See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/322356954

A REVIEW OF RECENT RESEARCH ON SAND DUNES FORMATION

Article · January 2018

CITATION READS

1 2,395

3 authors:

Eugen Dimofte Florin Popescu

Universitatea Dunarea de Jos Galati Universitatea Dunarea de Jos Galati
7 PUBLICATIONS   10 CITATIONS    53 PUBLICATIONS   101 CITATIONS   

SEE PROFILE SEE PROFILE

Ion V. Ion
"Dunarea de Jos" University of Galati
72 PUBLICATIONS   253 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Sustainability of MONITOX interdisciplinary cooperation network for toxics monitoring, protection of public health and environment, and research of advanced
materials - ENVIMATOX View project

Thermochemical conversion of biomass View project

All content following this page was uploaded by Ion V. Ion on 10 January 2018.

The user has requested enhancement of the downloaded file.

 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

A REVIEW OF RECENT RESEARCH ON SAND DUNES

FORMATION
Eugen DIMOFTE, Florin POPESCU, Ion V. ION
”Dunarea de Jos” University of Galati

ABSTRACT: This work aims at analysing the current stadium of research on building-up of the sand dunes and at being the
starting point for a much detailed analysis on the topic. Throughout this essay practical studies are presented on the erosion,
migration and building-up of sand dunes in the Sahara Desert; studies based on on-field measurements. Then there’s a
presentation of an experimental study on sand dunes erosion performed in the aero-dynamic tunnel; a study to determine the
friction speeds on each of the three models used. The practical studies are followed by a numeric modelling study whose
results are close enough to the experimental results so as to be validated. The one before last chapter describes methods and
techniques of stabilizing and anchoring sand dunes, their modality of putting them into practice, as well as the way in which
numeric modelling can be used in the study of mechanically anchoring and biologically stabilizing sand dunes.
KEYWORDS: sand dunes, erosion, deflation, wind tunnel, numerical modelling

1. INTRODUCTION processes of sand dunes erosion, migration, initiation

and their interaction. Such research requires long
According to the United Nation Convention to periods of field study to capture the processes
Combat Desertification (UNCCD), desertification is involved in sand dunes initiation and nucleation.
defined as being “the degradation of soil in the arid, Field research of dune formation is fundamental as it
semiarid and dry areas, as a result of the action of cannot be reproduced under controlled laboratory
different factors, including climatic changes, and of conditions without changing the density of the
human activities”. working fluid [3].
The purpose of this paper is to carry out an Dunes can arise spontaneously by evolving
analysis of recent research on sand dunes formation from proto-dunes [4]. The formation of small
and migration, setting the basis for a future work on a barchans from proto-dunes comprises five phases. In
numerical model to be used in the analysis and the initial phase, sand patches of a few cm high are
prediction of desertification. Relevant are the reviews generated. In the second phase, these sand patches are
performed by Livingstone et al. [1] on the exposed to aeolian waves of sand, which lead to their
geomorphological and numerical studies of the growth. The accumulation of sand continues, so that
formation of sand dunes and those of Durán et al. [2], during the third phase, these patches become 25 to 40
on research to find a model capable of simulating the cm thick, leading to the formation of proto-dunes
evolution of dunes. with a grain slip on their lee slopes. During the fourth
This paper contains five chapters and phase, a small dune of 1 to 1.5 m height with a grain
conclusions and analyses the phenomena both via flow on the lee side appears. The last phase is the
experimental studies as well as via those performed in formation of a crescent-shaped barchan 1 to 2 m high
the experimental aero-dynamic tunnel. Another [5].
important aspect is represented by the study of using Dune movement process is affected by various
numerical modelling within these researches and how sand transport processes, especially at the dune crest
they manage to offer results comparable with those of [6]. After many field measurements it was concluded
the experimental study. that the sand transport flux over the dune crest can be
expressed as an exponential function and a Gaussian
1. FIELD STUDY ON SAND DUNE function. A correlation coefficient analysis of the
EROSION, MIGRATION, AND complex relationship between the sand transport and
FORMATION wind conditions (mean velocity, maximum wind
velocity, shear velocity, potential sand transport, and
wind direction) and the surface properties (roughness
The field studies presented were carried out in
length, fetch length) indicated that the sand transport
the Sahara Desert and are aimed at investigating the
rate over the dune crest is positively correlated with
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

the mean velocity and shear velocity and is negatively decrease of the crest depends on the relative position
correlated with the wind direction [7]. of maximum sand flow. There is a saturation length
Measurements performed in the Badain Jaran which is proportional to the turbulent drag length
Desert in western China demonstrated that in addition [10]:
to average wind parameters, dune height is highly 𝜌𝑠
sensitive to local geology, subsurface characteristics, 𝐿𝑠𝑎𝑡 ≈ 4.4 𝑑 (2)
𝜌𝑓
and topography, and interactions between changing
climate conditions and aeolian and fluvial processes.
The simulation and interpretation of dunes should
take into consideration these additional factors.

