Nicaragua. Photogramm. Eng. Remote Sens. 77,
509–519.
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.
Dasho, O.A., Ariyibi, E.A., Akinluyi, F.O., Awoyemi, M.O., Adebayo, A.S., 2017.
Application of satellite remote sensing to groundwater potential modeling in Ejigbo
area, Southwestern Nigeria. Model. Earth Syst. Environ. 3, 615–633.
El-Magd, I.A., Mohy, H., Basta, F., 2015. Application of remote sensing for gold
exploration in the Fawakhir area, Central Eastern Desert of Egypt. Arab. J. Geosci. 8,
3523–3536. https://doi.org/10.1007/s12517-014-1429-4.
Engdahl, M.E., Hyyppa, J.M., 2003. Land-cover classication using multitemporal ERS-
1/2 InSAR data. IEEE Trans. Geosci. Remote Sens. 41, 1620–1628.
Hashim, M., Ahmad, S., Johari, M.A.M., Pour, A.B., 2013. Automatic lineament
extraction in a heavily vegetated region using Landsat Enhanced Thematic Mapper
(ETM+) imagery. Adv. Sp. Res. 51, 874–890.
Hong, S.H., Wdowinski, S., 2014. Multitemporal multitrack monitoring of wetland water
levels in the orida everglades using alos palsar data with interferometric
processing. IEEE Geosci. Remote Sens. Lett. 11, 1355–1359. https://doi.org/
10.1109/LGRS.2013.2293492.
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. In: Remote Sensing for Environmental Monitoring, GIS
Applications, and Geology V. International Society for Optics and Photonics,
p. 59830T.
Ibrahim, U., Mutua, F., 2014. Lineament extraction using landsat 8 (OLI) in Gedo,
Somalia. Int. J. Sci. Res. 3, 291–296.
Javhar, A., Chen, X., Bao, A., Jamshed, A., Yunus, M., Jovid, A., Latipa, T., 2019.
Comparison of multi-resolution optical Landsat-8, Sentinel-2 and radar Sentinel-1
A. Shebl and ´
A. Cs´
amer

Remote Sensing Applications: Society and Environment 24 (2021) 100617
12
data for automatic lineament extraction: a case study of Alichur area, SE Pamir. Rem.
Sens. 11, 778.
Koçal, A., 2004. A Methodology for Detection and Evaluation of Lineaments from
Satellite Imagery.
Koike, K., Ichikawa, Y., 2006. Spatial correlation structures of fracture systems for
deriving a scaling law and modeling fracture distributions. Comput. Geosci. 32,
1079–1095.
Lachaine, G., 1999. Structures g´
eologiques et lin´
eaments, Beauce (Qu´
ebec): apport de la
t´
el´
ed´
etection. Universit´
e de Sherbrooke.
Magaia, L.A., Goto, T., Masoud, A.A., Koike, K., 2018. Identifying groundwater potential
in crystalline basement rocks using remote sensing and electromagnetic sounding
techniques in central western Mozambique. Nat. Resour. Res. 27, 275–298.
Manuel, R., Brito, M.D.G., Chichorro, M., Rosa, C., 2017. Remote sensing for mineral
exploration in central Portugal. Minerals 7, 184.
Marghany, M., Hashim, M., 2010. Lineament mapping using multispectral 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.
Masoud, A.A., Koike, K., 2011. Auto-detection and integration of tectonically signicant
lineaments from SRTM DEM and remotely-sensed geophysical data. ISPRS J.
Photogramm. Remote Sens. 66, 818–832.
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.
Mohamed, A.E.-E.A., El-Hadidy, M., Deif, A., Elenean, K.A., 2019. Seismic hazard studies
in Egypt. https://doi.org/10.1016/j.nrjag.2012.12.008 1. https://doi.org/10.1016/
J.NRJAG.2012.12.008, 119-140.
Mostafa, M.E., Bishta, A.Z., 2005. Signicance of lineament patterns in rock unit
classication and designation: a pilot study on the Gharib-Dara area, northern
Eastern Desert, Egypt. Int. J. Rem. Sens. 26, 1463–1475.
Muhammad, M.M., Awdal, A.H., 2012. Automatic mapping of lineaments using shaded
relief images derived from digital elevation model (DEM) in Erbil-Kurdistan,
northeast Iraq. Adv. Nat. Appl. Sci. 6, 138–147.
Nath, B., Niu, Z., Acharjee, S., 2017. Pre-Earthquake Anomaly Detection and Assessment
through Lineament Changes Observation Using Multi-Temporal Landsat 8-OLI
Imageries: Case of Gorkha and Imphal. Multi-Purposeful Appl. Geospatial Data.
Nath, B., Niu, Z., Mitra, A.K., 2019. Observation of short-term variations in the clay
minerals ratio after the 2015 Chile great earthquake $$(8.3 M_ {\mathrm {w}}) $$
using Landsat 8 OLI data. J. Earth Syst. Sci. 128, 117.
