remote sensing
Article
Temporal Evolution and Regional Properties of Aerosol over the
South China Sea
Jie Chen 1,2,3 , Wenyue Zhu 1,3, * , Qiang Liu 1,3 , Xianmei Qian 1,3 , Xiaowei Chen 1,3 , Jianjie Zheng 1,3 ,
Tao Yang 1,2,3 , Qiuyi Xu 1,2,3 and Tengfei Yang 1,2,3
Citation: Chen, J.; Zhu, W.; Liu, Q.;
Qian, X.; Chen, X.; Zheng, J.; Yang, T.;
Xu, Q.; Yang, T. Temporal Evolution
and Regional Properties of Aerosol
over the South China Sea. Remote
Sens. 2023, 15, 501. https://doi.org/
10.3390/rs15020501
Academic Editor: Hanlim Lee
Received: 21 November 2022
Revised: 30 December 2022
Accepted: 7 January 2023
Published: 14 January 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Remote Sens. 2023, 15, 501. https://doi.org/10.3390/rs15020501
1 Key Laboratory of Atmospheric Optics, Anhui Institute of Optics and Fine Mechanics, HFIPS, Chinese
Academy of Sciences, Hefei 230031, China
2 Science Island Branch of Graduate School, University of Science and Technology of China, Hefei 230026, China
3 Advanced Laser Technology Laboratory of Anhui Province, Hefei 230037, China
* Correspondence: zhuwenyue@aiofm.ac.cn
Abstract: Aerosol robotic network (AERONET) data from Dongsha Island (20.699N, 116.729E) and
Taiping Island (35.90N, 3.03W) over the South China Sea (SCS) from January 2018 to December 2020
were used to analyze and discuss the temporal evolution properties of aerosols in the South China
Sea. Surrounding AERONET stations (Hong Kong, NSPO, Nha Trang and Singapore) were also
used to analyze regional characteristics. High aerosol loads over Dongsha were strongly associated
with the anthropogenic fine particle transport from the southeastern coast of China and occasional
advection of desert dust from Mongolian areas. The high fine aerosol loading in Taiping originates
from the region between Singapore and Indonesia. Compared with other marine islands in the
world, SCS was not a pure marine aerosol environment and was affected by terrestrial aerosols.
In the Taiping area, aerosol optical depth τ (500 nm) was 0.17 ± 0.13 and the average Ångström
exponent α (440–870 nm) was 0.96 ± 0.36. However, that of Dongsha shows the larger values of τ
(0.26 ± 0.21) and α (1.1 ± 0.38), indicating that there are large fluctuations in aerosol concentration
and size. Aerosol loads in different regions of the SCS due to uneven socioeconomic and complex
meteorological systems, such as those of the coastal cities of China, Singapore, and the region between
Singapore and Indonesia, contribute to the high optical depth. The special meteorological regime and
aerosol source mechanism in the SCS leads to the obvious seasonal cycle of aerosol optical depth and
Ångström index. Moreover, the loading variations of aerosols on Dongsha Island and Taiping Island
were highly consistent with those of coastal cities around them, suggesting the significant effect of
the aerosol in the SCS by the surrounding coastal cities, although the aerosol optical depth in these
two places was much lower than that in the surrounding cities.
Keywords: South China Sea; AERONET; aerosol optical depth
1. Introduction
The South China Sea (SCS) connects the Pacific Ocean with the Indian Sea and is one
of the largest marginal seas in Southeast Asia [1–3]. In such a low latitude sea, it is often
affected by aerosol eruptions from Asia, resulting in local climate change anomalies [3–5].
The complex physical, biological, geographical, monsoon climate and complicated economy
are of great importance in the combined effect of aerosol properties [6]. In particular, the
SCS receives substantial amounts of different types of aerosols from surrounding regions
due to various economic activity [5,7].
Many efforts have been made by many groups to characterize this area by using
two main methods: sampling and remote sensing. The sampling method of atmospheric
aerosols is one of the main methods used to research the pollutant sources and related
chemical properties in this area. Xiao et al. [8] found that fossil fuel combustion on Yongxing
Island (especially coal in the coastal areas of China) was an important source of NO3 − (56%)
https://www.mdpi.com/journal/remotesensing
Remote Sens. 2023, 15, 501 2 of 16