1.1. WORKING METHODOLOGY

A relevant field study of the erosion, migration
and initiation of sand dunes was performed by
Hicham [8] in the Sahara Desert. The dune contours
were measured with hand-held, 5 m-spaced GPS
receivers. The shape of the slip face was recorded in
photographs taken from the ground level.
The technique of stakes was used to obtain the
dune profiles. Thus, two lines of stakes were installed
on the windward side, with the first line parallel to the
direction of the prevailing winds and the second
perpendicular to the first one. After installation of
stakes, the distance between stakes along the profile
was measured with an accuracy of 5 mm and the
angle with an accuracy of 0.1°. These sets of Fig. 1. a) Small sand agglomeration; b) semilunar
measurements were combined in the following dune of elementary dimensions; c) and d) longitudinal
equation: profile showing the transformation of the sand
agglomeration into a dune; e) the wind rose for the
𝑝 studied area [8].
𝛿ℎ = 𝛿𝑠 (1)
√1+𝑝 2

where: ρs is the sand density, ρf is the density of the

where: p is the slope angle, δh is the height variation fluid and d is the sand grain diameter. Experimental
and δs is the distance between two consecutive stakes. measurements indicated the saturation length within
Once the profiles were obtained, the the studied area to be Lsat =1.7 m [8].
accumulation or the erosion at each post was There is a minimal wavelength at which the
recorded, the percentage evolution of the profile formation of dunes starts λN ≈ 7 Lsat. Sand
being then analysed by the same Hicham and Douday accumulations with a wavelength larger than λN will
[9]. Two methods have been used for the lead to the formation of sand dunes, whereas those
measurement of patch wavelength. The first consists with a wavelength equal or smaller than λN will be
in using a tape measure for direct measurements and eroded and will disappear [8].
ensures an accuracy of 5 cm. The second is based on Other studies have shown that there are other
aerial photographs allowing (0.2 – 1) m accuracy [9]. conditions which could initiate the process of barchan
Figure 1 shows the transformation of a proto- formation. One of these is represented by the
dune (Fig. 1.a.) with an initial length of 15 m into a existence of topographic barriers which may act as
small barchan dune during a period of five days. The genuine stoppers, reducing the wind velocity and
longitudinal profile of this dune (Fig. 1.c.) recorded leading to big sand pile or cliff top dune being
on the 9th day shows that the maximum thickness of created. Another condition is the existence of small
the dune is 30 cm. A comparison between this profile barchans in interaction which form, by collision, one
and a second one recorded two days after shows the big barchan. Numerical modelling used for studying
erosion of the windward slope and the accumulation dune size and position has shown that the variation in
of sand around the ridge, as illustrated in Fig. 1.d. wind direction and the neighbouring effect play a
Figure 1.e. shows the wind rose for the period in the decisive part in the formation of megabarchan by the
studied area [8]. remote interaction between barchans.
Either erosion or accumulation can occur,
depending on the increase or decrease of the
maximum sand flow. Furthermore, the increase or

