remote sensing Article
Characterisation of Sand Accumulations in Wadi
Fatmah and Wadi Ash Shumaysi, KSA, Using
Multi-Source Remote Sensing Imagery
ElSayed Hermas
1,2,
*, Omar Alharbi
1
, Abdullah Alqurashi
1
, Abdoul Jelil Niang
1
,
Khalid Al-Ghamdi
1
, Motirh Al-Mutiry
3
and Abudeif Farghaly
1,4
1
Geography Department, Umm Al-Qura University, Makkah 21955, Saudi Arabia;
[email protected] (O.A.); [email protected] (A.A.); [email protected] (A.J.N.);
[email protected] (K.A.-G.); [email protected] (A.F.)
2
Applications of Physical Geology Department, The National Authority for Remote Sensing and Space
Sciences (NARSS), Cairo 1564, Egypt
3
Faculty of Arts, Princess Nourah bint Abdulrahman University, Riyadh 84428, Saudi Arabia;
[email protected]
4
Geology Department, Faculty of Science, Sohag University, Sohag 82524, Egypt
*Correspondence: [email protected]; Tel.:+966-12-5494-5454
Received: 26 September 2019; Accepted: 25 November 2019; Published: 28 November 2019





Abstract:
The study area has three sand accumulations: Two in Wadi Fatmah and one in Wadi
Ash Shumaysi, midwest of Saudi Arabia. The spatial extents of these sand accumulations have
signicantly increased over the last few decades. Multi-source satellite imagery, such as CORONA
(1967, 1972), SPOT 5 (2013), LandSat TM (1986), and LandSat 8 OLI (2013), enabled monitoring and
analysis of the interplay between the changes in the anthropogenic activities and spatial expansion of
the areas of sand accumulation. The main driving force of the spatial expansion could be strongly
linked to extensive changes in the anthropogenic regimes in the middle zone of Wadi Fatmah and
its surrounding landforms and mountain masses. In this context, the once dominant agricultural
lands of the middle zone of Wadi Fatmah have been transformed into abandoned agricultural areas.
Extensive o-road driving has resulted in soil degradation. Excavation and mining activities for
urban spatial expansion are widespread over the valley oor, the adjacent bajada, and the mountain
blocks. These anthropogenic activities have remarkably induced strong wind erosion of the soil in
severe arid conditions in the middle zone of Wadi Fatmah and Wadi Ash Shumaysi. Wind erosion has
eventually produced a sucient sand budget to be transported into the areas of sand accumulation.
The primary consequence of the excess sand budget has been an increase in the spatial extents and
dune migration rates of sand accumulations in the study area. However, this increase varies from one
sand accumulation to another. In this study, we used multi-source remote sensing imagery and the
state-of-the-art COSI-Corr technology to characterize sand accumulations in the study area and to
determine the spatio-temporal changes in both the spatial extents and the dune migration rates. The
mean annual migration rates of sand dunes in the three sand accumulations ranged from 5.5 and 7.2
to 8.6 m/yr. Analysis of the spatial extent and migration rates of sand accumulations indicates that
the study area may have experienced desertication in response to changes in the anthropogenic
regimes through the last few decades.
Keywords:dune migration; COSI-Corr; hazards; remote sensing; Makkah; Saudi Arabia; ASTER
1. Introduction
In response to sparse vegetation cover and the availability of sand supply, wind acts in arid
regions to form sand accumulations [1]. They can develop in a variety of local environments, such as
Remote Sens.2019,11, 2824; doi:10.3390 /rs11232824 /journal/remotesensing

Remote Sens.2019,11, 2824 2 of 22
coastal areas [2], low-land depressions [3,4], and over valley oors in a mountainous landscape [5].
Dune elds in mountainous landscapes normally exist in association with major wide valleys and
move against or into the foothills of mountain escarpments [5,6]. Topography, landforms, and the
vegetation cover of the mountain landscape, along with human activities, can all play a critical role in
controlling wind and sand supply characteristics.
Topographic characteristics, such as the relative elevations and slope magnitude and aspect, can
greatly aect the intensity and direction of local wind [7,8], rather than their roles as potential obstacles
in front of winds, allowing sand deposition [9,10]. Alluvial landforms, such as the alluvial plains,
bajada, and the weathered sedimentary formation of the mountain masses, can represent potential
sources of sand supply to sand accumulation in mountain landscapes [1,11–15].
In addition to the impact of topography and landforms on the formation and development of sand
accumulations, both natural and agricultural vegetation signicantly inuence sand supply and dune
mobility. Vegetation can play a dual role in the aeolian processes—its absence may increase soil erosion
and dune mobility while its presence can minimize soil erosion and accordingly the magnitude of sand
mobility [16–19]. There are many factors that control the role vegetation will contribute in aeolian
processes. Anthropogenic activities are considered a potential factor that could result in signicant
disturbances to the plant community and soil in a given geographic area [20,21]. Anthropogenic
activities can drastically deteriorate both natural and agricultural vegetation, resulting in severe soil
erosion by winds [22–26]. These anthropogenic disturbances might extend to o-road driving and
overgrazing [27], and excavation and mining activities [28,29]. Both activities have been recorded
and documented in many areas across the world as direct driving forces of land degradation. These
activities signicantly contribute to the sand budget of a given sand accumulation. In this context,
the study area has experienced a severe retreat in natural and agricultural vegetation over the last
few decades in response to various and extensive unmanaged anthropogenic activities. Multi-source
remote sensing imagery and advanced digital image-processing techniques can help monitor and map
the morphologic and anthropogenic characteristics of a given geographic area and their inuences on
the spatio-temporal changes of sand accumulations.
Earth observation satellites have served the community by providing a systematic and long
archive of remote sensing images over the last 50 years. They have enabled scientists to carry out
many dune-related studies, such as dune mapping and taxonomy [30,31], pattern analysis of dune
elds [32–34], spatial analysis of dune elds [35,36], topographic and morphologic variations of
dune elds [37–39], mineralogic and sedimentologic characteristics [40], and climatic changes [41].
In addition to these studies, the determination of the rates of dune migration has received particular
emphasis from remote sensing application scientists in the eld of geomorphology (e.g., [42–52]). From
this perspective, two or more remote sensing images are accurately georeferenced and then dune
boundaries or axes are delineated for eventual comparison of the boundaries and axes to calculate
the lateral displacement of sand dunes. In contrast to in-situ eld measurements, the determination
of dune migration rates using georeferenced remote sensing images allows the dune migration to be
constrained over a wide spatial coverage. However, this classical approach to dening dune boundaries
using remote sensing images is highly aected by satellite geometry and illumination conditions [53],
along with the eect of personal bias through the denition and delineation of dune boundaries [36].
Recent advances in remote sensing techniques have submitted the co-registration of optically
sensed images and correlation (COSI-Corr) as an alternative pixel-wise quantitative approach to
dune migration. This approach relies on correlating two or more ortho-rectied images using spatial
cross-correlation to produce displacement images where the values and signs of each pixel express
the amount of surface displacement (in meters) and its direction, respectively. The approach was
submitted to the scientic community by Leprince et al. [54]. Recently, it has been used in many studies
for the purpose of determining dune migration rates and directions [55–60].
In this research, we aimed mainly to characterize the climatic, morphologic, and anthropogenic
parameters that contribute to the changes in the spatial extents and dune migration rates of sand

Remote Sens.2019,11, 2824 3 of 22
accumulations in the study area using multi-source remote sensing images and the COSI-Corr
technology. The results of this study will provide an outline of the temporal and spatial characteristics
of the aeolian activity in the study area, which can help local authorities and environmental planners
avoid its consequences in the future.
2. Study Area
Makkah City and its surroundings mainly comprise granitic basement rocks that belong to the
Precambrian. In the north and east of Makkah City, sedimentary rocks are represented as well. Complex
sets of faults and fractures characterize the area and imply a long tectonic history. These sets occur
either parallel or perpendicular to the direction of the Red Sea. Since the initiation of the tectonic
processes in the Tertiary Period, both basement and sedimentary rocks have been uplifted, resulting
in rugged topographic mountains that range in elevations from 200 to 400 m asl. Well-developed
surface stream networks have incised the rugged mountains, developing many drainage basins around
Makkah City. The largest is Wadi Fatmah, which bounds the city from the north and west. The study
area is located in the main transportation system of Wadi Fatmah drainage basin and along Wadi Ash
Shumaysi, which is a wide structural depression perpendicular to the transportation system of the
Wadi Fatmah drainage basin (Figure). Sand accumulations occupy signicant patches at the southern
edges of the Wadi Fatmah trunk valley, and along Wadi Ash Shumaysi (Figure). The spatial extents
of these sand accumulations have recently experienced signicant spatial changes. These changes
could be linked strongly to various human activities that have taken place over the last few decades.Remote Sens. 2019, 11, x FOR PEER REVIEW 4 of 24


Figure 1. Location of the study area in Saudi Arabia; A, B, and C are selected samples of sand
accumulations (SA1, SA2, SA3).
3. Data and Method
3.1. Data Acquisition and Preparation
We used different multi-source remotely sensed data in this research—namely, CORONA, SPOT
5, ASTER, Landsat images, and digital elevation model (DEM). The intelligence project CORONA,
which was declassified in 1990 [61], provides a potential historical source for long-term earth
observation. For this research, CORONA images taken in 1967 and 1972 were acquired over the study
area. As a panchromatic image, automatic extraction of the land features cannot be done using such
Figure 1.
Location of the study area in Saudi Arabia; A, B, and C are selected samples of sand
accumulations (SA1, SA2, SA3).

