International Journal of

Environmental Research
and Public Health

Article
Application of a High-Precision Aeolian Sand Collector in Field
Wind and Sand Surveys
Xinchun Liu 1 , Yongde Kang 2, * , Hongna Chen 3 and Hui Lu 4,5

1 Institute of Desert Meteorology, CMA, Taklimakan Desert Meteorology Field Experiment Station of CMA,
Xinjiang Laboratory of Tree Ring Ecology, Key Laboratory of Tree-Ring Physical and Chemical Research of
China Meteorological Administration, Urumqi 830002, China; liuxch@idm.cn
2 State Key Laboratory of Eco-Hydraulics in Northwest Arid Region of China, School of Water Resources and
Hydroelectric Engineering, Xi’an University of Technology, Xi’an 710048, China
3 Urumqi Environmental Monitoring Center, Urumqi 830001, China; liuxinchun2001@163.com
4 Key Laboratory of Ecology of Rare and Endangered Species and Environmental Protection, Guangxi Normal
University, Ministry of Education, Guilin 541004, China; luhui1008@163.com
5 College of Environment and Resources, Guangxi Normal University, Guilin 541004, China
* Correspondence: kyd0115@yeah.net; Tel.: +86-199-4623-5190

Abstract: Sand collectors are important for quantitatively monitoring aeolian sand activities. In
this paper, an automatic high-precision sand collector was designed. Based on the measured data
of aeolian transport performed with a piezoelectric saltation sensor (H11-Sensit) and a 10 m high
meteorological tower, the sampling efficiency of the automatic sand sampler and the horizontal dust
flux of the near surface were analyzed based on observed data. The results were as follows: the best-
fitting function between the number of impacting sand particles and the amount of collected sand was



a linear relationship. The average value of R2 was 0.7702, and the average sand collection efficiency


of the sand collector at a height of 5 cm was 94.3%, indicating good sand collection performance.
Citation: Liu, X.; Kang, Y.; Chen, H.; From all field tests conducted so far, it appeared that a high-precision sand sampler was a useful
Lu, H. Application of a device for making field measurements of horizontal dust fluxes and ascertaining the relationship
High-Precision Aeolian Sand
between transition particles and wind speed. In the future, the equipment costs and wind drive will
Collector in Field Wind and Sand
continue to be optimized.
Surveys. Int. J. Environ. Res. Public
Health 2021, 18, 7393. https://
Keywords: soil erosion; wind erosion; sand collection efficiency; dust horizontal flux
doi.org/10.3390/ijerph18147393

Academic Editor: Alan W. Gertler

Received: 17 June 2021

1. Introduction
Accepted: 5 July 2021 Wind is an important erosion force that shapes the surfaces of Earth, Venus, Mars,
Published: 10 July 2021 and Titan [1,2]. In arid, semi-arid, high latitude, and high-altitude regions on Earth, wind
erosion is usually considered to be the main driving force for soil loss and dust release [3–6].
Publisher’s Note: MDPI stays neutral Quantitative studies of wind-induced sand migration have played an important role in
with regard to jurisdictional claims in revealing the geomorphological processes, and soil losses and wind erosion dust have
published maps and institutional affil- been accurately simulated [7–9]. In some countries, the annual average particulate matter
iations. (e.g., PM2.5 and PM10) mass concentrations are high [10], where the major emission sources
of PM are the degradation of soil and the mismanagement of lands [10].
Many studies have been carried out on horizontal sand dust fluxes, such as in Europe,
China, Canada, Australia, and the United States, especially in arid and semi-arid regions.
Copyright: © 2021 by the authors. On Earth, this occurs mainly in deserts, on beaches, and in other sparsely vegetated
Licensee MDPI, Basel, Switzerland. areas, such as dry lake beds [11–15]. The lift-off velocities of differently sized particles
This article is an open access article obey different distribution functions. The mean particle velocity at different heights also
distributed under the terms and depends on momentum exchange and particle concentration.
conditions of the Creative Commons Because of the difficulty of obtaining relevant measurements, integrated dust samplers
Attribution (CC BY) license (https:// have historically been the most common method for measuring sand dust fluxes in the field
creativecommons.org/licenses/by/ and in laboratory investigations [16–18]. Different researchers have designed various types
4.0/).

Int. J. Environ. Res. Public Health 2021, 18, 7393. https://doi.org/10.3390/ijerph18147393 https://www.mdpi.com/journal/ijerph
Int. J. Environ. Res. Public Health 2021, 18, 7393 2 of 19

of sand samplers [19–22] to study the horizontal and vertical fluxes of dust transport in the
near strata [23–35], and they are widely used to obtain sand transport measurements in field
and in wind tunnel experiments [36–39]. Initial designs were improved by increasing the
sampling efficiency through taking account of the aerodynamics associated with blocking a
portion of the flow [17]. Additional improvements were made by increasing the temporal
resolution through automatic weighing systems [40,41] to retroactively apply a temporal
signature to the mass flux of sand dust [42]. However, the spatial and temporal resolution
of a mass-collecting sand dust traps remain insufficient for capturing most small-scale
aeolian processes. In addition, because sand traps obstruct aerodynamic processes to
varying degrees, the efficiency of sampling sand dust is variable, and differs with height
and wind conditions [43], but is generally around 80% [44]. In short, traditional sediment
collectors (such as the big spring number eight, BSNE, and modified Wilson and Cooke
(MWAC) samplers) cannot automatically collect and weigh sediment flux, nor can wind
erosion sensors (such as the Sensit and Wenglor laser fork).
The aim of this study was to understand the sand dust transport process and sampling
efficiency more fully. Thus, our research group designed an automatic, long-period, and
high-frequency sand sampler. The sand collector designed in this paper has two innova-
tions, focused at overcoming two problems: one is the interference of the sand collector on
the airflow in the measurement section and the other is the exhaust problem inside the sand
collector. The inlet of the sand sampler has been designed as a wedge, and a wind speed
guide groove and airflow guide outlet are also incorporated into the design. A hollow shaft
was designed as a special dustproof system (i.e., with first and second air leakage networks)
so that the airflow could be discharged into the atmosphere before entering the hollow shaft.
Consequently, the influence of the sand collector on the airflow was minimized, which
conformed to the principle of equal dynamic performance of the sand collector design,
which is of great significance to improving the sand collection efficiency. Additionally, the
dust collection of the sand collector realized the real-time dynamic monitoring of wind
erosion by combining dynamic dust collection with static weighing. Through this multi-
stage weighing approach, a historic breakthrough was achieved in terms of measurement
accuracy and resolution, and the measurement range has been greatly improved. Thus, the
high-precision, high-frequency, and long-period automatic monitoring of sand dust was
truly realized, which automatically recorded at high speeds and accurately reflected any
changes of the dust. In addition, the advantages of this sand collector were a low failure
rate, strong maintainability, and a wide range of use. It was not only able to observe the
wind and sediment activities at the height of several centimeters to several meters above
the surface, but was also able to observe these at heights of zero to several hundred meters.
Field experiments were conducted using the sand sampler, existing meteorological
data, and a Sensit wind erosion sensor to reveal the characteristics of the mass flux, mass
concentration, horizontal velocity, and impact number of particles, thus enhancing our
understanding of the characteristics and relationships of dust activities in the area. At the
same time, the results from this paper provided useful input data for theoretical analyses
and numerical model validation.

2. Materials and Methods

2.1. Study Area
The study area is located around a western station used to make wind drift sand flow
observations in Tazhong, China (83◦ 390 E, 38◦ 580 N), in the hinterland of the Taklimakan
Desert (Figure 1). The area has a typical temperate continental climate, with high tem-
peratures in dry summers, including little precipitation, rich sand resources, and scarce
vegetation. It has an average annual temperature of 13.6 ◦ C, temperature extremes of
46.0 ◦ C and −5.0 ◦ C, an average annual precipitation of only 25.9 mm, and an average
annual evaporation capacity of up to 3812.3 mm. The area is also characterized by frequent
blowing sands and sandstorms, with more than 500 annual sand windstorms (6.0 m/s,
measured at 11.4 m above the ground), average wind speeds of 2.5 m/s, a maximum
Int. J. Environ. Res. Public Health 2021, 18, x FOR PEER REVIEW 3 of 20

evaporation capacity of up to 3812.3 mm. The area is also characterized by frequent blow-
Int. J. Environ. Res. Public Health 2021, 18, 7393 3 of 19
ing sands and sandstorms, with more than 500 annual sand windstorms (6.0 m/s, meas-
ured at 11.4 m above the ground), average wind speeds of 2.5 m/s, a maximum instanta-
neous wind speed
instantaneous windofspeed
20.0 m/s, and
of 20.0 a blown-sand
m/s, activityactivity
and a blown-sand index of about
index of 8000.
about The
8000.main
The
wind directions are ENE, EN, and NNE. The frequency of sandstorms and
main wind directions are ENE, EN, and NNE. The frequency of sandstorms and blowing blowing sands
was high
sands wasfrom
highMarch to August.
from March The regional
to August. sandstorms
The regional in this in
sandstorms area feature
this long dura-
area feature long
tions and poor visibilities. Previous studies showed that 91% of the sand grains
durations and poor visibilities. Previous studies showed that 91% of the sand grains at a 5 at
cma
height were in the range of 63–250 μm, with most in the range of 125–250 μm [45,46].
5 cm height were in the range of 63–250 µm, with most in the range of 125–250 µm [45,46].

Figure
Figure 1.
1. Map
Mapof
ofthe
thestudy
studyarea,
area,showing
showingthe
thelocation
locationof
ofthe
theTazhong
Tazhong site
site in
in the
the Taklimakan
Taklimakan Desert
Desert
of
ofXinjiang
Xinjiang Province,
Province, China.
China.