2
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

2. WIND TUNNEL STUDY OF SAND 9.1; 9.9 and 10.7 m/s, which led to four different
DUNES EROSION velocity profiles, as illustrated in Fig. 3 [11].
The piles of triangular shape were tested for
Friction velocity is a very important parameter different windward slope angles, namely 10 o, 20o, and
for the understanding of aeolian and soil erosion 32o, the models being named S10, S20, and
phenomena. It is defined as the minimum friction respectively S32. The height of all models is 75 mm
velocity required for the aerodynamic forces to while, according to analysis, the prevailing particles
exceed the wind erosion. To study this parameter, have a diameter of approximately 0.5 mm.
Faria et al. [11] have performed both numerical
modelling and experimental study in a wind tunnel, as
well as through field research.
Soil erosion depends on factors such as soil
humidity, biological crusts covering the ground,
quantity and distribution of vegetation, surface
roughness, but also on the presence of non-erodible
elements [12]. To determine the influence of the sand
dune slope angle on the friction velocity, three
Fig. 3. Incident velocity profile for various
models were studied, basically shaped transverse
undisturbed wind velocities tested [11].
dunes with the windward slopes of 10°, 20° and 32°.
All three models had the crest height H = 75 mm, and
In order to analyse the influence of wind erosion on
the angle of the leeward slopes equal to the angle of
the evolution of sand pile shape, a series of erosion
repose of 32° [12]. Two series of tests have been
tests have been performed by laser measurements.
performed. The first series targeted the measurement
The sand model was exposed to the intended wind
of threshold friction velocity distribution across the
velocity for certain time periods t = 1; 2; 3; 5; 7; 10;
windward slope. During the second series of
15 and 20 minutes. Measurements of the wall shear
experiments the erosion over the longitudinal profiles
stress or friction velocity along the slope of different
of three models of triangular shape at specific time
piles and pressure have been recorded (Fig. 4). It can
intervals was recorded.
be noted that he friction velocity increases
The physical models were placed on the floor of
proportionally to the undisturbed wind speed, which
the working chamber inside the wind tunnel, equally
indicates that the entrainment of particles is enhanced
spaced from its sidewalls (Fig. 2).
at higher wind speeds and the friction velocity
increases along the stoss surface [11].

Fig. 2. Schematic diagram of the testing area of wind

tunnel and the location of model (sizes in mm) [11].

The average velocity profile of the airflow,

measured at the half-width of the empty settling
chamber and at 2.4 m from the nozzle exit is given by
the following equation:

𝑢 𝑧 𝛼
( ) (3) Fig. 4. Experimental distribution of the friction
𝑈0 𝛿
velocity along the three slopes, for different wind
where u (m/s) is the longitudinal wind velocity velocities [11].
component, U0 (m/s) is the reference wind velocity,
and z (m) is the vertical height measured from the 3. NUMERICAL MODELLING OF
ground. SAND DUNES EROSION
According to the given conditions, the boundary layer
thickness is δ = 0.1 m and the constant α is equal to Experimental studies and field observations
0.11. Several wind velocities were used, namely 8.3; provide information which is valid qualitatively to

3
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

describe processes and from a quantitative point of Figure 8 clearly shows that the numerical model
view, only for the studied cases. The results obtained accurately predicts the friction velocity distribution
by these methods cannot be used to simulate the and can be successfully used for future cases in order
evolution of erosion and dune migration in other real to predict the shape variation dynamics of the sand
cases. dunes.
The numerical modelling was performed in The shape at different time steps of the three eroded
Ansys CFX, the discretization grid was generated models and the evolution of the friction velocity
using CFX Mesh and CFX Solver was used to solve obtained through numerical methods can be observed
the proposed cases [11]. The computational domain is in the diagrams in Fig. 9. The agreement between
assumed to be two-dimensional, involving symmetry numerical and experimental results is better for low
conditions (Fig. 6). wind speeds, regardless of the windward slope angle.
This is due to the actual limits of Ansys CFX and
Ansys Fluent in modelling accurately this kind of
multiphasic flow being a problem which can be
improved. However, a lower wind velocity means, in
fact, less erosion, hence an inferior rate of variation of
the windward slope angle. For a given wind velocity,
an increase in the angle of the windward slope causes
the area exposed to erosion to approach the crest,
which leads to a decrease of eroded surface [11].

Fig. 6. The computational domain [11].

A series of numerical simulations were

performed for multiple values of sand-grain
roughness k. Good results compared to experimental
values being obtained for k = d/30. The benchmark
tests performed for the model S32 are presented in
Fig. 7, which shows the predicted values and the wind
tunnel results. it can be noted that the sand grain
roughness has some influence on the numerical
predictions, especially in the region xh/Ls > 0.5 [11].

Fig. 8. Comparison between experimental and

numerical results of friction velocity distribution at
different wind velocities for S20 [11].