Remote Sens.2019,11, 2824 4 of 22
3. Data and Method
3.1. Data Acquisition and Preparation
We used dierent multi-source remotely sensed data in this research—namely, CORONA, SPOT 5,
ASTER, Landsat images, and digital elevation model (DEM). The intelligence project CORONA, which
was declassied in 1990 [61], provides a potential historical source for long-term earth observation.
For this research, CORONA images taken in 1967 and 1972 were acquired over the study area. As a
panchromatic image, automatic extraction of the land features cannot be done using such data [62];
thus, CORONA images were only used to visually delineate the spatial characteristics of the various
spatial features, such as morphological and human features in the past. We used a SPOT 5 image
acquired in 2013 from the King Abdulaziz City for Science and Technology (KACST) to monitor the
changes of the spatial distribution and extent of sand accumulations over the last decades in the study
area. The Landsat Thematic Mapper (TM) of 1986 and the Landsat Operational Imager (OLI) of 2013,
acquired from the United States Geological Survey (USGS) Global Visualization (GloVis) site, were
used for visual interpretation. Heads-up digitizing of the various morphological features, such as
sand accumulations, landforms, and human-made features, was carried out using both CORONA and
SPOT 5 images. Topographic analysis of the study area was conducted as well, using 1 arc sec digital
Elevation model (DEM) acquired by the Shuttle Radar Topography Mission (SRTM). The migration
rates of sand dunes of the various sand accumulations in the study area were determined using ve
ASTER images of a spatial resolution of 15 m that were collected from the United States Geological
Survey (USGS) Global Visualization (GloVis) site. These images were acquired in November 2011,
2012, and 2013, and September 2014 and 2015. Table
acquired ASTER images. They have almost similar pointing and orientation angles, which is suitable
for achieving successful correlation results. Finally, we obtained meteorological data on the study area
between 1985 and 2014 from the General Authority of Meteorology and Environmental Protection, KSA.
Table 1.Technical specications of Advanced Spaceborne Thermal Emission and Reection Radiometer
(ASTER) images used in the study.
Acquisition Dates
Pointing
Angle
Orientation
Angle
Sun Azimuth
Angle
Sun elevation
Angle
Spatial
Resolution
November 09, 2011 8.58 8.80 155.79 48.64 15
November 27, 2012 8.59 8.80 157.48 43.94 15
November 14, 2013 8.59 8.80 156.87 47.20 15
September 30, 2014 8.58 8.80 142.38 60.15 15
September 24, 2015 2.87 8.80 141.049 62.855 15
3.2. Determination of Dune Migration
All ASTER images were accurately ortho-rectied for eventual sub-pixel spatial cross-correlation
using the typical processing chain of COSI-Corr technology. The COSI-Corr is a plug-in ENVI
®
digital
image processing software. The ortho-rectication process involved careful selection iteratively of the
tie points between a previously ortho-rectied image and a slave image (the raw image) [63]. The tie
points were then converted into ground control points (GCPs) and eventually optimized using the
reference image, a free sinks Shuttle Radar Topographic Mission digital elevation model (DEM), and
the ancillary le of each image containing the various parameters of ASTER LA1 sensor system to
produce ve ortho-rectied images (Figure). After the ortho-rectication processes, the COSI-Corr
correlation engine was then automated to correlate the ortho-rectied images consecutively. In this
way, we arrived at four correlation pairs: 2011–2012, 2012–2013, 2013–2014, and 2014–2015. For each
pair, we adopted the same correlation parameters, which were multi-scale window sizes ranging from
64 to 32 pixels, and a 2-pixel shifting step between the sliding windows. The correlation of each pair
produced a horizontal displacement image comprising three bands: East-west displacement band,

Remote Sens.2019,11, 2824 5 of 22
north-south displacement band, and signal-to-noise ratio band. To remove the noise values from the
displacement bands and the signal-to-noise ratio band, we applied the non-local means lter [64].
The pixels of the east-west (E-W) and the north-south (N-S) displacement bands comprised values
and signals. The values refer to the pixel displacement between the master and slave images of each
correlation pair, while the signals denote the direction of such displacement. For instance, the positive
values indicate the amount of pixel displacements in the east and north whereas the negative values
indicate the amount of pixel displacements in the west and south. The east-west and north-south
displacement bands were integrated to calculate the net annual migration rates, using the approach
proposed by Necsoiu, Leprince, Hooper, Dinwiddie, McGinnis, and Walter [57]. In addition, we
carried out two days of eldwork to explore the data extracted from visual interpretation and to collect
signals of land degradation, such as soil cracks. Figure
approach adopted to achieve the aims of this research.Remote Sens. 2019, 11, x FOR PEER REVIEW 6 of 24
between the master and slave images of each correlation pair, while the signals denote the direction
of such displacement. For instance, the positive values indicate the amount of p
ixel displacements in
the east and north whereas the negative values indicate the amount of pixel displacements
in the
west and south. The east-west and north-south
displacement bands were integrated to calculate the
net annual migration rates, using the a
pproach proposed by Necsoiu, Leprince, Hooper, Dinwiddie,
ts a flowchart demonstrating the whole approach
adopted to achieve the aims of this research.
Figure 2. Five ortho-rectified and co-registered ASTER images.
Figure 2.Five ortho-rectied and co-registered ASTER images.Remote Sens. 2019, 11, x FOR PEER REVIEW 7 of 24

Figure 3. A flowchart of the adopted approach in this research.
4. Results
4.1. Climatic Parameters of the Study Area
The study area is characterized by a typical desert environment where the temperature is high
and rainfall is low [65]. The climatic investigation carried out on the study area indicated the
dominance of arid conditions. The mean annual temperature rates ranged from 24.8 to 38.2 °C.
Summer is characterized by the highest seasonal values of temperatures (Figure 4). The mean of the
minimum temperature of summer was 29.2 °C, whereas the mean of the maximum temperature of
summer was high, reaching 43.2 °C. On the other hand, the mean seasonal rates of the temperature
in winter ranged from 19.5 to 31.5 °C. Al-Harbi [66] used the temperature data collected from all
climatic stations located within and around the Wadi Fatmah drainage basin to produce a
distribution map of the temperature. The map demonstrated that the highest temperature values
were limited to the middle zone of the Wadi Fatmah basin, which is where our study area is located.
Rainfall is another factor that controls arid conditions in a given geographic area. Spatially, the
amount of rainfall significantly varies in the Wadi Fatmah drainage basin. It increases in the upstream
reaches where elevations are high, and decreases downstream [67]. The rainfall in the upstream
reaches exceeds the arid rainfall threshold, reaching ~279 mm/year [67]. On the other hand, the
middle area of Wadi Fatmah basin where the study area is located is characterized by limited rainfall
amounts ranging from 80 to 100 mm/year [66–68] (Figure 4). The limited amounts of rainfall
associated with high temperature values outweigh the opportunity for arid conditions to dominate
in the study area. This setting could potentially be affirmed by the recorded high evaporation rates.
The recorded annual rate of evaporation was 3560 mm/year [66], which exceeds many times the
precipitation rate in the study area, denoting strong arid conditions. These conditions could
potentially result in a significant water deficit [69]. In this context, wind action transports sand
fragments to the study area and deposits them there. Exploration of wind data records indicated that
the study area experiences an unidirectional wind regime from the north direction (Figure 5). The
northern wind strongly increases in summer, winter, and spring in the order of ~83%, 62%, and 59%.
In autumn, the northern wind decreases to 32% in favor of the southern and southwestern winds,
which record 45% and 23% of the total wind, respectively [66]. In addition to the arid conditions that
trigger strong aeolian processes, geomorphic features have also helped initiate and develop sand
accumulations in the study area. Figure 3.A owchart of the adopted approach in this research.