2.2. Construction
2.2. Constructionofofthe
theSand
Sand Collector
Collector
The core
The core structures
structures of
of the
the sand
sand sampler
sampler (Figure
(Figure 2)
2) were
were aa dust
dust collection
collection system
system and
and
a dust measurement system. The dust collection system was installed
a dust measurement system. The dust collection system was installed above the device above the device
enclosure (components
enclosure (components1–6 1–6 inin Figure
Figure 2),
2), which
which mainly
mainly consisted
consisted of
of an
an empennage,
empennage, firstfirst
air leakage
air leakage network,
network, second
second airair leakage
leakage network,
network, sand
sand intake,
intake, hollow
hollow shaft,
shaft, and
and bearing
bearing
dust proof
dust proof seat.
seat. The
The static
static measurement
measurement system
system was
was in
in the
the device
device enclosure,
enclosure, which
which was
was
mainly composed of a small-range, high-precision dust weighting
mainly composed of a small-range, high-precision dust weighting system and a large-system and a large-
scaledust
scale dustweighting
weightingsystem.
system. TheThe small-range,
small-range, high-precision
high-precision dustdust weighting
weighting systemsystem
con-
consisted of an automatic sand release device and a small-scale, high-precision
sisted of an automatic sand release device and a small-scale, high-precision weighing sen- weighing
sensor.
sor. TheThe small-scale,
small-scale, high-precision
high-precision weighting
weighting sensor
sensor waswas mounted
mounted on the
on the inner
inner wallwall
of
of the device enclosure, and the automatic sand release tipping bucket
the device enclosure, and the automatic sand release tipping bucket was related to thewas related to the
small-scale, high-precision weighting sensor. The automatic sand release device consisted
small-scale, high-precision weighting sensor. The automatic sand release device consisted
of a small-capacity container (tipping bucket) and a container support shaft. The small
of a small-capacity container (tipping bucket) and a container support shaft. The small
capacity container and container support shaft were related to a flexible rod, which can
capacity container and container support shaft were related to a flexible rod, which can
rotate around the support shaft. The large-scale dust weighting system consisted of a
rotate around the support shaft. The large-scale dust weighting system consisted of a large
large capacity container and a large-scale weighting sensor. The large capacity container
capacity container and a large-scale weighting sensor. The large capacity container was
was mounted on the second platform. One end of the large-scale weighting sensor was
mounted on the second platform. One end of the large-scale weighting sensor was fixed
fixed under the second platform and the other end was fixed on the inner wall of the
under the second platform and the other end was fixed on the inner wall of the device
device enclosure. Conveniently, the large-capacity container was designed as an inverted
enclosure. Conveniently, the large-capacity container was designed as an inverted trun-
truncated cone-shaped container, which could help concentrate the center of the gravity of
cated cone-shaped container, which could help concentrate the center of the gravity of the
the collected objects to the sensing part of the weighting sensor.
collected objects to the sensing part of the weighting sensor.
To improve the measurement accuracy and range of the sand sampler, graded weight-
ing was adopted. The range of the first-grade weighting container was 0–300 g with an
accuracy of 20 mg, and the range of the second-grade weighting container was 0–10 kg.
Sampling was carried out with the created device for twelve months at the test site, and
automatic observations were realized, where an aspect sensor was used to monitor the
sand amounts arising from each direction. The data acquisition system had a high storage
Int. J. Environ. Res. Public Health 2021, 18, 7393 4 of 19

capacity (2 GB) and high frequency (≥1 Hz), which allowed us to dynamically and compre-
Int. J. Environ. Res. Public Health 2021,hensively record
18, x FOR PEER the processes
REVIEW and dynamic variations of dust emission arising from 4wind
of 20
erosion in real time and weigh the collected dust with high precision over a wide range.

Figure 2. Structure
Structurediagram
diagramofofthe thefully
fullyautomated
automatedhigh-precision
high-precisionsand
sandcollector:
collector:1. empennage;
1. empennage;2. first air leakage
2. first net-
air leakage
work; 3. second
network; air leakage
3. second network;
air leakage 4. sand
network; intakeintake
4. sand valve;valve;
5. hollow shaft; 6.
5. hollow bearing
shaft; dust proof
6. bearing dustseat;
proof7. seat;
small-capacity
7. small-
container;
capacity 8. container
container; support shaft;
8. container 9. large-capacity
support container;container;
shaft; 9. large-capacity 10. small-range, high-precision
10. small-range, weighing weighing
high-precision sensor; 11.sensor;
large-
scale weighing sensor; 12. first triangular plate; 13. second triangular plate; 14. first rectangular plate; 15. second
11. large-scale weighing sensor; 12. first triangular plate; 13. second triangular plate; 14. first rectangular plate; 15. second rectan-
gular plate.
rectangular plate.

To
Duringimprove the the
the tests, measurement
empennageaccuracy kept the andsandrange
intakeof theconsistent
rate sand sampler,with the graded
wind
weighting was adopted. The range of the first-grade weighting container
direction according to the wind force in the wind direction. At this point, the surface was 0–300 g with
an accuracy
sands entered of the
20 mg,
hollow andshaft
the range
through of the
the second-grade
sand intake valve.weighting container gas
Sand-carrying wasflows
0–10
kg. Sampling was carried out with the created device for twelve
could be discharged from the first air and second air leakage networks. Sands then months at the test site,
and automatic sank
automatically observations were realized,container
to the small-capacity where anvia aspect sensor
gravity, andwasthe used
sand to monitor
weighting
the sand
sensor amounts
began to sensearising from each
(accuracy: direction.
20 mg). The data
The weight dataacquisition system had
were subsequently a high
uploaded
storage
to the datacapacity (2 GB) and
acquisition highThen,
system. frequency (≥1 Hz),
the aspect whichuploaded
sensor allowed us thetoaspect
dynamically
data of andthe
comprehensively record the processes and dynamic variations of
acquired sand to the data acquisition system to obtain the instantaneous sand amount in the dust emission arising
from wind erosion
corresponding in realIf time
direction. and weigh
the collected sandsthewere
collected dustg,with
over 240 high precision
the tipping over a
bucket rotated
wide range. to pour the sands into the large-capacity container for weighing (weight
automatically
range:During the tests,
0–10 kg). Not onlythe empennage
was the total kept
sandtheamount
sand intake
in therate
wholeconsistent
processwith the wind
recorded, but
direction
also the dataaccording
measured to thebywind force capacity
the small in the wind direction.
sensor At this point,
were calibrated, the surface
realizing sands
long-term
entered
automatic themonitoring,
hollow shaftand through the sand
reducing intake valve.
the manual Sand-carrying
workload. gas flows could
The data acquisition system be
discharged
was connected from to the first air and second
a GPRS/CDMA air leakage
communication networks.
module Sands supply
and power then automatically
system, and
the GPRS/CDMA
sank communication
to the small-capacity containermodule was wirelessly
via gravity, and the sand connected
weighting with the computer,
sensor began to
with which
sense real-time
(accuracy: 20 mg).online
Theobservations
weight data were were achieved
subsequentlyoffsite or onsite.toThe
uploaded the observation
data acqui-
frequency
sition wasThen,
system. up tothe 10aspect
Hz. All the data
sensor measured
uploaded during
the aspect an of
data experiment,
the acquired including
sand to
wind
the dataspeed, wind direction,
acquisition system tosand transport
obtain aspect, instantaneous
the instantaneous sand amount aspect
in thesand flux, profile
corresponding
sand flux,Ifand
direction. thecumulative
collected sands totalwere
sandover flux,240
were g, uploaded
the tippingtobucket
the data acquisition
rotated system.
automatically
Through the wireless transmission network, the backstage carried
to pour the sands into the large-capacity container for weighing (weight range: 0–10 out real-time remote
kg).
automatic monitoring of the acquired data.
Not only was the total sand amount in the whole process recorded, but also the data meas-
ured by the small capacity sensor were calibrated, realizing long-term automatic monitor-
2.3. Data Collection and Processing
ing, and reducing the manual workload. The data acquisition system was connected to a
Fifteen dust
GPRS/CDMA events from July
communication moduleto August
and power in 2010 weresystem,
supply studied. andBlowing sands and
the GPRS/CDMA
communication module was wirelessly connected with the computer, with which data.
sandstorms are frequent in this area, which facilitated the collection of experimental real-
The automatic
time sand sampler
online observations weredeveloped by [46]or(Figure
achieved offsite onsite. 2)
The was used, which
observation had a sand
frequency was
intake
up to 10 area
Hz.ofAll
5 cmthe×data 2 cm and a collection
measured during an frequency
experiment,of 1 Hz. An H11-B
including windwindspeed, erosion
wind
direction, sand transport aspect, instantaneous aspect sand flux, profile sand flux, and
cumulative total sand flux, were uploaded to the data acquisition system. Through the
wireless transmission network, the backstage carried out real-time remote automatic
monitoring of the acquired data.
Int. J. Environ. Res. Public Health 2021, 18, 7393 5 of 19

sensor (H11-Sensit, American SENSIT Company: Valparaiso, IN, USA) (Figure 3) was used
to record the number of sand particles impacting the sensor every second. The sensor probe
was installed 5 cm above the ground. The sand sampler was mounted on a horizontal sand
surface with a straight-line distance of 50 m from the wind erosion sensor to ensure that
the sand sampler empennage could rotate normally on windy days. The sand intake height
Int. J. Environ. Res. Public Health 2021, 18, x FOR PEER REVIEW 6 of 20
of the sand sampler was the same as the height of the sensor probe. A national standard
test sieve was used for statistics.

Figure3.3.Wind
Figure Winderosion
erosionsensor
sensor(H11-Sensit).
(H11-Sensit).