The results of numerical modelling indicate the

existence of a large area of recirculation for all
models, as shown in Fig. 9. The friction velocity
within the recirculation bubble is quite low, leading to
almost null local aeolian erosion (Ferreira and
Fig. 7. The influence of roughness (k) on the Oliveira 2009). It was noticed that the location of
distribution of friction velocity for the S32 model at a reattachment point is independent of the wind
wind velocity U0 = 8.3 m/s [11]. velocity. This observation is valid for numerical
simulations that take into account only wind
The numerical model is validated by comparing velocities situated within the limits of transient flow
the experimental data for the friction velocity with the regime. Thus, only the results for a wind velocity U0
results obtained through numerical simulation. Figure = 8.3 m/s are presented in Fig. 9 [11].
8 illustrates the comparison between the experimental The study confirmed the experimental results
and numerical results of the friction velocity from [12], where the length of the recirculation zone
distribution along the windward slope for the S20 increases with the windward slope. For all the piles
model and considering different wind velocities. It the flow separation occurs at the pile crest. It was
can be noticed that the absolute maximum deviation observed that the location of the reattachment point
for the S20 model is situated between 12.7% and did not change with the undisturbed velocity (Fig.
13.8%, while the average one varies from 5.3% to 10).
6.5% for the different wind velocities tested. Similar
results were obtained for different models [11].

4
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

The Navier-Stokes equations for a 3D single

phase flow are given by the following equation:

𝜕𝑢𝑖 𝜕𝑢𝑖 𝜕𝑝 1 𝜕 2 𝑢𝑖 1
+ ∑3𝑘=1 𝑢𝑘 =− + ∑3𝑘=1 + 𝑔𝑖
𝜕𝑡 𝜕𝑥𝑘 𝜕𝑥𝑖 𝑅𝑒 𝜕𝑥𝑘 𝜕𝑥𝑘 𝐹𝑟
(4)
𝜕𝑢
∑3𝑘=1 𝑢𝑘 𝑘 =0
𝜕𝑥𝑘

where: u is the velocity, g is the gravitational force,

Re is the Reynolds number, Fr the Froude number.
NaSt3DGPF was coupled with Exner model,
which arises from the principle of conservation of
mass [20]. It states that the height h of the sediment
surface changes according to the incoming and
outgoing mass rates and is given by the equation [22]:

𝜕ℎ 𝜕𝑞𝑘 (𝜏(𝑢))
= − ∑2𝑘−1 (5)
𝜕𝑡 𝜕𝑥𝑘

where q is mass rate, τ(u) is the shear stress and u is

the fluid velocity.
Fig. 9. Very good agreement between experimental This model shows the morphological evolution
and numerical results in the determination of friction of the dune surface. The height variation of the sand
velocity and the contours of models [11]. dune changes the lower boundary of the fluid domain.
To complete the model, an equation for the sand
suspension load is introduced to model the suspension
in the fluid body due to advection and diffusion [21].
The sand quantity is given by Eq. (6) which gives the
sand mass concentration within the fluid [20]:

𝜕𝑐 𝜕𝑐 𝜕𝑐 𝜕2 𝑐
+ ∑3𝑘=1 𝑢𝑘 + 𝑤𝑔 − 𝐾 ∑3𝑘=1 =0 (6)
𝜕𝑡 𝜕𝑥𝑘 𝜕𝑥2 𝜕𝑥𝑘 𝜕𝑥𝑘

where: c is the mass concentration, K is the diffusion

coefficient, wg an additional velocity for the settling
of particles due to gravitation.
The NaSt3DGPF code package comprises both
the three dimensional fluid model and the sediment
model handling bed load as well as suspension load.
The three equations and, implicitly three models used
Fig. 10. Determination of recirculation areas on the are coupled in each time step calculation. The
leeward slopes for the three models at a wind velocity capability of the full model is illustrated by numerical
U0 = 8.3 m/s [11]. modelling of the evolution of a barchanoid sand dune
from a conic pile of sand initially put in the middle of
In recent studies of Burkow and Griebel [16] the the flow domain. The coupling of the three equations
mathematical model of the flow is given by the and implicitly of the three models used takes place at
Navier Stokes equations but an additional model is every step of time. An example of the practical
used, capable of reproducing the geomorphological efficiency of combining these models is presented by
features resulting from the fluid flow. Both models the numerical modelling of the formation of a barchan
are coupled into a 3D final model which allows dune from an initial set of conical sand disposed at
numerical simulation of sand dune erosion and the centre of the experimental flow domain. Air is the
dispersion. working fluid for which a parabolic profile is applied
A code (NaSt3DGPF), which is under steady as inflow velocity profile. Both types of transport for
development, was created for numerical simulations coarser sand grains and fine particles are taken into
[17], NaSt3DGPF uses the k-ε turbulence model and account. The following parameters were considered
a Smagorinski subgrid scale model [18]. The code is in the numerical simulation: Re=1000, Fr=1 and a
written in C/C++ and is employing domain grain size of 0.00002 m [18]. The critical angle of
decomposition methods [19].
5
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