and SO4 2− (22%), and biomass burning accounted for 41% of K+ in Asia. On Dongsha
Island, carbonaceous content and water-soluble ions are the dominant components of total
suspended particulates (TSP), which are affected by mobile vehicles and coal and biomass
burning [9]. Zhang et al. [10] reported that methanesulfonic acid measured at PM2.5
aerosol samples over the northern SCS was comparable to those over other coastal regions.
The sampling study mainly focused on the short-term analysis of pollutant composition
in this region and its nearby islands, but did not carry out the long-term exploration of
aerosol in the South China Sea. In addition, remote sensing provides an important tool
for understanding the function of aerosols in Earth’s radiation budget. From satellite
and ground remote sensing measurements, aerosol column characteristics can be obtained.
Smirnov et al. [11,12] conducted a number of studies on aerosol optical properties in various
maritime and coastal areas, except the SCS, by using AERONET. Studies gave a generally
accepted criteria for the determination of “pure marine aerosol”: AOD (500 nm) smaller
than 0.15 and Ångström exponent α (440–870 nm) less than 1. Itahashi et al. [13] and Zheng
et al. [14] used MODIS to analyse aerosol optical depths in the Seto Inland Sea and East
Sea near the SCS. Reid et al. [15] analyzed 7 Southeast Asian Studies (7-SEAS) and found
that the strong monsoon effected the transport and removal of aerosol particles by the
AERONET and Lidar data in SCS. Many scholars have studied local aerosol characteristics
and analysed microphysical properties, optical properties and aerosol distributions based
on ground-based measurements surrounding the SCS [16,17].
Although the abovementioned studies focused on the analysis of local pollution and
aerosol properties, they only targeted short-term experimental measurements or focused
mainly on high aerosol load cases. Therefore, it is necessary to study the aerosol properties
over the SCS to evaluate the aerosol regimes, which is significant with regard to researching
temporal evolution and regional properties and makes up for the absence of aerosol models
over this explored region. In this study, the aerosol loading, source and spatio-temporal
variation of Dongsha Island and Taiping Island in the South China Sea from 1 January 2018
to 31 December 2020 were presented by analyzing AERONET data. The optical properties
of aerosol over the SCS are analysed and discussed by considering the complexity of aerosol
origin and propagation, meteorology and geography, which is of great significance for the
study of the regional climate and atmospheric radiation transfer model.

2. Data Sources and Research Sites

2.1. AERONET Measurements
As a standard automatic solar photometer, a CE-318 was used to measure the char-
acteristics of columnar aerosols in the AERONET network [18]. The full viewing angle of
the instrument was 1.2◦ , and direct sunlight measurement was conducted at 340, 380, 440,
500, 670, 870, 940 and 1020 nm (nominal wavelength). Moreover, the τ at each wavelength
was retrieved from the direct sunlight measurements, except τ (940 nm), since it was used
to calculate water vapour absorption. A further description about the photometer can be
found in Holben et al. [18]. The AERONET collaboration provides spectral aerosol optical
depth (AOD), retrieval products, and global distribution observations of precipitable water
under different aerosol states. In order to understand the AOD corresponding to short
wavelength and long wavelength, the aerosol optical properties were studied based on the
wavelengths of 380, 500 and 1020 nm. In addition, the Ångström exponent α (440–870 nm)
can effectively show the variation of particle size [19]. Ångström exponent α was obtained
from 440 nm to 870 nm by exponential fitting of the aerosol optical depth at different
wavelengths. Fine aerosol optical depth at 500 nm (τF (500 nm)), coarse aerosol optical
depth at 500 nm (τC (500 nm)), and fraction of optical depth of fine mode particles (FMF)
were determined by de-convolution algorithm [20]. Currently, AERONET contains three
levels of data, among which level 1 data was unfiltered, level 1.5 data was quality controlled
and filtered, and the level 2 data had quality assurance [21].
 tained from 440 nm to 870 nm by exponential fitting of the aerosol optical depth at differ-
ent wavelengths. Fine aerosol optical depth at 500 nm (τF (500 nm)), coarse aerosol optical
depth at 500 nm (τC (500 nm)), and fraction of optical depth of fine mode particles (FMF)
were determined by de-convolution algorithm [20]. Currently, AERONET contains three
Remote Sens. 2023, 15, 501 levels of data, among which level 1 data was unfiltered, level 1.5 data was quality 3con-
of 16
trolled and filtered, and the level 2 data had quality assurance [21].