3. Results and Analysis

The average particle size of sands was measured to be 63–250 µm. Sands with particle
sizes greater than 74 µm accounted
3.1. Analysis of the Correlation forNumber
between the 75.57%,ofand ImpactedthoseParticles
with particle
and the sizes
Amount greater
of
than 50 µm accounted
Sediment Collected for 97.65%. The average of the two was 86.61%. The actual sand
collection multiplied by 86.61% was the effective weight corresponding to the number of
Fifteen typical sand–dust weather events from July to August 2015 were monitored
sand particles impacting the sensor. After the experiment, the SigmaPlot12.5 software was
to analyze the variation trend of the impacting particle number and sediment concentra-
used for data processing and plotting.
tion. It was found that the variation trend was consistent between sand–dust storms and
Yang and He showed that the number of sands particles impinging the sensor (N) and
sand-blown weather, especially on 19 July 2015 (8:35–21:35), 4 August 2015 (11:35–21:15),
the dust horizontal flux (F) have a linear relationship:
5 August 2015 (10:20–20:25), and 7 August 2015 (9:30–21:55) (Figure 4). It was inferred that
the variation trend of the impact particle F =number
0.0512Nand sediment concentration was almost (1)
consistent for any kind of sand–dust weather, indicating that the field sediment collection
performance
At the same of the automatic
time, high-precision
the sand collection of sand
the BSNEcollector sandwas good for
sampler at 5both sand–dust
cm above the
storm weather
ground (y) and and the sand-blown weather.
number of sand particles impacting the sensor (x) also had a linear
relationship:
100,000 y =Number
0.0002x 6(2)
July 9 of sand particles impacting on sensor Sand collection

Equations (1) and (2) were combined to obtain the following equation:
5
80,000 F = 256M (3)
Number of sand particles impacting on sensor

where M is the mass of sand collected within 5 cm above the ground (units: kg), N is the 4
number of sand particles impacting the sensor (in units of 103 ), and F is the horizontal flux
60,000
Sand collection (g)

of dust passing through the 5 cm × 2 cm section (units: kg).

In our experiments, the number of sand particles impacting the sensor and the mass3 of
collected sand were found to have a linear relationship, as found by [47,48] and described
by Equation
40,000 (2). Therefore, the dust horizontal flux in this paper was calculated using
Equation (3). Next, the sand collection efficiency is the ratio of the amount of sand collected
2
by the sand sampler over the actual amount of transported sand. The correlation coefficient
of the two was taken here as the average sand collection efficiency.
20,000
1

0 0
0 20 40 60 80 100
T (5min)
Int. J. Environ. Res. Public Health 2021, 18, 7393 6 of 19

3. Results and Analysis

3.1. Analysis of the Correlation between the Number of Impacted Particles and the Amount of
Sediment Collected
Fifteen typical sand–dust weather events from July to August 2015 were monitored
to analyze the variation trend of the impacting particle number and sediment concentra-
tion. It was found that the variation trend was consistent between sand–dust storms and
sand-blown weather, especially on 19 July 2015 (8:35–21:35), 4 August 2015 (11:35–21:15),
5 August 2015 (10:20–20:25), and 7 August 2015 (9:30–21:55) (Figure 4). It was inferred that
the variation trend of the impact particle number and sediment concentration was almost
consistent for any kind of sand–dust weather, indicating that the field sediment collection
performance of the automatic high-precision sand collector was good for both sand–dust
storm weather and sand-blown weather.
Figure 3. Wind erosion sensor (H11-Sensit).
Figure 5 shows the relationship between the total number of impacting particles and
the total sediment concentration corresponding to the 15 dust weather events. It was found
3. Results and Analysis
that there is a statistically significant linear relationship between the total number of impact
particles and the total sediment concentration, where the correlation coefficient, R, reached
3.1. Analysis of the Correlation between the Number of Impacted Particles and the Amount of
0.9608. If the sand sampler failed for a short while, the linear equation y = 0.0002x was used
Sediment Collected to estimate the sediment concentration at a height of 5 cm in the corresponding time using
the number ofweather
Fifteen typical sand–dust particles impacting
events from the sensor,
July toso August
as to verify the reliability
2015 of the weighing.
were monitored
to analyze the variation Atrendregression analysis was carried out on the treated sediment concentration data and
of the impacting particle number and sediment concentra-
impacting particle number data by fitting an exponential function, logarithmic function,
tion. It was found that thefunction,
power variation and trend
linearwas consistent
function between
to the data. sand–dustresults
The best-fitting storms areand
shown in
sand-blown weather, especially on 19 July 2015 (8:35–21:35), 4 August 2015 (11:35–21:15),
Table 1. From the 15 sets of fitting results, the fitting degrees of the linear function and
5 August 2015 (10:20–20:25), and 7were
power function August
higher2015 (9:30–21:55)
than those (Figurefunction
of the logarithmic 4). It was
andinferred
exponential that
function.
The linear function has 11 times higher results than the
the variation trend of the impact particle number and sediment concentration was almostpower function, and the average
was higher than the power function. Therefore, the linear function could better describe the
consistent for any kind of sand–dust weather, indicating that the field sediment collection
correlation between the number of sand particles impacting and the sediment concentration;
performance of the automatic high-precision
that is, the optimal relationshipsand collector
between was good
the number of sandfor both impacting
particles sand–dust and the
storm weather and sand-blown weather.during each five-minute period was linear.
sediment concentration

100,000 6
July 19（8:35-21:35） Number of sand particles impacting on sensor Sand collection

5
80,000
Number of sand particles impacting on sensor

60,000
Sand collection (g)

40,000
2

20,000
1

0 0
0 20 40 60 80 100
T (5min)

Figure 4. Cont.
blic Health 2021, 18, x FOR PEER REVIEW 7 of 21
Int. J. Environ. Res. Public Health 2021, 18, 7393 7 of 19

14,000 1.0
August 7（9:30-21:55） Number of sand particles impacting on sensor Sand collection

12,000
0.8
Number of sand particles impacting on sensor

10,000

0.6

Sand collection (g)

8,000

6,000
0.4

4,000

0.2
2,000

0 0.0
0 20 40 60 80 100 120 140
T (5min)
20,000 2.0
August 4（11:35-21:15） Number of sand particles impacting on sensor Sand collection

1.8

1.6
N um ber of sand particles im pacting on sensor

15,000
1.4

1.2
Sand collection (g)

10,000 1.0

0.8

0.6
5,000
0.4

0.2

0 0.0
0 20 40 60 80 100 120 140 160 180 200
T (5min)

Figure 4. Cont.
blic Health 2021, 18, x FOR PEER REVIEW 8 of 21
Int. J. Environ. Res. Public Health 2021, 18, 7393 8 of 19

14,000 0.7
August 5（10:20-20:25） Number of sand particles impacting on sensor Sand collection

12,000 0.6
Number of sand particles impacting the sensor

10,000 0.5

8,000 0.4

Sand collection (g)

6,000 0.3

4,000 0.2

2,000 0.1

0 0.0
0 20 40 60 80 100 120 140 160
T (5min)
ublic Health 2021, 18, x FOR PEER REVIEW 9 of 20
Figure 4. Trends
Figurein4.the number
Trends of impact
in the number particles
of impact andand
particles thethe
amount
amountof
ofsediment collected.
sediment collected.

Figure 5 shows the relationship between the total number of impacting particles and
1,000
the total sediment concentration corresponding to the 15 dust weather events. It was
found 900that there is a statistically significant linear relationship between the total number
of impact particles and the total sediment concentration, where the correlation coefficient,
R, reached
800 0.9608. If the sand sampler failed for a short while, the linear equation y =
0.0002x was used to estimate the sediment concentration at a height 1:1 of 5 cm in the corre-
sponding time using the number of particles impacting the sensor, so as to verify the reli-
700
ability of the weighing.
Collected sand (g)

600
1,000 2
y=0.0002x R =0.96
500
900
400
800
300 1:1
700
200
Collected sand (g)

600
100 2
y=0.0002x R =0.96
500
0
400 0 500,000 1,000,000 1,500,000 2,000,000 2,500,000 3,000,000 3,500,000
Number of sand particles impacting on sensor
300
Figure 5.Figure
Correlation between
5. Correlation thethe
between total
totalimpact
impact particle number
particle number andand
totaltotal sediment
sediment concentration.
concentration.

200
A regression analysis was carried out on the treated sediment concentration data and
impacting particle number data by fitting an exponential function, logarithmic function,
100
power function, and linear function to the data. The best-fitting results are shown in Table
Int. J. Environ. Res. Public Health 2021, 18, 7393 9 of 19

Table 1. Fitting analysis of saltation particle numbers and sand collection quantity.

Coefficient of Determination (R2 )

Observation Time Exponential Logarithmic Power Linear
Function Function Function Function
19 (8:35–21:25) 0.45 0.62 0.69 0.6
20 (19:35–21:25) 0.33 0.18 0.26 0.28
July 21 (20:20–22day5:25) 0.32 0.34 0.45 0.60
29 (22:15–30 day 2:15) 0.44 0.3 0.55 0.69
31 (13:10–8.1 day 7:10) 0.34 0.25 0.35 0.49
1 (21:40–23:25) 0.63 0.68 0.75 0.82
4 (11:35–21:15) 0.44 0.31 0.43 0.72
5 (10:20–20:25) 0.32 0.38 0.53 0.78
7 (9:30–21:55) 0.26 0.24 0.25 0.38
8 (10:35–20:25) 0.28 0.48 0.55 0.44
August
9 (13:15–18:05) 0.60 0.31 0.44 0.84
13 (10:30–21:25) 0.33 0.27 0.32 0.63
14 (10:00–21:15) 0.43 0.31 0.64 0.47
18 (10:50–20:50) 0.41 0.21 0.68 0.58
20 (13:40–19:55) 0.54 0.29 0.56 0.75
Average 0.41 0.34 0.50 0.61

According to the linear regression analysis of the number of sand particles impacting
every 5 min and the amount of sand collected during this period, it was found that the aver-
age correlation coefficient R is 0.7702. According to the judgment standard, 0.3 < |R| < 0.5
is referred to as a weak correlation, 0.5 < |R| < 0.8 is referred to as a significant correlation,
and 0.8 < |R| < 1 is referred to as high correlation. Thus, it can be seen that among the
15 fitting results, nine were significantly correlated, and six were highly correlated. This
showed that the linear correlation between the number of sand particles impacting the
sensor and the sediment mass was good.
From an energy point of view, the energy gained by impacting particles during wind-
blown sand events is extremely complex, especially in the field, where the impacting
particles generally acquire mechanical energy and horizontal kinetic energy in the airflow.
According to the conservation of energy theorem, the amount of energy lost by the mass
should yield reliable and accurate experimental results.