repose is modelled according to, by a heuristic various time steps [18]. Figure 12 shows vortex
approach [16]. forming around the dune at time step t =210 s. Figure
Figure 11 illustrates the numerical simulation 13 clearly shows the finer sand particles transported
results of the formation process of a typical crescent by air flow as suspension (t=210 s).
shaped barchan dune from the initial pile of sand at

Fig. 11. Transformation of initial sand pile into a typical crescent-shaped barchan dune [22].

6
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

4.1. MECHANICAL
STABILIZATION OF SAND DUNES
Mechanical stabilization of sand dunes is the
used as an initial stage during the sand dune
biological crusting. There are more methods of
achieving the sand dune mechanical stabilization
fixation, the efficiency depending largely on the
particularities of the area where being applied.
Halting or slowing the movement of sand can be
done by protecting fences 1 to 1.5 m high to cause a
build-up of sand, leading to the formation of an
artificial dune [24]. Protection fences could be of a
chess board pattern or fences aimed at preventing
Fig. 12. Vortex forming around the dune [22].
sand from advancing. Checkerboard fences are
generally built of wooden materials (straws, palm-
tree leaves) which can be easily acquired in the
vicinity of protected areas and they show a certain
permeability to wind.

Fig. 14. Fences in the form of chessboard [24].

Fig. 13. Transportation of finer sand particles as The size of rectangular patterns inside the
suspension load [22]. checkerboard largely depends on the land
particularities. According to William and
4. PREVENTION OF DESERT Mohammed [23], a typical checkerboard system
EXPANSION AND SAND DUNES built in Egypt is 50-70 cm high forming 3 x 3 m
FIXATION TECHNIQUES rectangles while, according to Guo et al. [25], the
typical size for a similar system but for different
If 50 years ago desertification was a problem land conditions is (10-15) cm high forming 1x1m
of major importance only in the arid regions of the rectangles. The fore dune fences are similar to the
world, nowadays, more and more countries lose vegetation belts except that they are made of palm
significant agricultural land annually due to soil leaves, wood, fibre glass, concrete, etc. The main
degradation and desertification. Whereas in the past purposes of the fore dune fences are either to reduce
research was limited only to experiments, with the the wind energy or to divert the direction of the
rapid growth of computer technology, it was wind from the area that needs to be protected.
replaced by numerical simulation, allowing good Depending on the position of the fence in
prediction of environmental modifications, the relation to the dominant wind, we can distinguish
experimental tests being used to validate the between two types of fences: vertical fences which
numerical models. aim to reduce the wind energy and provide sand
Keeping under control the sand dune trapping abilities and diverting fences, slanted at an
migration can be done by reducing saltation, which angle of 120o to 140o from the wind direction to
can be reduced by soil stabilization, or reducing the divert sand away from the area that needs to be
wind velocity at ground level. Another sand dune protected (Fig. 15).
migration control method would be modifying the
wind direction to remove and prevent sand deposits.
The techniques based on these principles can be
divided into two complementary categories:
mechanical stabilization and biological setting [23]. Fig. 15. Vertical fence. b) Diverting fence [23].

7
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

The mulch or protective screen technique planting trees and perennial vegetation (Watson
consists of covering the dune uniformly with a 1985). The choice of species to be planted depends
natural or artificial screen to prevent saltation. The on climatic and ecological conditions. These must
protection screen can be made of different materials, be resistant to drought, wind, high temperature
such as straws, branches, stalks, plastic or acrylic variations; they should grow fast and have a high
fibres foil and net. Surfaces can also be covered capacity of proliferation (Mohamed et al. 2010).
with gravel (Fig. 16), whose granulation and
quantity depend on the wind velocity in the
protected area (National Cooperative Highway
Research Program 1973).