Pirasteh, S., Pradhan, B., Safari, H.O., Ramli, M.F., 2013. Coupling of DEM and remote-
sensing-based approaches for semi-automated detection of regional geostructural
features in Zagros mountain, Iran. Arab. J. Geosci. 6, 91–99.
Pour, A.B., Hashim, M., 2015a. Structural mapping using PALSAR data in the central
gold belt, Peninsular Malaysia. Ore Geol. Rev. 64, 13–22.
Pour, A.B., Hashim, M., 2015b. 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, 170–188.
Pour, A.B., Hashim, M., 2014. Structural geology mapping using PALSAR data in the Bau
gold mining district, Sarawak, Malaysia. Adv. Sp. Res. 54, 644–654.
Pour, A.B., 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, 368–385.
Qari, M.H.T., 2011. Lineament extraction from multi-resolution satellite imagery: a pilot
study on wadi Bani Malik, Jeddah, Kingdom of Saudi Arabia. Arab. J. Geosci. 4,
1363–1371.
Rahnama, M., Gloaguen, R., 2014. TecLines: a MATLAB-based toolbox for tectonic
lineament analysis from satellite images and DEMs, part 1: line segment detection
and extraction. Rem. Sens. 6, 5938–5958.
Ramli, M.F., Yusof, N., Yusoff, M.K., Juahir, H., Shafri, H.Z.M., 2010. Lineament
mapping and its application in landslide hazard assessment: a review. Bull. Eng.
Geol. Environ. 692, 215–233. https://doi.org/10.1007/S10064-009-0255-5, 2009,
69.
Richards, J.A., others, 2009. Remote Sensing with Imaging Radar. Springer.
Saepuloh, A., Haeruddin, H., Heriawan, M.N., Kubo, T., Koike, K., Malik, D., 2018.
Application of lineament density extracted from dual orbit of synthetic aperture
radar (SAR) images to detecting uids paths in the Wayang Windu geothermal eld
(West Java, Indonesia). Geothermics 72, 145–155.
Sahoo, S., Das, P., Kar, A., Dhar, A., 2018. A forensic look into the lineament, vegetation,
groundwater linkage: study of Ranchi District, Jharkhand (India). Remote Sens.
Appl. Soc. Environ. 10, 138–152.
Sawires, R., Pel´
aez, J.A., Fat-Helbary, R.E., Ibrahim, H.A., 2015. A review of seismic
hazard assessment studies and hazard description in the building codes for Egypt.
Acta Geod. Geophys. 2015 512, 151–180. https://doi.org/10.1007/S40328-015-
0117-5, 51.
Shebl, A., Abdellatif, M., Elkhateeb, S.O., Cs´
amer, ´
A., 2021. Multisource data analysis for
gold potentiality mapping of Atalla area and its environs, central Eastern Desert,
Egypt. Miner. 2021 (11), 641. https://doi.org/10.3390/MIN11060641, 641 11.
Singh, P.P., Garg, R.D., 2013. Study of spectral reectance characteristics of asphalt road
surface using geomatics techniques. In: 2013 International Conference on Advances
in Computing, Communications and Informatics (ICACCI). IEEE, pp. 516–520.
Sukumar, M., Venkatesan, N., Nelson Kennedy Babu, C., 2014. A review of various
lineament detection techniques for high resolution satellite images. Int. J. Adv. Res.
Comput. Sci. Software Eng. 4, 72–78.
Takorabt, M., Toubal, A.C., Haddoum, H., Zerrouk, S., 2018. Determining the role of
lineaments in underground hydrodynamics using Landsat 7 ETM+data, case of the
Chott El Gharbi Basin (western Algeria). Arab. J. Geosci. 11, 76.
Tripathi, N.K., Gokhale, K., Siddiqui, M.U., 2000. Directional morphological image
transforms for lineament extraction from remotely sensed images. Int. J. Rem. Sens.
21, 3281–3292.
Zoheir, B., El-Wahed, M.A., Pour, A.B., Abdelnasser, A., 2019a. Orogenic gold in
transpression and transtension zones: eld and remote sensing studies of the
barramiya–mueilha sector, Egypt. Rem. Sens. 11, 2122.
Zoheir, B., Steele-MacInnis, M., Garbe-Sch¨
onberg, D., 2019b. Orogenic gold formation in
an evolving, decompressing hydrothermal system: genesis of the Samut gold deposit,
Eastern Desert, Egypt. Ore Geol. Rev. 105, 236–257.
Zoheir, B., Weihed, P., 2014. Greenstone-hosted lode-gold mineralization at Dungash
mine, Eastern Desert, Egypt. J. Afr. Earth Sci. 99, 165–187.
A. Shebl and ´
A. Cs´
amer