3.2. Sand Collection and Sand Collection Efficiency

The number of sand particles impacting the sensor during the 15 dust events and the
theoretical values of the collected sand masses in Table 2 were fitted, and, as before, the
moving particles in the airflow are approximately equal to the energy lost as they impact
the sensor; this energy can be ascertained by monitoring the total energy of the impacting
particles per second at a certain height. In different periods of the 15 dust events, different
wind drift sand flows were observed, which were influenced by the wind speed and sand
source supply condition, which resulted in different fitting functions for the number of
sand particles impacting the sensor and the amount of collected sand. Therefore, two or
more functions should be combined when studying the relationship between the number
of sand particles impacting the sensor, and the collected sand results were well-described
by a linear relationship. The coefficient of determination, R2 , was 0.6 (Figure 6). Therefore,
the theoretical value of the collected sand mass in each dust event could be calculated
according to the fitted linear equation. The calculated results are also shown in Table 2.
The statistical results showed that the theoretical values were smaller than the observed
values. The reason for this may be that the data described by the linear relationship do not
perfectly follow the best-fitting line, but are distributed about it. The field test and wind
tunnel test data were different, so the correlation coefficient was not 1 and it was inevitable
that the theoretical value of sand collection was smaller than the observed value.
Int. J. Environ. Res. Public Health 2021, 18, x FOR PEER REVIEW 11 of 20

Int. J. Environ. Res. Public Health 2021, 18, 7393 10 of 19

Table 2. Observed saltation particle numbers and observed and theoretical sand masses.

Table 2. Observed saltation

Impactparticle
Number numbers and observed and
Observed theoretical sand
Value masses. Value
Theoretical
Observation Time
of Particles of Sediment/g of Sediment Quantity/g
Theoretical Value
Impact Number Observed Value
Observation Time of Sediment
19 (8:35–21:25) 2,936,905 of Particles743.4966
of Sediment/g 587.3810
Quantity/g
20 (19:35–21:25) 3,407,661 154.126 681.5322
19 (8:35–21:25) 2,936,905 743.4966 587.3810
July 21 (20:20–22 day 5:25) 920,676 170.5492 184.1352
20 (19:35–21:25) 3,407,661 154.126 681.5322
29 (22:15–30 day 2:15) July 21 (20:20–22147,786
day 5:25) 920,676 34.6005 170.5492 29.5572
184.1352
31 (13:10–1 day 7:10) 1,050,073
29 (22:15–30 day 2:15) 147,786 179.586 34.6005 210.0146
29.5572
1 (21:40–23:25) 31 (13:10–1 day 7:10)
19,772 1,050,073 4.5764 179.586 210.0146
3.9544
4 (11:35–21:15) 195,696
1 (21:40–23:25) 19,772 26.6128 4.5764 39.1392
3.9544
5 (10:20–20:25) 4 (11:35–21:15)
116,005 195,696 12.3933 26.6128 39.1392
23.201
5 (10:20–20:25) 116,005 12.3933 23.201
7 (9:30–21:55) 231,534 43.2746 46.3068
7 (9:30–21:55) 231,534 43.2746 46.3068
8 (10:35–20:25) 1,048,891
8 (10:35–20:25) 1,048,891 118.5805 118.5805 209.7782
209.7782
August August
9 (13:15–18:05) 1,380,718
9 (13:15–18:05) 1,380,718 354.3815 354.3815 276.1436
276.1436
13 (10:30–21:25) 13 (10:30–21:25)
210,702 210,702 14.2759 14.2759 42.1404
42.1404
14 (10:00–21:15) 390,274 25.0633 78.0548
14 (10:00–21:15) 390,274 25.0633 78.0548
18 (10:50–20:50) 293,093 15.7758 58.6186
18 (10:50–20:50) 293,093
20 (13:40–19:55) 349,610 15.7758 4.5021 58.6186
69.922
20 (13:40–19:55) 349,610 4.5021 69.922

1000

900

800

700

600 1:1
Sediment load (g)

500
2
R =0.6
400

300

200

100

0
0 500,000 1,000,000 1,500,000 2,000,000 2,500,000 3,000,000 3,500,000 4,000,000
Number of sand particles impacting the sensor
Figure 6.
Figure 6. Change of sand collection
collection quantity
quantity with
with saltation
saltation particle
particle numbers.
numbers.

As shown in Figure 7, thethe theoretical

theoretical and
and observed
observed sand
sand masses
masseswere
werefitted
fittedand
andana-
an-
lyzed, and
alyzed, andthe
thetwo
twodid
didnot
notshow
showany
anylinear
linearrelationship, withaaRR22 of
relationship,with of only 0.40, indicating
the large difference between the observed and theoretical values. The reason for this large
discrepancy was
was that
thatthe
thetheoretical
theoreticalvalues
valueswere
werecalculated
calculatedbased
basedonon
thethe
actual amounts
actual amounts of
collected
of sand
collected andand
sand thethe
number of sand
number particles
of sand impacting
particles the sensor.
impacting In addition,
the sensor. the field
In addition, the
 example, one experiment showed that the BSNE sand sampler had a sand collection effi-
ciency of 100%–120% [48], while Xu et al. concluded that their BSNE sand sampler had a
sand collection efficiency of 90% [49]. Although it was impossible to calculate the sand
collection efficiency of each dust event directly based on the observed sand masses and
Int. J. Environ. Res. Public Health 2021,the theoretical values of the sand masses in this work, the average sand collection11effi-
18, 7393 of 19
ciency could be calculated according to the correlation coefficient between the two: the
average sand collection efficiency of the sand sampler at a height of 5 cm was 94.3%.
environment
Therefore, the was complex,
method which
of testing theresulted inof
efficiency errors
sand of the dataisacquisition
samplers system
a problem worth in the
further
transmission process, which reduced the final mass accuracies.
studying.

1000
950
900
850
800
750
Observed value of sand collection (g)

700
650
600
1:1
550
500
450 2
R =0.40
400
350
300
250
200
150
100
50
0
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000
Theoretical value of sand collection(g)
Figure7.7.Relationship
Figure Relationshipbetween
betweentheoretical
theoreticaland
andobserved
observedsand
sandmasses.
masses.

3.3. Sand Dust Horizontal

A similar study found Fluxthat
Calculation
their active sand sampler had a higher sand collection
efficiency
The dust horizontal fluxes weresampler
than their passive sand calculatedin wind tunnel experiments,
with Equation and found
(3). The variation that
of particle
the sand collection efficiency increased with an increase of grain size [49]. He
size and wind speed likely affect the distribution of the dust horizontal fluxes. When theet al. found
that the
sand sand size
particle collection efficiency
was large, of their active
the transition heightsand
wassampler wasand
also high, 105%
thewith the error of
corresponding
±
sand collection flux was also large. It can be seen from Figure 8 that during the85%,
5% [46]. The sand collection efficiency of their passive Leach sand sampler was and
15 dust
the collection efficiency for sands with particle sizes smaller than 10 µm was 70%. Their
Fryrear BSNE sand sampler had a 90% ± 5% collection efficiency for aeolian sands and a
40% collection efficiency for sands with particle sizes smaller than 10 µm. Goossens et al.
found that their same sand sampler had a higher efficiency for collecting aeolian sands
than for collecting dust [47–49], where the larger the size of the collected particles, the
higher the collection efficiency of the sand sampler. Therefore, even for the same type of
sand samplers, the sand collection efficiency may be different due to different test methods.
For example, one experiment showed that the BSNE sand sampler had a sand collection
efficiency of 100–120% [48], while Xu et al. concluded that their BSNE sand sampler had
a sand collection efficiency of 90% [49]. Although it was impossible to calculate the sand
collection efficiency of each dust event directly based on the observed sand masses and the
theoretical values of the sand masses in this work, the average sand collection efficiency
could be calculated according to the correlation coefficient between the two: the average
sand collection efficiency of the sand sampler at a height of 5 cm was 94.3%. Therefore, the
method of testing the efficiency of sand samplers is a problem worth further studying.