Fig. 18. The dynamics of vegetation on stabilized

dunes. Species with smaller contributions than 5%
Fig. 16. Stone protective coating used to stabilize are not shown [32].
sand dunes [27].
The following costs are envisaged for sand
In 1950, the Egyptian Desert Institute proved dune fixation:
that an area of 1600 ha of desert can be stabilized by  primary fixation: 0.65 USD / m on fences;
water wetting [28]. This brought about an increase  seedling production: 0.25 USD / plant;
in the percentage of the fines (silt and clay) particles  planting and watering: 0.35 USD / plant;
between the sand grains, a reduction of the salt  security: 15 USD / ha. year.
content in the soil, which enhanced the growth of Assuming an average fence length of 1000
more vegetation. m/ha, a planting density of 200 plants/ha, a rate of
Brines or saline waters rich in carbonate or 20% of repopulation and ensuring security for 5
sulphate salts such as brines can be a good years, the average cost would be 850 dollars/ha [32].
stabilizing agent. With the extended evaporation, the
deposition of the salts, present in the brines, tends to
cement the sand grains and resist wind erosion [28].
5.3. NUMERICAL SIMULATION
The chemical industry produces a range of RESEARCH IN SAND DUNE
chemicals that can be used as stabilizers to control STABILIZATION AND SETTING
erosion, such as: bitumen, polyelectrolyte, latexes,
etc. Chemical stabilizers produce adhesive bonding As already stated, numerical simulation is
of the sand granules, forming a film on the sand mainly used due to lower costs, faster and more
grains that increases their specific gravity and reliable results compared to field studies on large
consequently decrease the rate of erosion [23]. areas of land during long time periods.
To identify the most reliable and cost-effective
solution for mechanical dunes stabilization when
using checkerboard fences, the size of rectangular
patterns should be tailored to local environmental
conditions. The size of rectangular patterns within
the checkerboard fences can be determined
experimentally by building structures of different
sizes and studying the effects during long periods of
Fig. 17. Sticking of sand particles by chemical time or by numerical modelling which is much
stabilizer [27]. faster and cheaper. Guo et al. [25] used numerical
simulation to determine the optimal configuration
5.2. BIOLOGICAL FIXATION OF and size of the checkerboard fences. By using field
experiments and simulation with a computational
SAND DUNES fluid dynamics model was found that a height of
(10–20) cm has a substantial effect on dune fixation.
After dunes have been mechanically The practicable size of the checkerboard of 1m×1m
stabilized, they can then be permanently fixed by

8
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

has remarkable wind break and dune fixation effects

(Fig. 19).
A numerical simulation was performed to
simulate the distribution of wind velocities in the
different straw checkerboard sizes using Fluent
v.4.5. As in [24] they defined three types of
geometry conditions of 20 cm in height and 1m x
1m, 2m x 2m, 3m x 3 m in size. Simulated
distribution of wind velocity alone the wind contour
in the measuring square was performed for wind
velocities of 3, 6 and 9 m/s [24].

Fig. 20. Coastal sand dunes formation for different

vegetation growth rates, Vυ (m/year): (a) 2.0; (b)
12.0; (c) 24 and (d) 36, and u = 0.38 m/s [33].

6. CONCLUSIONS
This paper reviews recent developments in the
research of sand dune formation, as well as the
methods to control desertification. Hence, both
Fig. 19. Simulated distribution of wind velocity experimental and numerical modelling studies are
alone the wind contour in the measuring square by used, as well as field studies during long periods of
the CFD model. Wind velocity (e) was set at 10 m/s time.
and the size of the checkerboard at 1m x 1m [25] Sand dunes are analysed according to their
shape and structure, then the factors and conditions
The ideal dimensions of the chess board which concur to their initiation, expansion and
patterned system for the conditions of the land for migration are described together with stabilization
which the simulation was performed were methods and techniques. Field studies conducted in
determined as being made up of squares of 1 m x 1 large deserts of the world and wind tunnels
m. The results numerically achieved were also experiments are both essential in understanding the
experimentally confirmed, following the on-field processes involved in sand dunes formation,
studies [25]. migration and fixing. Field studies and wind tunnels
Numerical modelling can be also used to study experiments results are validated against the
the growth of vegetation used to stabilize dunes as computational predictions obtained by
well as the influence of the increasing plants height computational fluid dynamics model developed. The
on the sand accumulation. Luna et al. [33] used comparison between these sets of results reveals
numerical simulation to determine the influence of good agreement.
vegetation development and expansion on coastal Consequently, numerical modelling is a cost-
sand dune migration in the North-East of Brazil. effective, accurate and reliable research method in
Figure 20 shows an example of using numerical the study of sand dunes formation according to land
simulation to estimate coastal sand dune formation particularities and weather conditions.
depending on the vegetation growth rate.
The numerical simulation results given by the 5. BIBLIOGRAPHY
Fluent software were validated based on the
experimental field studies [33]. The advantages of [1] Livingstone I, Wiggs F.S.G, Weaver M.C.,
using numerical simulation in this field are obvious, Geomorphology of desert sand dunes: A review
significantly reducing the research costs and of recent progress, Earth-Science Reviews,
allowing the performance of an even higher number 2007, 80- 3, pp. 239-257.
of simulations, targeting at selecting the most [2] Durán O., Parteli E.J.R., Herrmann H.J., A
efficient practical solution for the analysed area. continuous model for sand dunes: Review, new
developments and application to barchan