2.2.
2.2.Sites
SitesSelection
Selection
The
The locationof
location ofeach
eachAERONET
AERONETsite siteisisshown
shownin inFigure
Figure1.1.ThisThisstudy
studyfocused
focusedon onthe
the
AERONET stations at Dongsha Island (20.699N, 116.729E)
AERONET stations at Dongsha Island (20.699N, 116.729E) and Taiping Island (35.90N, and Taiping Island (35.90N,
3.03W)
3.03W)overoverthe theSCS.
SCS.Dongsha
Dongsha Island
Island lieslies
in in
thethe
north of the
north SCSSCS
of the andandis surrounded
is surrounded by the
by
Philippines,
the Philippines,southern China
southern and the
China andIndochinese
the Indochinese Peninsula in theindistance
Peninsula (Figure
the distance 1). Tai-
(Figure 1).
ping Island
Taiping is located
Island in theinnorthwest
is located the northwest of theofNansha Islands,
the Nansha approximately
Islands, 740 km740
approximately away
km
from
awaythe fromXishatheIslands, and onlyand
Xisha Islands, 1000onlykm 1000
awaykm from Singapore
away at the eastern
from Singapore mouth
at the of
eastern
the Straitofofthe
mouth Malacca.
Strait ofIt is valuableItin
Malacca. is the aspectsinofthe
valuable channel
aspects safety, shipwreck
of channel safety,notification,
shipwreck
meteorological monitoring, international
notification, meteorological aviation information,
monitoring, international etc. NSPO etc.
aviation information, (24.784N,
NSPO
121.001E)
(24.784N,of121.001E)
AERONET station, located
of AERONET approximately
station, 628 km northeast
located approximately 628 km of Dongsha
northeastIs- of
land, is theIsland,
Dongsha abbreviation of the Taiwan
is the abbreviation of the Space Center.
Taiwan Space It Center.
was used to evaluate
It was the impact
used to evaluate the
impact
of urbanof urban of
sources sources
aerosols of on
aerosols
Dongsha on Dongsha
Island. The Island. The Hongkong
Hongkong station (22.483N,
station (22.483N, 114.17E)
114.17E)
and Nha andTrang Nha Trang(12.205N,
station station (12.205N,
109.206E) 109.206E)
are located are located
340 km 340 andkm 620and
km620 awaykm from
away
from Dongsha
Dongsha Island Island and Taiping
and Taiping Island,Island, respectively,
respectively, to assess to the
assess the distribution
spatial spatial distribution
of aer-
of aerosols.
osols. The Singapore
The Singapore station (1.298N,
station (1.298N, 103.780E) 103.780E)
is located is 1000
locatedkm 1000 km southwest
southwest of Taiping of
Taiping
Island toIsland
assess totheassess
source theofsource
aerosoloflong-distance
aerosol long-distance
transport. transport.
Since theSince the data from
data from
AERONETaerosol
AERONET aerosolsites
siteswere
werenotnotcontinuous
continuousand andsynchronous
synchronousin intime,
time,ininorder
orderto toobtain
obtain
therecent
the recent aerosol
aerosol optical
optical depth,
depth, seasonal
seasonaland and spatial
spatial distribution
distributioncharacteristics
characteristicsin inthe
the
SouthChina
South ChinaSea, Sea,our
ourstudy
studychose
chose2018
2018to to2020
2020as asthe
theresearch
research period.
period.