3.3. Sand Dust Horizontal Flux Calculation

The dust horizontal fluxes were calculated with Equation (3). The variation of particle
size and wind speed likely affect the distribution of the dust horizontal fluxes. When the
sand particle size was large, the transition height was also high, and the corresponding
sand collection flux was also large. It can be seen from Figure 8 that during the 15 dust
Int. J. Environ. Res. Public Health 2021, 18, x FOR PEER REVIEW 13 of 20
Int. J. Environ. Res. Public Health 2021, 18, 7393 12 of 19

events,
events,thethemaximum
maximum dust
dusthorizontal
horizontal flux is is
flux 190.335
190.335kg,kg,
which
whichwaswasmeasured
measured during
during a a
sandstorm. The other events were all blowing sand events, with
sandstorm. The other events were all blowing sand events, with a minimum value of a minimum value of
about
about1.21.2
kg.kg.
AsAsthese experiments
these experiments were
wereconducted
conducted at at
thethe
same
samesampling
sampling site, and
site, andover a a
over
continuous
continuous sampling
sampling period, thethe
period, impact
impact of of
particle
particlesize was
size wasnot taken
not into
taken intoconsideration
consideration
in in
ourour
study; instead,
study; onlyonly
instead, the impact of wind
the impact speed speed
of wind was considered. Unfortunately,
was considered. more
Unfortunately,
moreinformation
useful useful information was lost
was lost after after speeds
the wind the wind werespeeds were Therefore,
averaged. averaged. theTherefore,
transportthe
of transport
most dustof most dust
materials materials
happened in happened
the surfacein the and
layer surface
theylayer
landedandon
they
the landed
Earth’s onsur-the
face by short-distance transportation. Even during a sandstorm, only a small amountsmall
Earth’s surface by short-distance transportation. Even during a sandstorm, only a of
amount
dust of dust wasat
was transported transported at higher
higher altitudes. altitudes.
Thus, Thus, whenthe
when calculating calculating
horizontal thedust
horizontal
flux,
itsdust flux, its
functional functional with
relationship relationship with height
height should should beInestablished.
be established. In relationship
this study, the this study, the
relationship
between between
the collected the within
sand collected sand within
a height of 5 cmafromheighttheofground
5 cm from thesand
by the ground by the
sampler
andsand
thesampler
horizontal anddust
the flux
horizontal dustspecifically,
was used; flux was used; the specifically,
former was the former
used was used
to calculate the to
latter using Equation (3) via a simplified calculation method that reduced the computa-the
calculate the latter using Equation (3) via a simplified calculation method that reduced
computational
tional requirements. requirements.

300

280

260

240

220
horizontal sand-dust flux(kg)

200

180

160

140

120

100

80

60

40

20

0
8) 6) 4) 0) 6)
26） 3:0 9:5 1:2 9:2 6:4
2–18: 15–2 01–1 53–2 50–1 10–1
4 : : : : :
14: (20 (12 (9 (15 (13
21 t4 st 7 t8 t9
ly 19（ J uly u gus A ugu u gus u gus
Ju A A A

Figure 8. Horizontal sand–dust fluxes on selected dates.

Figure 8. Horizontal sand–dust fluxes on selected dates.
The dust horizontal fluxes were calculated with Equation (3). The variation of particle
The
size anddust horizontal
wind fluxesaffect
speed likely werethecalculated with of
distribution Equation
the dust(3). The variation
horizontal of When
fluxes. particlethe
size and wind speed likely affect the distribution of the dust horizontal fluxes.
sand particle size was large, the transition height was also high, and the corresponding When the
sand
sandparticle size flux
collection was was
large, thelarge.
also transition
It canheight was
be seen alsoFigure
from high, 8and
thatthe corresponding
during the 15 dust
sand
events, the maximum dust horizontal flux is 190.335 kg, which was measured15
collection flux was also large. It can be seen from Figure 8 that during the dust a
during
events,
sandstorm. The other events were all blowing sand events, with a minimum valuea of
the maximum dust horizontal flux is 190.335 kg, which was measured during
sandstorm. TheAs
about 1.2 kg. other events
these were allwere
experiments blowing sand at
conducted events, withsampling
the same a minimum value
site, and of a
over
about
continuous sampling period, the impact of particle size was not taken into considerationa in
1.2 kg. As these experiments were conducted at the same sampling site, and over
continuous
our study;sampling
instead, period,
only thethe impact
impact of of particle
wind speedsize
was was not taken Unfortunately,
considered. into considerationmore
in useful
our study; instead, was
information onlylost
the after
impact theofwind
windspeeds
speed were
was considered. Unfortunately,
averaged. Therefore, more
the transport
useful information
of most was lost
dust materials after theinwind
happened speedslayer
the surface wereandaveraged. Therefore,
they landed on thethe transport
Earth’s surface
of by
most dust materials
a short distance. happened in the surface layer and they landed on the Earth’s sur-
face by a short distance.
3.4. Sand Transportation Rate and Wind Speed
3.4. Sand Transportation
Figure 9 shows the Rate and Wind Speed
relationship between the sand transportation rate and wind speed
under different dusty weather conditions; (a) and (b) were strong and typical sandstorm
 Int. J. Environ. Res. Public Health 2021, 18, x FOR PEER REVIEW 14 of 20

Int. J. Environ. Res. Public Health 2021, 18, 7393 13 of 19

Figure 9 shows the relationship between the sand transportation rate and wind speed
under different
weather events,dusty weather and
respectively, conditions;
(c) to (j)(a) andblowing
were (b) were sand
strongweather
and typical sandstorm
events. In general,
weather events, respectively, and (c) to (j) were blowing sand weather events.
the rate of sand–dust transport increased with an increase of wind speed, but the changing In general,
the rateof
trends of the
sand–dust transport
wind speed andincreased
sand–dust with an increase
transport rateofduring
wind speed,
somebut the changing
weather processes
were inconsistent. For example, for some blowing sand events (e, g, i, and processes
trends of the wind speed and sand–dust transport rate during some weather j) and strong
were inconsistent.
sandstorm events For example,
(a and for somethe
b), although blowing sand events
wind speeds were(e, g, i, and
small, j) and strong
the variations of the
sandstorm events (a and b), although the wind speeds were small, the variations of the
sand transportation rate were large. The soil temperature at 0 cm during the correspond-
sand transportation rate were large. The soil temperature at 0 cm during the correspond-
ing events showed that the soil temperature rapidly increased with an increase of solar
ing events showed that the soil temperature rapidly increased with an increase of solar
radiation intensity. In the surface layer, a strong decreasing temperature structure formed
radiation intensity. In the surface layer, a strong decreasing temperature structure formed
in the surface layer, and the thermal turbulence became stronger and stronger, resulting
in the surface layer, and the thermal turbulence became stronger and stronger, resulting
in enhanced heat transfer and unusually active sand saltation. When the wind speed
in enhanced heat transfer and unusually active sand saltation. When the wind speed was
was small, but the sand transportation rate was relatively large, the corresponding soil
small, but the sand transportation rate was relatively large, the corresponding soil tem-
temperature at 0 cm was hot. This showed that the increased soil temperature at 0 cm could
perature at 0 cm was hot. This showed that the increased soil temperature at 0 cm could
promote
promote thethe saltation
saltation of
ofsands.
sands.This
Thisfinding
findingisisconsistent
consistentwith
withthe
theresults of of
results [48], who
[48], found
who
that
found that the “synchronization of wind and temperature” was in favor of the co-occur- of
the “synchronization of wind and temperature” was in favor of the co-occurrence
wind and
rence of saltation.
wind The specific
and saltation. relationship
The specific of the
relationship ofsand transportation
the sand transportation rate with
rate wind
with
speed and temperature needs to be further
wind speed and temperature needs to be further studied.studied.

-1 -1
200 48 times sand collection/g sand transportaion rate/g.cm .s 1.0 54 12
° -1
46 soil temperature of 0 cm/C wind speed of 0.5m /m.s
52
44
42 50 11
40 0.8
150 48
38
36 46 10
34
0.6 44
32
30 42
100 9
28
26 40
0.4
24 38
22 8
20 36
50
18 0.2 34
16 7
14 32
12 30
0 10 0.0 6-1
g -1 -1 °
m.s
A B c d e f g h i j g.cm .s C

weather processes
Figure 9.
Figure 9. Sediment
Sedimentofofdifferent
differentdust
dustevents.
events.

4. Discussion
Discussion
There
There are
aremany
manyfactors affecting
factors thethe
affecting horizontal flux of
horizontal fluxsand dust. In
of sand addition
dust. to dif- to
In addition
ferences in collection
differences instruments,
in collection there
instruments, are are
there many manyother factors,
other especially
factors, in terms
especially of of
in terms
physical mechanisms,
physical mechanisms,whichwhicharearesummarized
summarized inin
thethe
following.
following.
The
The movement
movementofofsand sanddust
dustininresponse
response to to
wind
windis the most
is the important
most importantfeature of of
feature
aeolian sediment transport. Wind accelerates particles and ejects more particles
aeolian sediment transport. Wind accelerates particles and ejects more particles when when im-
pacting a surface,
impacting which
a surface, leadsleads
which to increases in the horizontal
to increases dust fluxdust
in the horizontal in theflux
initial stages
in the initial
stages [50,51]. However, this rapid increase in the impact particle concentration aproduces
[50,51]. However, this rapid increase in the impact particle concentration produces cor-
aresponding increase
corresponding in the in
increase drag
theofdrag
dustofparticles on the fluid,
dust particles on the thereby
fluid,retarding the wind the
thereby retarding
speed. This in turn reduces the speeds of the horizontal dust flux particles, such that a
wind speed. This in turn reduces the speeds of the horizontal dust flux particles, such
that a steady state is reached when the speeds of the impacting particles are reduced to a
value at which there is a single particle leaving the soil surface for each particle impacting
Int. J. Environ. Res. Public Health 2021, 18, 7393 14 of 19