9
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

dunes and barchan dune fields, Earth Surf. [15] Parsons D.R., Walker I.J., Wiggs G.F.S.,
Process. Landforms, 2010, 35, pp. 1591–1600. Numerical modelling of flow structures over
[3] Fryberger S.G., Dune forms and wind regime, idealized transverse aeolian dunes of varying
In: A Study of Global Sand Seas, Mc Kee ED geometry, Geomorphology, 2004, 59, pp. 149–
(Ed.), Geological Survey Professional Paper, 164.
U.S. Govt. Printing Office, Washington, 1979, [16] Burkow M., Griebel M., Towards a three-
pp. 137–169. dimensional numerical simulation of fluvial
[4] Kocurek G., Townsley M., Yeh E., Havholm geomorphological processes and the
K., Sweet M.L., Dune and dune-field reconstruction of past event, Zeitschrift für
development on Padre Island, Texas, with Geomorphologie, 2015, 59, pp. 15-32.
implications for interdune deposition and [17] Adelsberger J., Esser P., Griebel M., Groß
water-table-controlled accumulation, Journal S., Klitz M., Rüttgers A., 3D incompressible
of Sedimentary Research, 1992, 62-4, pp. 622 - two-phase flow benchmark computations for
635. rising droplets, Proceedings of the 11th World
[5] Wasson R.J., Hyde R., Factors determining Congress on Computational Mechanics
desert dunes type, Nature, 1983, 304, pp. 337– (WCCM XI), Barcelona, Spain, 2014, pp. 274-
339. 5285.
[6] Zhang Z., Dong Z., Wu G., Field observations [18] Smagorinsky J., General circulation
of sand transport over the crest of a transverse experiments with the primitive equations,
dune in northwestern China Tengger Desert, Montly Weather Review, 1963, 91, pp. 99–164.
Soil and Tillage Research, 2017, 166, pp. 67– [19] Griebel M., Zaspel P., A multi-GPU accelerated
75. solver for the three-dimensional two-phase
[7] Yang X., Scuderi L., Liu T., Paillou Ph., incompressible Navier-Stokes equations,
Hongwei Li., Dong J., Zhu B., Jiang W., Computer Science Research and Development,
Jochems A., Weissmann G., Formation of the 2010, 25, pp. 65-73.
highest sand dunes on Earth, Geomorphology, [20] Exner F., Uber die Wechselwirkung zwischen
2011, 135(1–2), pp. 108-116. Wasser und Geschiebe in Flussen,
[8] Hicham E., Initiation and early development of Sitzungsbericht Akademie der Wisssenschaft
barchans dunes: A case study of the Moroccan Wien, 134, 1925, pp. 165-180.
Atlantic Sahara Desert, Geomorphology, 2012, [21] Malcherek A., Sedimenttransport und
138. pp. 181–188. Morphodynamik. Lecture notes, Institut fur
[9] Hicham E., Douday S., Equilibrium versus Wasserwesen, Universitat der Bundeswehr,
disequilibrium of barchan dunes, Munchen, 2004.
Geomorphology, 2011, 125, pp. 558–568. [22] Burkow M., Numerische Simulation
[10] Claudin P., Andreotti B., A scaling law for strömungsbedingten Sedimenttransports und
aeolian dunes on Mars, Venus, Earth, and for der entstehenden Gerinnebettformen.
subaqueous ripples, Earth and Planetary Diplomarbeit, Institut für Numerische
Science Letters, 2006, 252, pp. 30–44. Simulation, Universität Bonn, 2010.
[11] Faria R., Ferreira A.D., Sismeiro J.L., [23] William M. Shehata, Mohammed H. Al-
Mendes J.C.F., Sousa A.C.M., Wind tunnel Rehaili, A Review on Sand Dune Stabilization,
and computational study of the stoss slope Saudi Geological Survey, Saudi Arabia, 2013.
effect on the aeolian erosion of transverse sand [24] De Boever M., Khlosi M., Delbecque N., De
dunes, Aeolian Research, 2011, 3, pp. 303–314. Pue J., Ryken N., Verdoodt A., Cornelis
[12] Ferreira A.D., Oliveira R.A., Wind erosion of W.M., Gabriels D., Desertification and land
sand placed inside a rectangular box, Journal degradation processes and mitigation,
of Wind Engineering and Industrial UNESCO Chair of Eremology, Ghent
Aerodynamics, 2009, 97, pp. 1–10. University, Belgium, 2013.
[13] White B.R., Laboratory simulation of aeolian [25] Guo Yu Qiu, In-Bok Lee, Hideyuki Shimizu,
sand transport and physical modeling of flow Yong Gao, Guodong Ding, Principles of sand
around dunes, Annals of Arid Zone, 1996 35, dune fixation with straw checkerboard
pp. 187–213. technology and its effects on the environment,
[14] Qian G., Dong Z., Luo W., Wang H., Journal of Arid Environments, 2004, 56, pp.
Variations of horizontal and vertical velocities 449–464.
over two-dimensional transverse dunes: a wind [26] *** US Highway Research Board, Erosion
tunnel simulation of the effect of windward control on highway construction, Highway
slope, Journal of Arid Environments, 2009, 73, Research Board, National Cooperative
pp. 1109–1116. Highway Research Program of highway
Practice No. 18, 1973.
10
 THE ANNALS OF "DUNAREA DE JOS” UNIVERSITY OF GALATI
FASCICLE IV REFRIGERATING TECHNIQUE, INTERNAL COMBUSTION ENGINES,
BOILERS AND TURBINES, ISSN 1221-4558
2015