Remote Sens.2019,11, 2824 6 of 22
4. Results
4.1. Climatic Parameters of the Study Area
The study area is characterized by a typical desert environment where the temperature is high and
rainfall is low [65]. The climatic investigation carried out on the study area indicated the dominance
of arid conditions. The mean annual temperature rates ranged from 24.8 to 38.2

C. Summer is
characterized by the highest seasonal values of temperatures (Figure). The mean of the minimum
temperature of summer was 29.2

C, whereas the mean of the maximum temperature of summer was
high, reaching 43.2

C. On the other hand, the mean seasonal rates of the temperature in winter ranged
from 19.5 to 31.5

C. Al-Harbi [66] used the temperature data collected from all climatic stations located
within and around the Wadi Fatmah drainage basin to produce a distribution map of the temperature.
The map demonstrated that the highest temperature values were limited to the middle zone of the
Wadi Fatmah basin, which is where our study area is located.Remote Sens. 2019, 11, x FOR PEER REVIEW 8 of 24



Figure 4. The monthly mean of temperature (up) and the accumulated annual amounts of rainfall
(down).
20.0
22.0
24.0
26.0
28.0
30.0
32.0
34.0
36.0
38.0
Temperature (°C )
0.0
50.0
100.0
150.0
200.0
250.0
300.0
1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013
Accumulated annual rainfall (mm)
Figure 4.
The monthly mean of temperature (up) and the accumulated annual amounts of rainfall
(down).

Remote Sens.2019,11, 2824 7 of 22
Rainfall is another factor that controls arid conditions in a given geographic area. Spatially, the
amount of rainfall signicantly varies in the Wadi Fatmah drainage basin. It increases in the upstream
reaches where elevations are high, and decreases downstream [67]. The rainfall in the upstream reaches
exceeds the arid rainfall threshold, reaching ~279 mm/year [67]. On the other hand, the middle area of
Wadi Fatmah basin where the study area is located is characterized by limited rainfall amounts ranging
from 80 to 100 mm/year [66–68] (Figure). The limited amounts of rainfall associated with high
temperature values outweigh the opportunity for arid conditions to dominate in the study area. This
setting could potentially be armed by the recorded high evaporation rates. The recorded annual rate
of evaporation was 3560 mm/year [66], which exceeds many times the precipitation rate in the study
area, denoting strong arid conditions. These conditions could potentially result in a signicant water
decit [69]. In this context, wind action transports sand fragments to the study area and deposits them
there. Exploration of wind data records indicated that the study area experiences an unidirectional
wind regime from the north direction (Figure). The northern wind strongly increases in summer,
winter, and spring in the order of ~83%, 62%, and 59%. In autumn, the northern wind decreases to
32% in favor of the southern and southwestern winds, which record 45% and 23% of the total wind,
respectively [66]. In addition to the arid conditions that trigger strong aeolian processes, geomorphic
features have also helped initiate and develop sand accumulations in the study area.Remote Sens. 2019, 11, x FOR PEER REVIEW 9 of 24


Figure 5. The frequency of winds through the various wind directions.
4.2. Morphology of the Study Area
The study area comprises various geomorphic landforms, including mountain ridges, folded
mountains masses, isolated hills, bajada, and alluvial plains (Figure 6). The mountain ridges are
located in the east, southeast, and southwest of the study area. Lithologically, the mountain ridges
mainly comprise hard basement rocks of granitic composition along with metamorphic rocks, such
as schists and gneisses. The majority of the mountain ridges are structurally controlled, and aligned
either in the NE-SW, N-S, or E-W directions (Figure 6). The ridges are considerably eroded to various
extents, and mainly separated from each other by intermittent streams (Figure 6). The SW mountain
ridges are separated from the SE mountain ridges by a structurally controlled wide depression, which
is known locally as Wadi Ash Shumaysi (Figure 6). The depression trends in the NW-SE direction
and is filled with alluvial deposits transported from the adjacent eastern and western mountain
blocks by infrequent surface runoff. To the north and almost perpendicular to the trend of Wadi Ash
Shumaysi lies the alluvial plain of Wadi Fatmah. It trends in the NE-SW direction (Figure 6). North
to the alluvial plain of Wadi Fatmah, three salient mountain masses occur, which are Jabal Sidr, Jabal
Mikassar, and Jabal Daf (Figure 6). These folded mountain masses consist of un-metamorphosed to
low-grade metamorphosed sedimentary rocks known as the Fatmah Group [70]. At the foot-slope of
each mountain mass, bajada surfaces have been developed as prominent morphologic features in the
study area (Figure 6). The bajada surfaces occupy a considerable spatial extent in the order of ~24
km
2
. Although the bajadas occupy extensive areas in front of the three northern mountain masses,
they are produced from small catchments (Figure 6). The production of these considerable bajada
surfaces from small catchments clearly implies that the three mountain masses are highly erodible
surfaces, which could represent potential source areas for detrital sand to be transported by the
dominant northern wind in the study area. This hypothesis could be supported not only by the
lithologic composition of these mountains but also by the topographic characteristics of the study
area.
Figure 5.The frequency of winds through the various wind directions.
4.2. Morphology of the Study Area
The study area comprises various geomorphic landforms, including mountain ridges, folded
mountains masses, isolated hills, bajada, and alluvial plains (Figure). The mountain ridges are located
in the east, southeast, and southwest of the study area. Lithologically, the mountain ridges mainly
comprise hard basement rocks of granitic composition along with metamorphic rocks, such as schists
and gneisses. The majority of the mountain ridges are structurally controlled, and aligned either in the
NE-SW, N-S, or E-W directions (Figure). The ridges are considerably eroded to various extents, and
mainly separated from each other by intermittent streams (Figure). The SW mountain ridges are
separated from the SE mountain ridges by a structurally controlled wide depression, which is known
locally as Wadi Ash Shumaysi (Figure). The depression trends in the NW-SE direction and is lled
with alluvial deposits transported from the adjacent eastern and western mountain blocks by infrequent
surface runo. To the north and almost perpendicular to the trend of Wadi Ash Shumaysi lies the
alluvial plain of Wadi Fatmah. It trends in the NE-SW direction (Figure). North to the alluvial plain of
Wadi Fatmah, three salient mountain masses occur, which are Jabal Sidr, Jabal Mikassar, and Jabal Daf

Remote Sens.2019,11, 2824 8 of 22
(Figure). These folded mountain masses consist of un-metamorphosed to low-grade metamorphosed
sedimentary rocks known as the Fatmah Group [70]. At the foot-slope of each mountain mass, bajada
surfaces have been developed as prominent morphologic features in the study area (Figure). The
bajada surfaces occupy a considerable spatial extent in the order of ~24 km
2
. Although the bajadas
occupy extensive areas in front of the three northern mountain masses, they are produced from small
catchments (Figure). The production of these considerable bajada surfaces from small catchments
clearly implies that the three mountain masses are highly erodible surfaces, which could represent
potential source areas for detrital sand to be transported by the dominant northern wind in the study
area. This hypothesis could be supported not only by the lithologic composition of these mountains
but also by the topographic characteristics of the study area.Remote Sens. 2019, 11, x FOR PEER REVIEW 10 of 24



Figure 6. The major morphologic features in the study area demonstrated on a 30 m Digital Elevation
Model (DEM); A-B, C-D, and E-F are the spatial locations of the transects of topographic profiles
shown in Figure 7.
Three N-S topographic transects were extracted from the digital elevation model of the study
area (Figure 7). All topographic transects disclose significant variations in the elevation where the
northern mountain masses are notably higher than the southern mountain ridges in the study area.
The elevation differences, along with the high erodibility of the northern mountain masses, qualify
these masses to be potential sources of detrital sand to be transported southward. While the northern
mountain masses function as source areas of aeolian sands, the southern mountain ridges with their
orientations could efficiently function as topographic obstacles for wind-transported sands (Figure
7). In addition, the satellite imagery, field observation, and topographic transects all assert to the
presence of many other obstacles for wind-transported sand from the northern mountain masses.
One of these potential obstacles is the isolated hills (Figure 8). The role of the isolated hills as obstacles
that could initiate and help in the development of various dune forms is well monitored and
Figure 6.
The major morphologic features in the study area demonstrated on a 30 m Digital Elevation
Model (DEM); A-B, C-D, and E-F are the spatial locations of the transects of topographic proles shown
in Figure.

Remote Sens.2019,11, 2824 9 of 22Remote Sens. 2019, 11, x FOR PEER REVIEW 11 of 24

documented [10,71]. In front of all mountain ridges that flank the southern edges of the alluvial plain
of Wadi Fatmah, more than 100 NE-SW-trending isolated hills exist (Figure 8). Climbing dunes and
leeward-extended dunes occur windward and leeward of the isolated hills, respectively. The
topographic transects denote that the alluvial plains of Wadi Fatmah and Wadi Ash Shumaysi are
undulated surfaces as well (Figure 7), which qualifies them as potential obstacles against wind-
transported sands. Over these undulated surfaces, especially in low topography areas, wide areas of
bushes (scrub) were monitored using the Corona satellite image acquired in 1972 (Figure 8).