it [52]. Because of the finite response time of dust flux particle speeds to the wind speed,
the horizontal dust flux can “overshoot” the eventual steady-state mass flux [53–55].
Kok, Parteli, and Michaels found that the wind profile was modified through momen-
tum transfer by impacting particles [14]. On the one hand, it is the retardation of the wind
profile through drag by the saltating particles that ultimately limits the number of particles
that can be saltated under given conditions. On the other hand, the drag produced by
saltating particles reduces the horizontal momentum flux carried by the wind. At the same
time, although the particle speeds at the surface were once thought to remain constant,
numerical simulations and wind-tunnel measurements have shown that the particle speeds
increase as a function of the shear velocity above the surface. This means that the mass
flux higher up in the saltation layer increases relative to that in lower layers, producing
a slight increase in the saltation layer height with shear velocity. The extraction of wind
momentum by saltating particles produces a steady-state wind profile that accelerates
saltating particles to a saltate speed that, on average, results in a single particle leaving the
soil bed for each particle impacting upon it.
Rice, Willetts, and McEwan found that, when the surface shear stress fell below the
impact threshold, fewer saltate particles were entrained by the wind. This in turn reduced
the transfer of momentum from the fluid to the impacting particles, thereby increasing the
surface shear stress back to its threshold value. Conversely, when the surface shear stress
exceeded the threshold value, more saltating particles were entrained, again restoring the
surface shear stress to its critical value. Experiments have indicated that particles with
a larger cross-sectional area have a correspondingly higher chance of being “splashed”
by a saltate particle [56]. Saltate particles may have short lifetimes because their high
inertia causes them to attain lower speeds when splashed, leading them to quickly settle
back to the dust surface [57]. Actually, winds cannot directly lift dust particles because
the corresponding interparticle cohesive forces are large compared to aerodynamic forces.
Instead, these dust particles mainly collide with each other under wind erosion [58,59].
During a dust storm, lifted particles are accelerated by the wind, and the resulting impacts
on the dust bed can eject, or splash, new erosion particles into the fluid stream. This
process produces an exponential increase in the particle concentration [60], which leads to
increasing drag on the wind, thereby retarding the wind speed in the erosion layer [61]. It
is this slowing of the wind that acts as a negative feedback process that reduces particle
speeds, which ultimately limits the number of impacting particles [62].
Dust flux is predominantly emitted by the saltates of eroding particles at the near
surface [58,59]. Several authors have also argued that mid-air collisions can affect particle
trajectories and the mass flux at large shear velocities [63–66]. Wind erosion produced
by colliding particles during a single hop is of the order of 10–50%, which increases with
wind speed [66]. That is, even at large impact speeds, not all saltating particles rebound
from the surface, because some will dig into the particle bed [67]. Moreover, recent mea-
surements indicated that the restitution coefficient is a declining function of the saltator
impact speed [68]. Numerical simulations [50,53,69,70], laboratory experiments [71], and
theory [57] have all indicated that the number of splashed particles scales with the impact-
ing momentum. Experiments have also indicated that the fraction of the average impacting
momentum spent on splashing surface particles is of the order of 15% for a bed of loose
sand particles [56,57,59]. Most splashed particles, therefore, move in low-energy reptation
trajectories and quickly settle back to the soil bed. However, some splashed particles do
gain enough momentum from the wind to participate in saltation and splash up more
particles. This multiplication process produces a rapid increase in the particle concentration
upon initiation of saltation, where the distribution of splashed speeds at a given ejected
velocity follows either an exponential or a lognormal distribution [72,73].
Sand transport has been measured by many instruments developed mainly to deter-
mine the rates of transported material. Accurately measuring these processes has been an
ongoing challenge since the first known discrete measurements were made by Bagnold.
In general, impact sensors suffer from poor sensitivity to small sand grains [74], and in
Int. J. Environ. Res. Public Health 2021, 18, 7393 15 of 19

the case of Safire, poor interinstrument repeatability [75]. Over the past decade, although
instruments vary in design and complexity, they can be split into two categories: inte-
grating and real-time electronic instruments. As achieved in this study, there have been
incremental improvements in this latter category of devices that have been motivated, in
part, by the observation that sediment transport occurred on spatial scales smaller than
5 cm and temporal scales shorter than 1 s. Initial designs were improved by increasing
the efficiency by taking account of the aerodynamics associated with blocking a portion of
the flow [13]. Additional improvements were made by increasing the temporal resolution
through automatic weighing systems [27,28]. However, the spatial and temporal resolu-
tions of a mass-collecting sediment trap remain insufficient for capturing most small-scale
aeolian processes. In addition, because sand traps obstruct the flow to varying degrees,
the efficiency of sampling saltating grains varies with height and wind conditions [17] but
is generally around 80% [44]. Sand particle sizes not only differ regionally [76], but also
vary significantly with height [77–79]. This variation of sand particle size with location and
height inevitably influences the sampling efficiency of different samplers, leading to errors
in the amount of aeolian sand collected. However, in previous experiments, this has often
been neglected, or a fixed sampling efficiency has been applied [48,80–82].
Some research has shown that an exchange of sand is expected between adjacent
dunes of different heights, since larger dunes are more likely to disperse sand than smaller
ones [83]. Changes in individual dunes relative to sand flux are unstable, meaning that in
one area, dunes that evolve only through sand flux should eventually merge into larger
dunes [84]. It is well-known that dune patterns are mainly determined by the amount of
available sand and the behavior of the wind in the year. At the same time, changes in the
wind and its direction may destroy the stability of large dunes, thereby affecting sediment
flux [85]. However, wind erosion processes are often random or strongly dependent on
local environmental conditions, making them difficult to incorporate into general analyses
or simulations of dune fields. Additionally, collisions between sand particles can adjust the
dune shape and the distance between them. When sand particles collide with each other,
sand will be redistributed between the dunes [86]. There is not yet a general theoretical
model that can self-consistently explain this diversity of dune patterns.
In order to understand the mechanism of sand–particle collisions between sand dunes,
it is necessary to use a continuum model for numerical simulations. Continuum dune
models have been used to quantitatively connect dune characteristics to environmental
conditions and physical processes, including interactions within a single sand dune and
between multiple sand dunes. At larger scales, each sand dune is regarded as a single
particle, and the results of the large-scale continuum model are used to explain the collisions
between particles, i.e., interactions between sand dunes. The pioneering work of [87] is
the basis of most sand dune formation and evolution continuum models. These models
have been mainly applied to study the influences of wind and sand flux on details of
dune morphology [88,89], and the influence of vegetation and induration in stabilizing
dunes [88]. Many studies have examined the characteristic scales of dune formation and
evolution in different environments, such as underwater or on Mars [90,91]. Jump particles
on the back of a windward dune hit the bottom of the dune, causing sand deposition, and
ultimately remain in the windward dune. With the transport of windward dune sand, the
dune’s contour becomes shorter, which leads to a speed increase in the sand flux. Similarly,
windward dunes accumulate sand, become taller, and thus reduce the sand flux speed [87].
It seems that the size ratio between dunes, not the individual size of dunes, determines the
impact of sand collision on sediment transport [92–94].
In future studies, we will attempt to apply a continuum model to fields composed of
hundreds of dunes. However, due to the large temporal and spatial scales involved and
current technical constraints, its feasibility may be limited [87]. In other words, following
the methods adopted in other research and the obtained basic parameters, we hope that the
future measurements of sand flux in areas between desert sand dunes will help to develop
better models for dune fields, based on previous agent-based simulations [83,87,95]. Such
Int. J. Environ. Res. Public Health 2021, 18, 7393 16 of 19

models rely on the accurate description of interdune sand transport, which is still poorly
known, and should thus constitute a topic of further investigation in future work.
To sum up, in the complex field environment, differences in the horizontal dust
fluxes and dust collection efficiencies were caused by differences in the type, accuracy,
and sampling periods of the sand samplers, as well as the sand particle size and height.
Therefore, in future research, developing a sand sampling device with a high sand collection
efficiency that is sensitive to small disturbances to the flow fields in complicated fields is an
important requirement to accurately estimate dust fluxes. Secondly, we aim to develop a
numerical model, and then calibrate the model through field-measured data, so as to carry
out numerical simulations of large-scale sediment flux and dust flux.

5. Conclusions
The cumulative collected sand mass and number of sand particles impacting the
sensor of the designed sand sampler were highly consistent in the complex field test
environment, which had a linear relationship with an R2 value of 0.6053, reflecting good
performance in sand collection. The sand collection efficiency was about 94.3% for the
sand sampler, indicating that the sand sampler could effectively monitor the movement
of sand particles. The data obtained from field experiments allowed us to deduce a
simple formula for calculating the dust horizontal flux: F = 256M. We also estimated the
maximum and minimum dust horizontal flux within a height of 5 cm near the surface of
the mobile sand surface in the tower during 15 dust events, finding values of 190.335 kg
and 1.2 kg, respectively.
Based on the simultaneous observations of wind speed and temperature in our attempt
to analyze the transitional sediment transport rate at a height of 5 cm, it was found that
with an increase of wind speed, the sediment transport rate generally increased, but
during some weather periods, the sediment transport rate changed with the wind speed.
Inconsistent, and in this case, the landmark temperature during the corresponding period
is significantly higher.

Author Contributions: Conceptualization, X.L. and Y.K.; data curation, H.C.; formal analysis, H.L.;
funding acquisition, H.C.; investigation, Y.K.; project administration, X.L.; software, Y.K.; supervision,
H.L.; validation, H.C.; visualization, X.L.; writing—original draft, X.L.; writing—review and editing,
Y.K. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by the Tianshan Youth Talents (Xinjiang) Plan Project (2019Q037).
Institutional Review Board Statement: Not applicable for studies not involving humans.
Informed Consent Statement: Not applicable for studies not involving humans.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Anderson, R.S.; Haff, P.K. Simulation of eolian saltation. Science 1988, 241, 820–824. [CrossRef]
2. Kok, J.F.; Parteli, E.J.; Michaels, T.I.; Karam, D.B. The physics of wind-blown sand and dust. Rep. Prog. Phys. Physical Soc. 2012,
75, 106901. [CrossRef]
3. Shao, Y.P. Physics and Modelling of Wind Erosion; Kluwer Academic Publishers: Dordrecht, The Netherlands, 2008.
4. Ravi, S.; D’Odorico, P.; Breshears, D.D.; Field, J.P.; Goudie, A.; Huxman, T.E.; Li, J.; Okin, G.; Swap, R.; Thomas, A.; et al. Aeolian
process and the biosphere. Rev. Geophys. 2011, 49, 45. [CrossRef]
5. Bullard, J.E.; Baddock, M.; Bradwell, T.; Crusius, J.; Darlington, E.; Gaiero, D.M.; Gassó, S.; Gisladottir, G.; Hodgkins, R.;
McCulloch, R.; et al. High-latitude dust in the Earth system. Rev. Geophys. 2016, 54, 447–485. [CrossRef]
6. Jiang, F.; Ren, B.; Hursthouse, A.S.; Zhou, Y. Trace Metal Pollution in Topsoil Surrounding the Xiangtan Manganese Mine Area
(South-Central China): Source Identification, Spatial Distribution and Assessment of Potential Ecological Risks. Int. J. Environ.
Res. Public Health 2018, 15, 2412. [CrossRef]
7. Shao, Y.; Wyrwoll, K.H.; Chappell, A.; Huang, J. Dust cycle: An emerging core theme in earth system science. Aeolian Res. 2011, 2,
181–204. [CrossRef]
Int. J. Environ. Res. Public Health 2021, 18, 7393 17 of 19