China NSPO

Hongkong
Dongsha
South Pacific
China Sea
Nha Trang Sea
Taiping

Singapore

Locationsofofthe
Figure1.1.Locations
Figure theobservation
observationsites
sitesininthis
thisstudy.
study.Dongsha
DongshaIsland
Islandand
andTaiping
TaipingIsland
Islandsites
sites
areshown
are shownas asred
redstars,
stars,and
andother
othersites
sitesare
are shown
shown as as blue
blue squares.
squares. All
Allstations
stationsare
areaffiliated
affiliatedwith
with
AERONET.
AERONET.

2.3. EAC4 and ERA5

EAC4 (ECMWF Atmospheric Composition Reanalysis 4) is the fourth generation
European Centre for Medium-Range Weather Forecasts (ECWMF) global atmospheric
composition reanalysis. The reanalysis uses atmospheric models based on physical and
chemical laws to combine the model data with observation data from all over the world to
form a global complete and consistent dataset. This principle, known as data assimilation,
is based on the method used by the numerical weather prediction center and the air quality
prediction center. MODIS observations of AOD (550 nm) over ocean and land (except
bright surfaces) was also assimilated in EAC. Tropospheric aerosols were classified into
five types: sea salt, organic matter, desert dust, sulfate aerosols and black carbon. All
aerosol species are regarded as tracers, and are incorporated into the vertical diffusion and
convection scheme of the integrated forecasting system (IFS) [22]. Detailed methods for
EAC4 to obtain various aerosol optical depths can be found in many studies [23,24]. The
 osol species are regarded as tracers, and are incorporated into the vertical diffusion and
convection scheme of the integrated forecasting system (IFS) [22]. Detailed methods for
EAC4 to obtain various aerosol optical depths can be found in many studies [23,24]. The
AOD and the mass concentration in Dongsha and Taiping were obtained through EAC4
in this study [25]. In addition, the high consistency (R2 = 0.92) of AOD obtained by EAC
Remote Sens. 2023, 15, 501 and AERONET indicates that the data of EAC4 can be used to effectively verify the anal- 4 of 16

ysis results (Figure 2).

ERA5 is the ECMWF’s fifth generation reanalysis of global climate and weather over
the
AOD past
and40the
to mass
70 years. Hourly estimates
concentration of large
in Dongsha amounts
and Taiping of atmosphere,
were oceanEAC4
obtained through waves in
this land
studysurfaces
[25]. In was
addition, the high consistency 2 = 0.92) of AOD obtained by EAC and
and provided by ERA5 [26]. In(R
this study, the wind speed, wind direc-
AERONET
tion and wave indicates
height that
of 10the data were
metres of EAC4 can betoused
obtained to meteorological
access effectively verify the analysis
characteristics
results
of (Figure
the South 2). Sea.
China

1.5
EAC
AERONET

1.0
t (500nm)

0.5

0.0
2018/1/1 2018/7/1 2019/1/1 2019/7/1 2020/1/1 2020/7/1 2021/1/1
Day
Figure
Figure 2.
2. Comparison
Comparison of
of ττ(500
(500nm)
nm)obtained
obtained by
by AERONET
AERONET and
and EAC4
EAC4 at
at Dongsha
Dongsha Island.
Island.

ERA5
2.4. Air MassisTrajectories
the ECMWF’s fifth generation reanalysis of global climate and weather over
the past
HYSPLIT70is years.
40 to Hourly
a complex andestimates
widely of
usedlarge amounts
system thatofuses
atmosphere, ocean waves
the Lagrangian and
approach
land surfaces was provided by ERA5 [26]. In this study, the wind speed,
and Eulerian methodology for the calculation of air encapsulation paths, complex wind direction
and wave height of 10 metres were obtained to access meteorological characteristics of the
transport, pollutant dispersion, chemical transformation, and particle deposition simula-
South China Sea.
tions [27]. By calculating five-day backwards trajectories ending at 00:00 UTC at Dongsha
Island
2.4. Airand
MassTaiping Island for 500 m above ground level using the HYSPLIT model every
Trajectories
day for three years, the source and transport of air mass to our target AERONET sites
HYSPLIT is a complex and widely used system that uses the Lagrangian approach and
were characterized. We chose five days as the compute cycles for this backwards trajec-
Eulerian methodology for the calculation of air encapsulation paths, complex transport,
tory analysis to ensure trajectory integrity and reduce computational burden. In addition,
pollutant dispersion, chemical transformation, and particle deposition simulations [27]. By
calculating five-day backwards trajectories ending at 00:00 UTC at Dongsha Island and
Taiping Island for 500 m above ground level using the HYSPLIT model every day for three
years, the source and transport of air mass to our target AERONET sites were characterized.
We chose five days as the compute cycles for this backwards trajectory analysis to ensure
trajectory integrity and reduce computational burden. In addition, the analysis height of
500 metres can effectively cover multiple air mass sources and transport trajectories at other
altitudes that are similar to this height.