Figure 7. The topographic profiles in the study area; A-B, C-D, and E-F represents of the spatial
locations of the topographic transects shown in Figure 6.
The total area of these bushes is ~19 km
2
. The density of the bushes is spatially varied. It increases
westward where the main trunk course of Wadi Fatmah exists, and significantly decreases eastward
Figure 7.
The topographic proles in the study area; A-B, C-D, and E-F represents of the spatial
locations of the topographic transects shown in Figure.
Three N-S topographic transects were extracted from the digital elevation model of the study
area (Figure). All topographic transects disclose signicant variations in the elevation where the
northern mountain masses are notably higher than the southern mountain ridges in the study area.
The elevation dierences, along with the high erodibility of the northern mountain masses, qualify
these masses to be potential sources of detrital sand to be transported southward. While the northern
mountain masses function as source areas of aeolian sands, the southern mountain ridges with their
orientations could eciently function as topographic obstacles for wind-transported sands (Figure).
In addition, the satellite imagery, eld observation, and topographic transects all assert to the presence

Remote Sens.2019,11, 2824 10 of 22
of many other obstacles for wind-transported sand from the northern mountain masses. One of these
potential obstacles is the isolated hills (Figure). The role of the isolated hills as obstacles that could
initiate and help in the development of various dune forms is well monitored and documented [10,71].
In front of all mountain ridges that ank the southern edges of the alluvial plain of Wadi Fatmah, more
than 100 NE-SW-trending isolated hills exist (Figure). Climbing dunes and leeward-extended dunes
occur windward and leeward of the isolated hills, respectively. The topographic transects denote that
the alluvial plains of Wadi Fatmah and Wadi Ash Shumaysi are undulated surfaces as well (Figure),
which qualies them as potential obstacles against wind-transported sands. Over these undulated
surfaces, especially in low topography areas, wide areas of bushes (scrub) were monitored using the
Corona satellite image acquired in 1972 (Figure).Remote Sens. 2019, 11, x FOR PEER REVIEW 12 of 24

where dunes dominate (Figure 8). The topographic maps of 1971 also demonstrate significant
agricultural areas in the order of 14 km
2
occupying the middle zone of the Wadi Fatmah alluvial plain
around the main course of the trunk valley (Figures 6 and 8). The extensive areas of both agricultural
and natural vegetation clearly indicate that the alluvial plain was receiving an appropriate water
supply that kept the soil moist enough for plants to grow. Maintaining soil moisture would reduce
wind erosion, and accordingly minimize the production of wind-transported sands from the alluvial
plain of Wadi Fatmah. Therefore, sand supply to the area of sand accumulations was mostly
attributed to the elevated and highly erodible mountain masses in the north of the alluvial plain.
Under such conditions, the initial sand accumulations developed in the study area (Figure 6).

Figure 8. The impact of both isolated hills and mountain ridges on sand accumulations (as shown
from the comparison between the upper left image A-Corona 1967 and the upper right image B-SPOT
5 2015) and the shrinking of both natural and agricultural vegetation in favor of the formation of sand
accumulations (as shown from the comparison between the lower left image C-Corona 1972 and the
lower right image D-SPOT 5 2015).
The initial sand accumulations that were determined from the Corona images can be labelled
SA1 (sand accumulation 1), SA2 (sand accumulation 2), and SA3 (sand accumulation 3) (Figure 6).
The entire spatial extent of all the initial sand accumulations was in the order of 40 km
2
. SA1 is the
smallest and SA3 is the largest, extending over 18.7 km
2
. Both SA1 and SA2 occupy low elevations
Figure 8.
The impact of both isolated hills and mountain ridges on sand accumulations (as shown from
the comparison between the upper left image A-Corona 1967 and the upper right image B-SPOT 5
2015) and the shrinking of both natural and agricultural vegetation in favor of the formation of sand
accumulations (as shown from the comparison between the lower left image C-Corona 1972 and the
lower right image D-SPOT 5 2015).

Remote Sens.2019,11, 2824 11 of 22
The total area of these bushes is ~19 km
2
. The density of the bushes is spatially varied. It increases
westward where the main trunk course of Wadi Fatmah exists, and signicantly decreases eastward
where dunes dominate (Figure). The topographic maps of 1971 also demonstrate signicant
agricultural areas in the order of 14 km
2
occupying the middle zone of the Wadi Fatmah alluvial plain
around the main course of the trunk valley (Figures). The extensive areas of both agricultural
and natural vegetation clearly indicate that the alluvial plain was receiving an appropriate water
supply that kept the soil moist enough for plants to grow. Maintaining soil moisture would reduce
wind erosion, and accordingly minimize the production of wind-transported sands from the alluvial
plain of Wadi Fatmah. Therefore, sand supply to the area of sand accumulations was mostly attributed
to the elevated and highly erodible mountain masses in the north of the alluvial plain. Under such
conditions, the initial sand accumulations developed in the study area (Figure).
The initial sand accumulations that were determined from the Corona images can be labelled
SA1 (sand accumulation 1), SA2 (sand accumulation 2), and SA3 (sand accumulation 3) (Figure).
The entire spatial extent of all the initial sand accumulations was in the order of 40 km
2
. SA1 is the
smallest and SA3 is the largest, extending over 18.7 km
2
. Both SA1 and SA2 occupy low elevations and
slopes in the southern ank of the alluvial plain of Wadi Fatmah. They extend against the NE-SW
mountain ridges and the numerous isolated hills in front of these ridges (Figure). While SA1 and
SA2 have developed in front of the mountain ridges in the east of the study area, SA3 has developed in
the wide depression of Wadi Ash Shumaysi between the mountain blocks of the SE and SW reaches
of the study area. In the depression, a low-energy wind is possibly stimulated, which allows aeolian
processes to develop sand dunes on the oor of the depression. Although sand accumulations in the
study area occupy dierent places, they extend in the southeast direction against hillslopes of the
mountain ridges and the isolated hills, indicating the impact of the local topographic setting on the
dominant northern wind regime. The above characteristics of the climatic and morphologic features
have controlled the formation of the initial sand accumulations in the study area. Through the last
few decades and in response to massive changes in the anthropogenic changes, these initial sand
accumulations have extensively expanded, with signicant high rates of dune migration.
5. Discussion
5.1. Anthropogenic Activities in the Study Area
A review of the previous literature, along with an analysis of the available satellite images, might
help dene the relationship between human activities and the increasing sand budget in the study
area over the last few decades. The principal human activity in Wadi Fatmah drainage basin for
a long time has been agriculture. Until 40 years ago, Wadi Fatmah valley was the main source of
various agricultural crops for the major surrounding big cities, such as Makkah, Taif, and Jeddah [72].
In this context, the middle zone of Wadi Fatmah, in which the study area is located, was producing
more than 40% of all the agricultural farms of Wadi Fatmah [73]. The primary source of water for the
agricultural and other domestic activities in the middle zone of Wadi Fatmah was the underground
water discharged from shallow alluvial aquifers [67,73]. Water discharge for agricultural activities
and associated domestic practices primarily relied on water springs, which were well-known features
in the Wadi Fatmah drainage basin. Over the last six decades, the number of these springs has
decreased signicantly from about 400 to only 5 [73]. The drastically decreasing number of water
springs is mainly attributed to a massive unmanaged discharging strategy adopted in the 1970s to
meet the domestic fresh-water requirements of Makkah and Jeddah cities. This strategy required the
construction of new water wells to discharge fresh water from the alluvial aquifers through water
pipelines to these cities [73]. The direct consequence of this water-discharging strategy was the
decreasing water level of the alluvial aquifers, and accordingly, the shrinkage of the number of water
springs. In addition, the alluvial aquifers have experienced a severe lack of replenishment over the
last decades as a result of the construction of the Wadi Fatmah Dam in 1985 (Figure). The dam is

Remote Sens.2019,11, 2824 12 of 22
located at Abu Hasaniyah village where surface runoof three major upstream sub-basins accumulates
(Figure). According to the Ministry of Water, Environment, and Agriculture, the dam is designed to
store 20 million m
3
of fresh water annually. The surface runoof these upstream sub-basins was the
principal source of replenishment of the alluvial aquifers in the middle zone of Wadi Fatmah drainage
basin, including the study area [73]. Under such conditions, water required for agricultural practices
has undergone a drastic loss over the last few decades, eventually resulting in the transformation
of ~19 km
2
of agricultural lands into abandoned agricultural areas (Figure), allowing the era of
land degradation to start. Therefore, vegetation cover has diminished, and decit soil moisture levels
have been sustained [23,74–76]. The reduction in both vegetation cover and soil moisture content
has resulted in intense wind erosion [22,23,25,26] that has made the sand grains easily detach and be
transported by the wind [17]. Visual interpretation of the 1972, 1986, and 2013 images demonstrates a
substantial reduction in both natural and agricultural vegetation cover (Figure). Observations from
eldwork reveal the presence of drought signals, such as soil cracks (Figure). These cracks are widely
distributed in the area of bushes, implying a sustained decit in soil moisture. These indices could
potentially indicate that the study area might have experienced serious drought conditions through
the last decades in response to the common arid conditions and the changes in the un-managed and
planned anthropogenic changes. Accordingly, most natural vegetation cover has been extensively
transformed into mobile sand dunes, while the majority of the agricultural lands have become bare
land or been replaced by other anthropogenic activities (Figure).
In addition, many other anthropogenic activities can increase the rate and amount of sand supply
in a given geographic region. These activities may include o-road driving [27], intense grazing [77],
and excavation for construction purposes [28,29]. Figure
number of o-roads on both the northern and southern alluvial plains around Wadi Fatmah trunk
valley, along with the presence of recent o-roads over the surfaces of the alluvial fans in the northwest
of the study area. A comparison of the satellite images discloses a tangible urban spatial expansion
either on the valley oor, over the surface of the alluvial fans, or over the mountain masses in the
northwest reaches of the Wadi As Shumayi depression (Figure). Urban spatial expansion over the
mountain masses normally requires fragmentation and removal of detrital sediments before urban
construction (Figure). This process of excavation could have released a signicant amount of ne
sands that may have participated in the expansion of sand accumulations [29] in the study area.
The above characterization of the various climatic, morphologic, and anthropogenic activities in the
study area shows intense soil erosion constraints that could result in a signicant sand supply for
aeolian saltation processes [23,24]. With the increase of the sand supply, the spatial extent of sand
accumulations is highly expected [10], and sand dune development and migration rates can intensify
as well [29,78]. Both the spatial extent and dune movement of the three sand accumulations in the
study area were monitored using COSI-Corr technology.