8. Zou, X.; Zhang, C.; Cheng, H.; Kang, L.; Wu, Y. Cogitation on developing a dynamic model of soil wind erosion. Sci. China Earth
Sci. 2015, 58, 462–473. [CrossRef]
9. Sharratt, B.S.; Mcguire, A.; Horneck, D. Early-season wind erosion influenced by soil-incorporated green manure in the Pacific
Northwest. Soil Sci. Soc. Am. J. 2018, 82, 678–684. [CrossRef]
10. WHO. Air Quality Guidelines: Global Update 2005: Particulate Matter, Ozone, Nitrogen Dioxide, and Sulfur Dioxide; WHO: Geneva,
Switzerland, 2006.
11. Li, J.; Shang, J.; Huang, D.; Tang, S.; Zhao, T.; Yang, X.; Zhang, Q.; Liu, K.; Shao, X. Grazing and Cultivated Grasslands Cause
Different Spatial Redistributions of Soil Particles. Int. J. Environ. Res. Public Health 2019, 16, 2639. [CrossRef]
12. Zhang, S.; Ding, G.-D.; Yu, M.-H.; Gao, G.-L.; Zhao, Y.-Y.; Wu, G.-H.; Wang, L. Effect of Straw Checkerboards on Wind Proofing,
Sand Fixation, and Ecological Restoration in Shifting Sandy Land. Int. J. Environ. Res. Public Health 2018, 15, 2184. [CrossRef]
13. Peri, P.L.; Bloomberg, M. Wind breaks in southern Patagonia, Argentina: A review of research on growth models, windspeed
reduction, and effects oncrops. Agrofor. Syst. 2002, 56, 129–144. [CrossRef]
14. Wang, L.-Y.; Xiao, Y.; Rao, E.-M.; Jiang, L.; Xiao, Y.; Ouyang, Z.-Y. An Assessment of the Impact of Urbanization on Soil Erosion in
Inner Mongolia. Int. J. Environ. Res. Public Health 2018, 15, 550. [CrossRef]
15. Alemu, M.M. Ecological benefits of trees as wind breaks and shelter belts. Phys. Geogr. 2016, 6, 10–13.
16. Tshehla, C.E.; Wright, C.Y. Spatial and Temporal Variation of PM10 from Industrial Point Sources in a Rural Area in Limpopo,
South Africa. Int. J. Environ. Res. Public Health 2019, 16, 3455. [CrossRef] [PubMed]
17. Nickling, W.G.; McKenna Neuman, C. Wind tunnel evaluation of a wedgeshaped aeolian sediment trap. Geomorphology 1997, 18,
333–345. [CrossRef]
18. Ono, D.; Weaver, S.; Richmond, K. Quantifying Particulate Matter Emissions from Wind Blown Dust using Real-Time Sand Flux
Measurements; Owens Emissions Papers, Paper-C; EPA: Washington, DC, USA, 2003.
19. Zobeck, T.M.; Sterk, G.; Funk, R.; Rajot, J.L.; Stout, J.E.; Van Pelt, R.S. Measurement and data analysis methods for field-scale
wind erosion studies and model validation. Earth Surf. Proc. Land. 2003, 28, 1163–1188. [CrossRef]
20. Van Pelt, R.S.; Peters, P.; Visser, S. Laboratory wind tunnel testing of three com-monly used saltation impact sensors. Aeolian Res.
2009, 1, 55–62. [CrossRef]
21. Webb, N.P.; Herrick, J.E.; Van Zee, J.W. The national wind erosion research network: Building astandardized long-term data
resource for aeolian research, modeling and land man-agement. Aeolian Res. 2016, 22, 23–36. [CrossRef]
22. Goossens, D.; Nolet, C.; Etyemezian, V. Field testing, comparison, and discussion of five aeolian sand transport measuring devices
operating on different measuring principles. Aeolian Res. 2018, 32, 1–13. [CrossRef]
23. Leatherman, S.P. Short communication: A new aeolian sand trap design. Sedimentology 1978, 25, 303–306. [CrossRef]
24. Fryberger, S.G.; Al-Sari, A.M.; Clisham, T.J. Wind sedimentation in the Jafurah sand sea, Saudi Arabia. Sedimentology 1984, 31,
413–431. [CrossRef]
25. Zhong, H.Y.; Lin, C.; Sun, Y.; Kikumoto, H.; Ooka, R.; Zhang, H.L.; Hu, H.; Zhao, F.Y.; Jimenez-Bescos, C. Boundary layer wind
tunnel modeling experiments on pumping ventilation through a three-story reduce-scaled building with two openings. Build.
Environ. 2021, 202, 108043. [CrossRef]
26. Wang, R.D.; Li, Q.; Chang, C.P.; Guo, Z.L. Sampling efficiency of a new flat opening collector for particles with various sizes. J.
Des. Res. 2018, 38, 734.
27. Wilson, S.J.; Cooke, R.U. Wind erosion. In Soil Erosion; Kirkby, M.J., Morgan, R.P.C., Eds.; Wiley: Chichester, UK, 1980; pp. 217–252.
28. Fryrear, D.W. A field dust sampler. J. Soil Water Conserv. 1986, 41, 117–120.
29. Dong, Z.; Sun, H.; Zhao, A. WITSEG sampler: A segmented sand sampler for wind tunnel test. Geomorphology 2004, 59, 119–129.
[CrossRef]
30. Zhao, Y.L.; Ma, S.S.; Chen, Z.; Sun, Y.C. The calibration of the whirl type separation sand sampler. J. Agr. Mech. Res. 2007, 4, 95–98.
31. Gu, Z.; Guo, L. Experimental investigation on trap stagnant effect and sand flux in aeolian sand transport. Phys. Lett. A 2007, 368,
435–441. [CrossRef]
32. Basaran, M.; Erpul, G.; Uzun, O.; Gabriels, D. Comparative efficiency testing for a newly designed cyclone type sediment trap for
wind erosion measurements. Geomorphol. 2011, 130, 343–351. [CrossRef]
33. Rotnicka, J. Aeolian vertical mass flux profiles above dry and moist sandy beach surfaces. Geomorphol. 2013, 187, 27–37. [CrossRef]
34. Sherman, D.; Swann, C.; Barron, J. A high-efficiency, low-cost aeolian sand trap. Aeolian Res. 2014, 13, 31–34. [CrossRef]
35. Hilton, M.; Nickling, B.; Wakes, S.; Sherman, D.; Konlechner, T.; Jermy, M.; Geoghegan, P. An efficient, self-orienting, vertical-array,
sand trap. Aeolian Res. 2017, 25, 11–21. [CrossRef]
36. Goossens, D.; Offer, Z.; London, G. Wind tunnel and field calibration of five aeolian sand traps. Geomorphology 2000, 35, 233–252.
[CrossRef]
37. Kang, Y.; He, Q.; Yang, X. Based on the earth’s surface layer dust flux field observation and study on collection efficiency. Am. J.
Med. Sci. Med. 2017, 5. [CrossRef]
38. An, M.; Li, S.; Liu, Z.; Yan, B.; Li, L.; Chen, Z. Galloping vibration of stay cable installed with a rectangular lamp: Field
observations and wind tunnel tests. J. Wind. Eng. Ind. Aerodyn. 2021, 215, 104685.
39. Shannak, B.; Corsmeier, U.; Kottmeier, C.; Al-Azab, T. Wind tunnel study of twelve dust samples by large particle size. Atmos.
Environ. 2014, 98, 442–453. [CrossRef]
Int. J. Environ. Res. Public Health 2021, 18, 7393 18 of 19