[27] Vanoni V.A., Sedimentation Engineering, [31] Mohamed M.O., Saleck M.O., Berte Ch.J.,
Manuals and Reports on Engineering Practice, Fighting sand encroachment: lessons from
ASCE, No. 54, 1975. Mauritania, FAO Forestry paper 158, Roma,
[28] *** Egyptian Desert Institute, Sand dunes in 2010.
Egypt, (in Arabic), M. Monir (Ed.), Egyptian [32] Rozé F., Lemauviel S., Sand Dune Restoration
Desert Institute Publication, 1983. in North Brittany, France: A 10-Year
[29] *** King Fahd University of Petroleum and Monitoring Study, Society for Ecological
Minerals (KFUPM), Research Institute, Restoration International, Restoration Ecology,
Laboratory investigation of chemicals for sand 2004, 12-1, pp. 29-35.
stabilization, Technical Report, Project No. [33] Luna M.C.M. de M, Parteli E.J.R., Durán
23046, 1987. O., Herrmann H.J., Model for the genesis of
[30] Watson A., The control of wind-blown sand coastal dune fields with vegetation.
and moving dunes: a review of the methods of Geomorphology, 2011, 129, pp. 215–224.
sand control in deserts, with observations from
Saudi Arabia, Quarterly Journal of Engineering
Geology & Hydrogeology, 1985, 18, pp. 237-
252.

Rezumat: Această lucrare analizează stadiul actual al

cercetărilor privind formarea dunelor de nisip. Sunt
prezentate studii practice privind eroziunea, migrația și
formarea dunelor de nisip în deșertul Sahara; studii
bazate pe măsurători pe teren. Se prezintă, de asemenea,
un studiu experimental privind eroziunea dunelor de
nisip realizat în tunelul aerodinamic; un studiu pentru
determinarea vitezelor de frecare pe fiecare dintre cele
trei modele utilizate. Studiile practice sunt urmate de un
studiu de modelare numerică având rezultate suficient de
apropiate de rezultatele experimentale pentru a fi validat.
Sunt descrise metodele și tehnicile de stabilizare și
ancorare a dunelor de nisip, modalitatea de punere în
practică a acestora, precum și modul în care modelarea
numerică poate fi utilizată în studiul ancorării mecanice
și stabilizării biologice a dunelor de nisip.
Cuvinte cheie: modelare numerică, tunel de vânt, dune
de nisip, eroziune, migrație

11

View publication stats