Remote Sens.2019,11, 2824 13 of 22Remote Sens. 2019, 11, x FOR PEER REVIEW 14 of 24

the un-managed and planned anthropogenic changes. Accordingly, most natural vegetation cover
has been extensively transformed into mobile sand dunes, while the majority of the agricultural lands
have become bare land or been replaced by other anthropogenic activities (Figure 9).

Figure 9. Multi-source satellite imagery at different dates show significant shrinkage of both
agricultural and natural vegetation (row A and B) and spatial urban activities at the expense of the
adjacent mountain block (row C).
In addition, many other anthropogenic activities can increase the rate and amount of sand
supply in a given geographic region. These activities may include off-road driving [27], intense
grazing [77], and excavation for construction purposes [28,29]. Figure 6 demonstrates the presence of
a considerable number of off-roads on both the northern and southern alluvial plains around Wadi
Fatmah trunk valley, along with the presence of recent off-roads over the surfaces of the alluvial fans
in the northwest of the study area. A comparison of the satellite images discloses a tangible urban
spatial expansion either on the valley floor, over the surface of the alluvial fans, or over the mountain
Figure 9.
Multi-source satellite imagery at dierent dates show signicant shrinkage of both agricultural
and natural vegetation (row A and B) and spatial urban activities at the expense of the adjacent mountain
block (row C).
5.2. Dune Migration Rates of Sand Accumulations in the Study Area
The correlation process between ve ortho-rectied ASTER images produced four correlation
images, which are the 2011–2012, 2012–2013, 2013–2014, and the 2014–2015 images (Figure). Each
correlation image comprises E-W and N-S displacement bands (Figure) along with the signal-to-noise
ratio (SNR) band. The pixel values of the E-W and N-S bands express pixel displacements between
the input ortho-rectied ASTER images. Table
values of the E-W and N-S displacement bands. The mean values of pixel displacements of the E-W
bands recorded higher values than the N-S displacement bands of all the correlation images (Table).
In addition, the mean values of the E-W displacements bands are characterized by positive signs
whereas the mean values of the N-S displacements values are characterized by negative signs. The mean
values and signs of the E-W and N-S bands indicate that the overall pixel displacements tend to be in
the southeast. To determine the net annual displacement in the study area, we integrated the E-W and
N-S displacement bands of each correlation image using the approach proposed byNecsoiu et al.[79].
The integration produced four images representing the annual rates of pixel displacements in the

Remote Sens.2019,11, 2824 14 of 22
study area from November 2011 to September 2015 (Figure). The gure indicates that the highest
values of the annual rates of pixel displacement are associated with the three sand accumulations in
the study area, indicating dune migration on an annual scale. The pixel displacements of the three
sand accumulation areas were averaged to produce dune eld statistics for the purpose of studying
the spatial and temporal patterns of dune migration in the study area.Remote Sens. 2019, 11, x FOR PEER REVIEW 16 of 24


Figure 10. The E-W and N-S bands produced from the correlation of the ortho-rectified and co-
registered ASTER images.
Figure 10.
The E-W and N-S bands produced from the correlation of the ortho-rectied and co-registered
ASTER images.

Remote Sens.2019,11, 2824 15 of 22
Table 2.
Statistics of the values of pixel displacements of the produced E-W and N-S displacement
bands of the correlation pairs.
Statistics
Correlation Pairs
2011–2012 2012–2013 2013–2014 2014–2015
E-W N-S E-W N-S E-W N-S E-W N-S
Min. 13.4724.4819.2216.2730.4427.524.1624.61
Max. 24.94 19.75 18.48 15.18 29.54 28.46 25.61 23.95
Mean 1.22 0.38 0.76 0.38 0.79 1.03 1.82 0.83
SD 4.22 2.52 2.97 2.29 3.63 2.91 3.79 3.01Remote Sens. 2019, 11, x FOR PEER REVIEW 17 of 24


Figure 11. The net annual migration rates through the time period 2011–2015 of the three sand
accumulations in the study area.
Table 3 presents the statistics of dune movements within the three sand accumulations. The table
rows represent the spatial changes in the magnitude of the annual rates of dune movements within
sand accumulations. The table columns represent the temporal changes in the magnitude of the
annual rates of dune movements through all sand accumulations.
Table 3. Statistics of the annual migration rates of sand dunes in the three studied sand
accumulations.
Correlation pairs of annual
rates
Statistics of annual
rates
Dune areas
SA1 SA2 SA3 Temporal ranges and
means
Figure 11.
The net annual migration rates through the time period 2011–2015 of the three sand
accumulations in the study area.

Remote Sens.2019,11, 2824 16 of 22
Table
rows represent the spatial changes in the magnitude of the annual rates of dune movements within
sand accumulations. The table columns represent the temporal changes in the magnitude of the annual
rates of dune movements through all sand accumulations.
Table 3.
Statistics of the annual migration rates of sand dunes in the three studied sand accumulations.
Correlation Pairs of Annual
Rates
Statistics of
Annual Rates
Dune Areas
SA1 SA2 SA3
Temporal Ranges
and Means
2011–2012
Min 0.0 0.0 0.0 0.0
Max 17.0 26.5 24.3 22.6
Average 8.1 9.8 5.0 7.6
2012–2013
Min 0.0 0.0 0.0 0.0
Max 14.8 23.0 24.4 20.7
Average 6.1 7.2 4.6 6.0
2013–2014
Min 0.0 0.0 0.0 0.0
Max 22.0 37.8 38.8 32.9
Average 8.2 10.1 6.2 8.2
2014–2015
Min 0.0 0.0 0.0 0.0
Max 28.6 30.9 34.1 31.2
Average 6.5 7.3 6.1 6.6
Accumulative mean (in 4 years) 28.9 34.4 21.6
Spatial ranges and means
Min 0.0 0.0 0.0
Max 20.6 29.6 30.4
Means 7.2 8.6 5.5
Spatial variations in the magnitudes of dune movements are a well-known aspect of sand
accumulations worldwide. They vary spatially from one dune area to another, and even among
individual dunes in the same dune area [80,81]. This is well demonstrated in the study area where SA2
is characterized by the highest values of the mean annual rates of dune movements (Table), ranging
from 7.2 to 10.1 m/year, with an average of 8.6 m/year. The second-highest values of the mean annual
rates of dune movements were recorded in SA1, ranging from 6.1 to 8.2 m/year, with an average of
7.2 m/year. The lowest magnitudes of the annual rates of dune movements were measured in SA3,
ranging from 4.6 to 6.2 m/year, with an average of 5.5 m/year.
Figure
rates of dune movement in the various sand accumulations. Various implications can be outlined
from the analysis of Table. First, the values of the mean annual rates indicate that
sand accumulations in the study area are considered active dune areas of moderate magnitude [82].
Second, although the ranges of the mean annual rates between the three dune areas are limited to a
few meters annually, the accumulated magnitudes of the mean annual rates over the four-year interval
refer to a signicant magnitude of dune movements (Table). The accumulated means at the end
of the four-year interval were 21.6, 28.9, and 34.4 m for SA3, SA1, and SA2, respectively. Third, the
magnitudes of the mean annual rates are characterized by the same spatial pattern annually, where
SA2 is always the highest, SA1 is intermediate, and SA3 is the lowest (Figure). Many factors can
be responsible for the spatial pattern of the magnitude of the annual rates. These factors include the
spatial locations of sand accumulations and their proximity to sand supply areas. Both SA1 and SA2
are located in the southern reaches of the Wadi Fatmah alluvial plain whereas SA3 is located along
Wadi Ash Schumaysi. This exposes SA1 and SA2 of Wadi Fatmah to the predominant northern winds
before SA3 of Wadi Ash Shumaysi. Accordingly, SA1 and favorably SA2 have greater opportunity
to receive a high sand supply than SA3. In addition, the principal sources of sand supply are more