40. Martin, R.L.; Kok, J.F. Size-independent susceptibility to transport in aeolian saltation. J. Geophys. Res. Earth Surf. 2019, 124,
1658–1674. [CrossRef]
41. Namikas, S.L. Field evaluation of two traps for high-resolution aeolian transport measurements. J. Coast. Res. 2002, 18, 136–148.
42. Haustein, K.; Washington, R.; King, J. Testing the performance of state-of-the-art dust emission schemes using DO4Models field
data. Geosci. Model Dev. 2015, 8, 341–362. [CrossRef]
43. Li, Z.S.; Ni, J.R. Sampling efficiency of vertical array aeolian sand traps. Geomorphology 2003, 52, 243–252. [CrossRef]
44. Lin, Y.; Hang, J.; Yang, H.; Chen, L.; Chen, G.; Ling, H.; Sandberg, M.; Claesson, L.; Lam, C.K.C. Investigation of the Reynolds
number independence of cavity flow in 2D street canyons by wind tunnel experiments and numerical simulations. Build. Environ.
2021, 201, 107965. [CrossRef]
45. Zhao, C.M. Analysis of Sand Grain Morphology in Taklimakan Desert; Xinjiang Normal University: Urumqi, China, 2012; (In Chinese
with English Abstract).
46. He, Q.; Yang, X.H. Study on wind-induced sand observations in Taklimakan Desert—Preliminary study of test and observation
results. J. China Desert 2011, 31, 56–61, (In Chinese with English Abstract).
47. Yang, X.H.; He, Q. A field experiment on dust emission by wind erosion in the Taklimakan desert. Acta Meteor, Sinica 2012, 26,
241–249. [CrossRef]
48. Kang, Y.; He, Q. Urumqi. J. Arid Land Resour. Environ. 2017, 31, 119–125, (In Chinese with English Abstract).
49. Xu, J.; Xiao, Y.; Xie, G.; Zhen, L.; Wang, Y.; Jiang, Y. The spatio-temporal disparities of areas benefitting from the wind erosion
prevention service. Int. J. Environ. Res. Public Health 2018, 15, 1510. [CrossRef]
50. Anderson, R.S.; Haff, P.K. Wind modification and bed response during saltation of sand in air. Acta Mech. Suppl. 1991, 1, 21–51.
51. McEwan, I.K.; Willetts, B.B. Adaptation of the near-surface wind to the development of sand transport. J. Fluid Mech. 1993, 252,
99–115. [CrossRef]
52. Ungar, J.E.; Haff, P.K. Steady-state saltation in air. Sedimentology 1987, 34, 289–299. [CrossRef]
53. Ma, G.; Wang, Y.; Zheng, J. Numerical analysis of the influence of the near ground turbulence on the wind-sand flow under the
natural wind. Granular Matter 2021, 23, 40. [CrossRef]
54. Huang, N.; He, P.; Zhang, J. Large-eddy simulation of sand transport under unsteady wind. Geomorphol. 2020, 358, 107105.
[CrossRef]
55. Shao, Y.; Raupach, M.R. The overshoot and equilibration of saltation. J. Geophys. Res. 1992, 97, 20559–20564. [CrossRef]
56. Rice, M.A.; Willetts, B.B.; McEwan, I.K. An experimental study of multiple grain-size ejecta produced by collisions of saltating
grains with a flat bed. Sedimentology 1995, 42, 695–706. [CrossRef]
57. Kok, J.F.; Renno, N.O. A comprehensive numerical model of steady state saltation (COMSALT). J. Geophys. Res. 2009, 114, D17204.
[CrossRef]
58. Gillette, D.A. On the production of soil wind erosion having the potential for long range transport. J. Rech. Atmos. 1974, 8,
734–744.
59. Shao, Y.; Raupach, M.R.; Findlater, P.A. Effect of saltation bombardment on the entrainment of dust by wind. J. Geophys. Res. 1993,
98, 12719–12726. [CrossRef]
60. Dur’an, O.; Claudin, P.; Andreotti, B. On aeolian transport: Grain-scale interactions, dynamical mechanisms and scaling laws.
Aeolian Res. 2011, 3, 243–270. [CrossRef]
61. Bagnold, R.A. The movement of desert sand Proc. R. Soc. Lond. A 157 0594–620 Owen P R 1964 Saltation of uniform grains in air.
J. Fluid Mech. 1936, 20, 225–242.
62. Owen, P.R. Saltation of uniform grains in air. J. Fluid Mech. 1964, 20, 225–242. [CrossRef]
63. Sorensen, M.; McEwan, I. On the effect of mid-air collisions on aeolian saltation. Sedimentology 1996, 43, 65–76. [CrossRef]
64. Dong, Z.; Liu, X.; Li, F.; Wang, H.; Zhao, A. Impact-entrainment relationship in a saltating cloud. Earth Surf. Process. Landf. 2002,
27, 641–658. [CrossRef]
65. Dong, Z.B.; Huang, N.; Liu, X.P. Simul of the probability of midair interparticle collisions in an aeolian saltating cloud. J. Geophys.
Res. 2005, 110, D24113. [CrossRef]
66. Huang, N.; Zhang, Y.L.; D’Adamo, R. A model of the trajectories and midair collision probabilities of sand particles in a steady
state saltation cloud. J. Geophys. Res. 2007, 112, D08206. [CrossRef]
67. Rice, M.A.; Willetts, B.B.; McEwan, I.K. Observations of collisions of saltating grains with a granular bed from high-speed
cine-film. Sedimentology 1996, 43, 21–31. [CrossRef]
68. Gordon, M.; Neuman, C.M. A comparison of collisions of saltating grains with loose and consolidated silt surfaces. J. Geophys.
Res. 2009, 114, F04015. [CrossRef]
69. Oger, L.; Ammi, M.; Valance, A. Discrete Element Method studies of the collision of one rapid sphere on 2D and 3D packings. Eur.
Phys. J. E 2005, 17, 467–476. [CrossRef]
70. Oger, L.; Ammi, M.; Valance, A. Study of the collision of one rapid sphere on 3D packings: Experimental and numerical results.
Comput. Math. Appl. 2008, 55, 132–148. [CrossRef]
71. Werner. B.T. A steady-state model of wind-blown sand transport. J. Geol. 1990, 98, 1–17.
72. Mitha, S.; Tran, M.Q.; Werner, B.T. The grain-bed impact process in aeolian saltation. Acta Mech. 1986, 63, 267–278. [CrossRef]
73. Ho, T.D.; Dupont, P.; El Moctar, A.O.; Valance, A. 2012 Particle velocity distribution in saltation transport. Phys. Rev. E 2012,
85, 052301. [CrossRef] [PubMed]
Int. J. Environ. Res. Public Health 2021, 18, 7393 19 of 19

74. Fu, G.; Xu, X.; Qiu, X.; Xu, G.; Shang, W.; Yang, X.; Zhao, P.; Chai, C.; Hu, X.; Zhang, Y.; et al. Wind tunnel study of the effect of
planting Haloxylon ammodendron on aeolian sediment transport. Biosyst. Eng. 2021, 208, 234–245. [CrossRef]
75. Baas, A.C.W. Evaluation of saltation flux impact responders (Safires) for measuring instantaneous aeolian sand transport intensity.
Geomorphology 2004, 59, 99–118. [CrossRef]
76. McKee, E.D. A Study of Global Sand Seas; US Government Printing Office: Washington, DC, USA, 1979.
77. Dong, Z.B.; Man, D.Q.; Luo, W.Y. Horizontal aeolian sediment flux in the Minqin area, a major source of Chinese dust storms.
Geomorphology 2010, 116, 58–66. [CrossRef]
78. Sharratt, B. Size distribution of windblown sediment emitted from agricultural fields in the Columbia Plateau. Soil Sci. Soc. Am. J.
2011, 75, 1054–1060. [CrossRef]
79. Farrell, E.J.; Sherman, D.J. Vertical distribution of grain size for wind blown sand. Aeolian Res. 2012, 7, 51–61. [CrossRef]
80. Zhang, Z.C.; Dong, Z.B.; Zhao, A.G. The characteristics of aeolian sediment flux profiles in the south-eastern Tengger Desert.
Sedimentology 2011, 58, 1884–1894. [CrossRef]
81. Pierre, C.; Bergametti, G.; Marticorena, B. Modeling wind erosion flux and its seasonality from a cultivated sahelian surface: A
case study in Niger. Catena 2014, 122, 61–71. [CrossRef]
82. Zhang, Z.C.; Dong, Z.B.; Qian, G.Q. An investigation into the pro-cesses and quantity of dust emissions over gravel and sand
deserts in north-western China. Bound. Layer Meteorol. 2017, 163, 523–535. [CrossRef]
83. EParteli, J.R.; Herrmann, H.J. A simple model for a transverse dune field. Physica A 2003, 327, 554–562. [CrossRef]
84. Hersen, P.; Andersen, K.H.; Elbelrhiti, H.; Andreotti, B.; Claudin, P.; Douady, S. Corridors of barchan dunes: Stability and size
selection. Phys. Rev. E 2004, 69, 011304. [CrossRef] [PubMed]
85. Elbelrhiti, H.; Claudin, P.; Andreotti, B. Field evidence for surface-wave-induced instability of sand dunes. Nature 2005, 437,
720–723. [CrossRef]
86. Hersen, P.; Douady, S. Collision of barchan dunes as a mechanism of size regulation. Geophys. Res. Lett. 2005, 32, 21403. [CrossRef]
87. Diniega, S.; Glasner, K.; Byrne, S. Long-time evolution of models of aeolian sand dune fields: Influence of dune formation and
collision. Geomorphol. 2010, 121, 55–68. [CrossRef]
88. Herrmann, H.J.; Durán, O.; Parteli, E.J.R.; Schatz, V. Vegetation and Induration as Sand Dunes Stabilizators. J. Coast. Res. 2008,
246, 1357–1368. [CrossRef]
89. Parteli, E.J.R.; Durán, O.; Herrmann, H.J. The shape of the barchan dunes in the Arkhangelsky Crater on Mars. In Proceedings of
the 37th Annual Lunar and Planetary Science Conference, League City, TX, USA, 13–17 March 2006.
90. Claudin, P.; Andreotti, B. A scaling law for aeolian dunes on Mars, Venus, Earth, and for subaqueous ripples. Earth Planet. Sci.
Lett. 2006, 252, 30–44. [CrossRef]
91. Parteli, E.J.R.; Herrmann, H.J. Dune formation on the present Mars. Phys. Rev. E 2007, 76, 041307. [CrossRef] [PubMed]
92. Katsuki, A.; Nishimori, H.; Endo, N.; Taniguchi, K. Collision Dynamics of Two Barchan Dunes Simulated Using a Simple Model.
J. Phys. Soc. Jpn. 2005, 74, 538–541. [CrossRef]
93. Endo, N.; Taniguchi, K.; Katsuki, A. Observation of the whole process of interaction between barchans by flume experiments.
Geophys. Res. Lett. 2004, 31, 12503. [CrossRef]
94. Durán, O.; Schwämmle, V.; Herrmann, H. Breeding and solitary wave behavior of dunes. Phys. Rev. E 2005, 72, 021308. [CrossRef]
95. Génois, M.; Du Pont, S.C.; Hersen, P.; Grégoire, G. An agent based model of dune interactions produces the emergence of patterns
in deserts. Geophys. Res. Lett. 2013, 40, 3909–3914. [CrossRef]