Remote Sens.2019,11, 2824 17 of 22
proximal to SA2 and SA1 than to SA3. These sources include the abandoned agricultural areas that are
directly located in the northern reaches of these sand accumulations, along with the areas of intensive
urban activities and o-road driving. All of these areas represent signicant sources of loose sand to
be transported by northern winds and deposited where various obstacles are available in the areas
of SA1 and SA2. Although SA1 and SA2 have higher magnitudes of dune movements than SA3,
SA2 in particular is considered the highest active dune area in the study area. Its spatial extent has
expanded remarkably from 15 km
2
, as monitored from the Corona image acquired in 1972, to 26 km
2
,
as mapped from the SPOT 5 image acquired in 2013. The vast spatial expansion of SA2 throughout
these few decades strongly implies persistent and potential sand transportation rates [83] that have
eventually resulted in a high magnitude of dune movement [1]. Fourth, along with the spatial pattern
variations in the magnitude of the annual dune movements, the magnitudes of the mean annual
rates show a temporal pattern of dune movements through all dune areas (Figure). The temporal
variations of the mean annual rates of the three dune areas show consistent patterns, which uctuate
between high and low values through the time interval of the study (Figure). The consistency in the
uctuation of temporal magnitudes of dune movements could potentially point to consistent temporal
changes in wind intensity on an annual basis. However, the degree of temporal uctuation in the
annual dune migrations of both SA1 and SA2 is higher than that of SA3 (Figure), which could be
attributed to variations in the local conditions of wind systems [84–86]. Finally, although the degree
of temporal uctuation and mean annual rates of dune movements are generally low for SA3 in the
rst two sequential years, the last two years have higher magnitudes and almost the same degree of
uctuation in the annual rates of dune movements. The relatively high magnitudes of the annual
rates of dune movements in SA3 in the last two sequential years could be linked to the accelerated
excavation processes in the mountain blocks located in the northwest corner of Wadi Ash Shumaysi.Remote Sens. 2019, 11, x FOR PEER REVIEW 19 of 24

movements than SA3, SA2 in particular is considered the highest active dune area in the study area.
Its spatial extent has expanded remarkably from 15 km
2
, as monitored from the Corona image
acquired in 1972, to 26 km
2
, as mapped from the SPOT 5 image acquired in 2013. The vast spatial
expansion of SA2 throughout these few decades strongly implies persistent and potential sand
transportation rates [83] that have eventually resulted in a high magnitude of dune movement [1].
Fourth, along with the spatial pattern variations in the magnitude of the annual dune movements,
the magnitudes of the mean annual rates show a temporal pattern of dune movements through all
dune areas (Figure 12). The temporal variations of the mean annual rates of the three dune areas show
consistent patterns, which fluctuate between high and low values through the time interval of the
study (Figure 12). The consistency in the fluctuation of temporal magnitudes of dune movements
could potentially point to consistent temporal changes in wind intensity on an annual basis.
However, the degree of temporal fluctuation in the annual dune migrations of both SA1 and SA2 is
higher than that of SA3 (Figure 12), which could be attributed to variations in the local conditions of
wind systems [84–86]. Finally, although the degree of temporal fluctuation and mean annual rates of
dune movements are generally low for SA3 in the first two sequential years, the last two years have
higher magnitudes and almost the same degree of fluctuation in the annual rates of dune movements.
The relatively high magnitudes of the annual rates of dune movements in SA3 in the last two
sequential years could be linked to the accelerated excavation processes in the mountain blocks
located in the northwest corner of Wadi Ash Shumaysi.

Figure 12. The spatio-temporal variation in the mean annual rates of dune migration of the three sand
accumulations in the study area.
6. Conclusions
The previous literature and long archived remote sensing images indicate that the study area in
the middle zone of Wadi Fatmah was once a dominant agricultural rural community. There was
enough water for agriculture to flourish. Soil moisture ensured widespread bushes in the southern
bank of the trunk valley of Wadi Fatmah drainage basin. Sand accumulations were limited to areas
adjacent to the western mountain block and slowly interfered with bushes westward. Earth
observation satellites allowed us to retrieve the spatial features of the landscapes in the study area,
and to analyze their spatial relationships. The unmanaged water discharge and water harvesting
strategy adopted in the study area along with unplanned agricultural practices represented the main
driving forces that resulted in extensive changes in the landscape of the study area. Over the last five
decades, surface runoff from the upstream basins of Wadi Fatmah has been controlled by Wadi
Fatmah Dam, depriving the alluvial deposits in the middle zone where the study area is located from
Figure 12.
The spatio-temporal variation in the mean annual rates of dune migration of the three sand
accumulations in the study area.
6. Conclusions
The previous literature and long archived remote sensing images indicate that the study area in the
middle zone of Wadi Fatmah was once a dominant agricultural rural community. There was enough
water for agriculture to ourish. Soil moisture ensured widespread bushes in the southern bank of the
trunk valley of Wadi Fatmah drainage basin. Sand accumulations were limited to areas adjacent to the
western mountain block and slowly interfered with bushes westward. Earth observation satellites

Remote Sens.2019,11, 2824 18 of 22
allowed us to retrieve the spatial features of the landscapes in the study area, and to analyze their spatial
relationships. The unmanaged water discharge and water harvesting strategy adopted in the study
area along with unplanned agricultural practices represented the main driving forces that resulted
in extensive changes in the landscape of the study area. Over the last ve decades, surface runo
from the upstream basins of Wadi Fatmah has been controlled by Wadi Fatmah Dam, depriving the
alluvial deposits in the middle zone where the study area is located from intermittent water recharge.
Extensive pumping of the shallow alluvial aquifers in the area to meet the water requirements of the
adjacent big cities has maximized water deciency as well. Under such circumstances, along with the
severe arid conditions, agricultural areas have been abandoned, allowing intensive wind erosion of the
soil to occur. These processes have increased sand supply to sand accumulations in the study area.
The conversion of the landscape from simple rural agricultural land into a more urbanized landscape
has allowed more o-road driving and excavation activities to occur over the alluvial plains and
bajada surfaces as well as at the top of the many mountain blocks adjacent to the alluvial plains. These
activities have created loose sands capable of being transported by the wind into sand accumulations
in the study area, which has resulted in e the spatial extents of these accumulations being enlarged and
enabled signicant rates of dune migration. These processes indicate that unmanaged and unplanned
anthropogenic interferences have caused the environmental system in the study area to be transformed
from one state to another. The massive retreat of vegetation cover associated with extensive expansion
of sand accumulations that characterize the recent state of the environmental system indicate that the
study area may undergo desertication processes. The results of this current research are useful in
supporting decision-making processes that may be interested in combating desertication processes
and conserving the natural resources in the study area.
Author Contributions:
Conceptualization, E.H.; methodology, E.H., O.A., A.A. and A.J.N.; software, E.H. and
A.J.N.; validation, E.H., O.A., A.A. and A.J.N.; formal analysis, E.H.; investigation, E.H.; resources, E.H., O.A., A.A.,
A.J.N., K.A.-G., M.A.-M. and A.F.; data curation, E.H., A.J.N. and K.A.-G.; writing—original draft preparation,
E.H.; writing—review and editing, E.H.; visualization, E.H.
Funding:This research received no external funding.
Acknowledgments:
Authors would like to thank the King Abdulaziz City for Science and Technology (KACST)
for providing SPOT-5 Data.
Conicts of Interest:The authors declare no conict of interest.
References
1.
Lancaster, N. Dune Morphology and Dynamics. InGeomorphology of Desert Environments; Parsons, A.J.,
Abrahams, A.D., Eds.; Springer: Dordrecht, The Netherlands, 2009; pp. 557–595. [CrossRef]
2. Coastal Dunes: Form and Process; Wiley: New York, NY, USA, 1990.
3. Sediment. Geol.1973,10, 77–106. [CrossRef]
4.
Fryberger, S.G.; Ahlbrandt, T.S.; Andrews, S. Origin, sedimentary features, and signicance of low-angle
eolian “sand sheet” deposits, Great Sand Dunes National Monument and vicinity, Colorado.J. Sediment. Res.
1979,49, 733–746. [CrossRef]
5.
Whitney, J.W.Erosional History and Surcial Geology of Western Saudi Arabia. Technical Report USGS-TR-04-1;
Ministry of Petroleum and Mineral Resources: Jeddah, Saudi Arabia, 1983.
6.
Al-Harthi, A. Geohazard assessment of sand dunes between Jeddah and Al-Lith, western Saudi Arabia.
Environ. Geol.2002,42, 360–369. [CrossRef]
7.
White, B.R.; Tsoar, H. Slope eect on saltation over a climbing sand dune.Geomorphology1998,22, 159–180.
[CrossRef]
8.
Bauer, B.O.; Davidson-Arnott, R.G.D. A general framework for modeling sediment supply to coastal dunes
including wind angle, beach geometry, and fetch eects.Geomorphology2003,49, 89–108. [CrossRef]
9.
Iversen, J.D.; Rasmussen, K.R. The eect of wind speed and bed slope on sand transport.Sedimentology1999,
46, 723–731. [CrossRef]

Remote Sens.2019,11, 2824 19 of 22
10.
Qian, G.; Dong, Z.; Luo, W.; Lu, J. Mean airow patterns upwind of topographic obstacles and their
implications for the formation of echo dunes: A wind tunnel simulation of the eects of windward slope.
J. Geophys. Res. Earth Surf.2011,116. [CrossRef]
11. Aeolian Geomorphology: An Introduction; Longman: London, UK, 1996.
12.
Ren, J.; Zhang, S.; Meigs, A.J.; Yeats, R.S.; Ding, R.; Shen, X. Tectonic controls for transverse drainage and
timing of the Xin-Ding paleolake breach in the upper reach of the Hutuo River, north China.Geomorphology
2014,206, 452–467. [CrossRef]
13.
Lorenz, R.D.; Zimbelman, J.R.Dune Worlds: How Windblown Sand Shapes Planetary Landscapes; Springer
Science & Business Media: New York, NY, USA, 2014.
14.
Muhs, D.R. Evaluation of simple geochemical indicators of aeolian sand provenance: Late Quaternary dune
elds of North America revisited.Quat. Sci. Rev.2017,171, 260–296. [CrossRef]
15.
Al-Naji, A.J.; Al-Shammery, T.A. Mineralogical study of Sand Dunes Fields in Najaf Governorate, Southern
Iraq.Iraqi J. Sci.2019,60, 239–254.
16.
Lancaster, N.; Baas, A. Inuence of vegetation cover on sand transport by wind: Field studies at Owens
Lake, California.Earth Surf. Process. Landf.1998,23, 69–82. [CrossRef]
17.
Tsoar, H. Sand dunes mobility and stability in relation to climate.Phys. A Stat. Mech. Appl.2005,357, 50–56.
[CrossRef]
18.
Forman, S.L.; Oglesby, R.; Webb, R.S. Temporal and spatial patterns of Holocene dune activity on the Great
Plains of North America: Megadroughts and climate links.Glob. Planet. Chang.2001,29, 1–29. [CrossRef]
19.
Amir, R.; Kinast, S.; Tsoar, H.; Yizhaq, H.; Zaady, E.; Ashkenazy, Y. The eect of wind and precipitation on
vegetation and biogenic crust covers in the Sde-Hallamish sand dunes.J. Geophys. Res. Earth Surf.2014,119,
437–450. [CrossRef]
20.
Wu, J.; Loucks, O.L. From Balance of Nature to Hierarchical Patch Dynamics: A Paradigm Shift in Ecology.
Q. Rev. Biol.1995,70, 439–466. [CrossRef]
21.
Seifan, M. Long-term eects of anthropogenic activities on semi-arid sand dunes.J. Arid Environ.2009,73,
332–337. [CrossRef]
22.
Tsoar, H. The ecological background, deterioration and reclamation of desert dune sand.Agric. Ecosyst.
Environ.1990,33, 147–170. [CrossRef]
23.
Schlesinger, W.H.; Reynolds, J.F.; Cunningham, G.L.; Huenneke, L.F.; Jarrell, W.M.; Virginia, R.A.;
Whitford, W.G. Biological feedbacks in global desertication.Science1990,247, 1043–1048. [CrossRef]
24.
Briske, D.D.; Fuhlendorf, S.D.; Smeins, F.E. Vegetation dynamics on rangelands: A critique of the current
paradigms.J. Appl. Ecol.2003,40, 601–614. [CrossRef]
25.
Mar½n, L.; Forman, S.L.; Valdez, A.; Bunch, F. Twentieth century dune migration at the Great Sand Dunes
National Park and Preserve, Colorado, relation to drought variability.Geomorphology2005,70, 163–183.
[CrossRef]
26.
Danielsen, R.; Castilho, A.M.; Dinis, P.A.; Almeida, A.C.; Callapez, P.M. Holocene interplay between a dune
eld and coastal lakes in the Quiaios—Tocha region, central littoral Portugal.Holocene2012,22, 383–395.
[CrossRef]
27.
Al-Hurban, A.E. Eects of recent anthropogenic activities on the surface deposits of Kuwait.Arab. J. Geosci.
2014,7, 665–691. [CrossRef]
28.
Cabrera-Vega, L.L.; Cruz-Avero, N.; Hern¡ndez-Calvento, L.; Hern¡ndez-Cordero, A.I.; Fern¡ndez-Cabrera, E.
Morphological changes in dunes as an indicator of anthropogenic interferences in arid dune elds.J. Coast.
Res.2013,65, 1271–1276. [CrossRef]
29.
Roskin, J.; Sivan, D.; Shtienberg, G.; Roskin, E.; Porat, N.; Bookman, R. Natural and human controls of the
Holocene evolution of the beach, aeolian sand and dunes of Caesarea (Israel).Aeolian Res.2015,19, 65–85.
[CrossRef]
30.
McKee, E.D. Introduction to a study of global sand seas. InA Study of Global Sand Seas; U.S. Geological
Survey Professional Paper: Reston, VA, USA, 1979; Volume 1052, pp. 1–19.
31.
Breed, C.S.; Fryberger, S.G.; Andrews, S.; McCauley, C.; Lennartz, F.; Gebel, D.; Horstman, K. Regional
studies of sand seas using Landsat (ERTS) imagery. InA Study of Global Sand Seas; U.S. Geological Survey
Professional Paper: Reston, VA, USA, 1979; Volume 1052, pp. 305–397.
32.
Ewing, R.C.; Kocurek, G.; Lake, L.W. Pattern analysis of dune-eld parameters.Earth Surf. Process. Landf.
2006,31, 1176–1191. [CrossRef]

Remote Sens.2019,11, 2824 20 of 22
33.
Ewing, R.C.; Kocurek, G. Aeolian dune interactions and dune-eld pattern formation: White Sands Dune
Field, New Mexico.Sedimentology2010,57, 1199–1219. [CrossRef]
34.
Ewing, R.C.; Kocurek, G. Aeolian dune-eld pattern boundary conditions.Geomorphology2010,114, 175–187.
[CrossRef]
35.
Hugenholtz, C.H.; Barchyn, T.E. Spatial analysis of sand dunes with a new global topographic dataset:
New approaches and opportunities.Earth Surf. Process. Landf.2010,35, 986–992. [CrossRef]
36.
Hugenholtz, C.H.; Levin, N.; Barchyn, T.E.; Baddock, M.C. Remote sensing and spatial analysis of aeolian
sand dunes: A review and outlook.Earth Sci. Rev.2012,111, 319–334. [CrossRef]
37.
Blumberg, D.G. Remote Sensing of Desert Dune Forms by Polarimetric Synthetic Aperture Radar (SAR).
Remote Sens. Environ.1998,65, 204–216. [CrossRef]
38.
Levin, N.; Ben-Dor, E.; Karnieli, A. Topographic information of sand dunes as extracted from shading eects
using Landsat images.Remote Sens. Environ.2004,90, 190–209. [CrossRef]
39.
Ojeda, J.; Vallejo, I.; Malvarez, G.C. Morphometric evolution of the active dunes system of the Doñana
National Park, Southern Spain (1977–1999).J. Coast. Res.2005,49, 40–45.
40.
Kocurek, G.A. Desert aeolian systems. InSedimentary Environments: Processes, Facies and Stratigraphy, 3rd ed.;
Reading, H.G., Ed.; Blackwell: Oxford, UK, 1996.
41.
Hugenholtz, C.H.; Wolfe, S.A. Biogeomorphic model of duneeld activation and stabilization on the northern
Great Plains.Geomorphology2005,70, 53–70. [CrossRef]
42. J. Geol.1959,67, 614–647. [CrossRef]
43.
Merk, G.P. Dune Form and Structure at Great Sand Dunes National Monument. Doctoral Dissertation,
University of Colorado, Boulder, CO, USA, 1962.
44.
Long, J.T.; Sharp, R.P. Barchan-dune movement in imperial valley, California.Geol. Soc. Am. Bull.1964,75,
149–156. [CrossRef]
45.
Calkin, P.E.; Rutford, R.H. The sand dunes of Victoria Valley, Antarctica.Geogr. Rev.1974,64, 189–216.
[CrossRef]
46.
Wiegand, J.P. Dune Morphology and Sedimentology at Great Sand Dunes National Monument. Doctoral
Dissertation, Colorado State University, Fort Collins, CO, USA, 1977.
47.
Maxwell, T.A.; Haynes, C.V. Sand sheet dynamics and Quaternary landscape evolution of the Selima Sand
Sheet, southern Egypt.Quat. Sci. Rev.2001,20, 1623–1647. [CrossRef]
48.
Aql, M.T. Movement of sands east Suez Canal and its impact on human activities: A study in applied
geomorphology.Al Ensaniat2002,19, 1–75. (In Arabic)
49.
Mitasova, H.; Overton, M.; Harmon, R.S. Geospatial analysis of a coastal sand dune eld evolution: Jockey's
Ridge, North Carolina.Geomorphology2005,72, 204–221. [CrossRef]
50.
Bourke, M.C.; Ewing, R.C.; Finnegan, D.; McGowan, H.A. Sand dune movement in the Victoria Valley,
Antarctica.Geomorphology2009,109, 148–160. [CrossRef]
51.
Tsoar, H.The Dynamics of Longitudinal Dunes: Final Technical Report; U.S. Army; European Research Oce:
London, UK, 1978; p. 171.
52.
Tsoar, H.; Blumberg, D.G.; Stoler, Y. Elongation and migration of sand dunes.Geomorphology2004,57,
293–302. [CrossRef]
53.
Liu, J.G.; Capes, R.; Haynes, M.; Moore, J.M. ERS SAR multi-temporal coherence image as a tool for sand
desert study (sand movement, sand encroachment and erosion). In Proceedings of the 12th International
Conference and Workshop on Applied Geologic Remote Sensing, Denver, CO, USA, 17–19 November 1997;
pp. I-478–I-485.
54.
Leprince, S.; Barbot, S.; Ayoub, F.; Avouac, J. Automatic and precise orthorectication, coregistration, and
subpixel correlation of satellite images, application to ground deformation measurements.IEEE Trans. Geosci.
Remote Sens.2007,45, 1529–1558. [CrossRef]
55.
Hermas, E.A. Monitoring the spatial occurrences and migration rates of sand dunes around Makkah city
using remote sensing technology.Remote Sens. GIS2015,3, 14–24.
56.
Vermeesch, P.; Drake, N. Remotely sensed dune celerity and sand ux measurements of the world's fastest
barchans (Bod²l², Chad).Geophys. Res. Lett.2008,35. [CrossRef]
57.
Necsoiu, M.; Leprince, S.; Hooper, D.M.; Dinwiddie, C.L.; McGinnis, R.N.; Walter, G.R. Monitoring migration
rates of an active subarctic dune eld using optical imagery.Remote Sens. Environ.2009,113, 2441–2447.
[CrossRef]

Remote Sens.2019,11, 2824 21 of 22
58.
Hermas, E.A.; Leprince, S.; El-Magd, I.A. Retrieving sand dune movements using sub-pixel correlation of
multi-temporal optical remote sensing imagery, northwest Sinai Peninsula, Egypt.Remote Sens. Environ.
2012,121, 51–60. [CrossRef]
59.
Al-Ghamdi, K.A.; Hermas, E.A. Assessment of dune migration hazards against landuse northwest Al-lith
City, Saudi Arabia, using multi-temporal satellite imagery.Arab. J. Geosci.2015,8, 11007–11018. [CrossRef]
60.Al-Mutiry, M.; Hermas, E.A.; Al-Ghamdi, K.A.; Al-Awaji, H. Estimation of dune migration rates north
Riyadh City, KSA, using SPOT 4 panchromatic images.J. Afr. Earth Sci.2016,124, 258–269. [CrossRef]
61.
Hamandawana, H.; Eckardt, F.; Ringrose, S. The use of step-wise density slicing in classifying high-resolution
panchromatic photographs.Int. J. Remote Sens.2006,27, 4923–4942. [CrossRef]
62.
Herrmann, S.M.; Wickhorst, A.J.; Marsh, S.E. Estimation of Tree Cover in an Agricultural Parkland of Senegal
Using Rule-Based Regression Tree Modeling.Remote Sens.2013,5, 4900–4918. [CrossRef]
63.
Ayoub, F.; Leprince, S.; Keene, L.User's Guide to COSI-CORR Co-Registration of Optically Sensed Images and
Correlation; California Institude of Tecnology: Pasadena, CA, USA, 2009; p. 38.
64.
Buades, A.; Coll, B.; Morel, J. Nonlocal Image and Movie Denoising.Int. J. Comput. Vis.2008,76, 123–139.
[CrossRef]
65.
Simons, M.The Changing World Deserts: The Problem of Water in Arid Lands; Oxford University Press: London,
UK, 1967.
66.
Al-Harbi, N.M. Geomorphic Units and Their Relationships with Land Use and Land Cover in the Drainage
Basin of Wadi. Ph.D. Thesis, Umm Al-Qura University, Makkah, Saudi Arabia, 2017. (In Arabic).
67.
Sadhan, A.S. Water Plan for Wadi Fatimah Basin, Saudi Arabia. Doctoral Dissertation, University of Wyoming,
Laramie, WY, USA, 1980.
68.
Alrhiliy, A.A. Climate Characteristics in Makkah Administrative Area. Ph.D. Thesis, Umm Al-Qura
University, Makkah, Saudi Arabia, 2010. (In Arabic).
69.
Al-Mealem, A.A. Geomorphology of Hassan Basin in Yemen. Ph.D. Thesis, Baghdad University, Baghdad,
Iraq, 2004. (In Arabic).
70.
Hamimi, Z.; Matsah, M.; El-Shafei, M.; El-Fakharani, A.; Shujoon, A.; Al-Gabali, M. Wadi Fatima thin-skinned
foreland FAT belt: A post amalgamation marine basin in the Arabian Shield.Open J. Geol.2012,2, 271.
[CrossRef]
71.
Pye, K.; Tsoar, H. Aeolian Bed Forms. InAeolian Sand and Sand Dunes; Springer: Berlin/Heidelberg, Germany,
2009; pp. 175–253. [CrossRef]
72.
AlJumaie, Z.M. Possibilities of Agricultural Development in Wadi Fatmah of Makkah Al-Mukarramah:
An Applied Study on Agricultural Capabilities. Master's Thesis, Umm Al-Qura University, Makkah,
Saudi Arabia, 1990. (In Arabic).
73.
El-Barodi, M.S. Water Balance Estimates of Wadi Fatmah Basin.J. Kuwait Geogr. Soc.1986,88, 5–60. (In Arabic)
74.Schlesinger, W.H.; Fonteyn, P.J.; Reiners, W.A. Eects of Overland Flow on Plant Water Relations, Erosion,
and Soil Water Percolation on a Mojave Desert Landscape.Soil Sci. Soc. Am. J.1989,53, 1567–1572. [CrossRef]
75.Muhs, D.R.; Been, J.; Mahan, S.A.; Burdett, J.; Skipp, G.; Rowland, Z.M.; Staord, T.W., Jr. Holocene eolian
activity in the Minot dune eld, North Dakota.Can. J. Earth Sci.1997,34, 1442–1459. [CrossRef]
76.
Huenneke, L.F.; Anderson, J.P.; Remmenga, M.; Schlesinger, W.H. Desertication alters patterns of
aboveground net primary production in Chihuahuan ecosystems.Glob. Chang. Biol.2002,8, 247–264.
[CrossRef]
77.
Meir, A.; Tsoar, H. International borders and range ecology: The case of Bedouin transborder grazing.
Hum. Ecol.1996,24, 39–64. [CrossRef]
78.
Tsoar, H. Critical Environments: Sand Dunes and Climate Change. InTreatise on Geomorphology, Aeolian
Geomorphology; Shroder, J.F., Lancaster, N., Sherman, D.J., Baas, A.C.W., Eds.; Academic Press: San Diego,
CA, USA, 2013; Volume 11, pp. 414–427.
79.
Necsoiu, M.; Hooper, D.; Roseberry, J. Physical properties of volcanic material (tephra) using visible
near-infrared spectroscopy. In Proceedings of the AGU Fall Meeting Abstracts, San Francisco, CA, USA,
9–13 December 2019.
80.
Ali, A. The Geomorphology of Sand Dunes of South and Southeast Qattara Depression. Ph.D. Thesis,
Ain Shams University, Cairo, Egypt, 1993.
81.
Tsoar, H. Dynamic processes acting on a longitudinal (seif) sand dune.Sedimentology1983,30, 567–578.
[CrossRef]

Remote Sens.2019,11, 2824 22 of 22
82.
Misak, R.F.; El Ghazawy, M.M. Desertication Processes in the Sinai Peninsula, Egypt. In Proceedings of the
International Meeting on Environmental Disasters and Desertication, Palermo, Italy, 16 June 1989.
83.
Scheidt, S.P.; Lancaster, N. The application of COSI-Corr to determine dune system dynamics in the southern
Namib Desert using ASTER data.Earth Surf. Process. Landf.2013,38, 1004–1019. [CrossRef]
84.
Tsoar, H. Desert dunes morphology and dynamics, El Arish, northern Sinai.Z. Geomorphol.1974,20, 41–61.
85.Assal, E.M. Sedimentological Studies on the Quaternary Sand Dunes and Sabkhas. Master's Thesis, Damiatta
Faculty of Science, University of Mansura, Northern Sinai, Egypt, 1999; p. 268.
86. Landscapes and Landforms of Egypt; Springer: Cham, Switzerland, 2018.
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http: //creativecommons.org/licenses/by/4.0/).