3. Results and Analysis

3.1. Time Evolution of Aerosol Properties over Dongsha Island and Taiping Island
Figure 3(a1 ,b1 ) shows the time evolution of daily mean values of τ (380, 500 and
1020 nm) and α (440–870 nm) measured at Dongsha Island in the SCS from 1 January 2018
to 31 December 2020. Much of the missing data in the τ (λ) and α (440–870 nm) series is due
to invalid data. All of the daily average values of Dongsha Island are presented in Table 1.
A distinct feature is the large variability of τ (λ) (τ (500 nm) that ranged from 0.04 (Min) to
1.3 (Max), which is closely linked to the diversity of air mass sources in the study area, as
mentioned below. In order to evaluate the magnitude of variation in the various data sets,
the coefficient of fluctuation (FOC) was expressed by dividing the standard deviation of
the data set by the mean. As shown in Table 1, τ (λ) at 1020 nm showed smaller FOC than
 to invalid data. All of the daily average values of Dongsha Island are presented in Table
1. A distinct feature is the large variability of τ (λ) (τ (500 nm) that ranged from 0.04 (Min)
to 1.3 (Max), which is closely linked to the diversity of air mass sources in the study area,
as mentioned below. In order to evaluate the magnitude of variation in the various data
Remote Sens. 2023, 15, 501 sets, the coefficient of fluctuation (FOC) was expressed by dividing the standard deviation 5 of 16
of the data set by the mean. As shown in Table 1, τ (λ) at 1020 nm showed smaller FOC
than that at 380 nm. It is well known that τ (λ) is more susceptible to fine particles (radius
below
that at 0.5
380µnm.
m) atIt short
is wellwavelengths,
known that τwhereas it issusceptible
(λ) is more more sensitive to coarse
to fine particles
particles (radius(radius
below
above 0.5 µ m) at long wavelengths [28]. In accordance with larger
0.5 µm) at short wavelengths, whereas it is more sensitive to coarse particles (radius FOC of τF (500 nm)
above
compared to τ C (500 nm) (Table 1), the higher variability of τ (λ) for Fshorter wavelengths
0.5 µm) at long wavelengths [28]. In accordance with larger FOC of τ (500 nm) compared
indicates
to τC (500strong variability
nm) (Table 1), theinhigher
the fine particle of
variability load over
τ (λ) forDongsha Island. In this
shorter wavelengths region,
indicates
the large
strong τ (λ) fluctuations
variability in the fine corresponding
particle load over to Dongsha
long wavelengths
Island. Inmay be caused
this region, the by theτ in-
large (λ)
trusion of marine
fluctuations aerosols to
corresponding and dust,
long as described
wavelengths may below. In addition,
be caused coarse particles
by the intrusion of marine in
the atmosphere have a shorter lifetime than fine particles, which may
aerosols and dust, as described below. In addition, coarse particles in the atmosphere have also be a reason for
athe large lifetime
shorter coarse particle
than finefluctuations τC (λ).
particles, which Thealso
may α (440–870
be a reason nm)foralso
theshowed greatparticle
large coarse varia-
bility, ranging C from 0.15 to 1.95 with an average of 1.1 ± 0.38, indicating
fluctuations τ (λ). The α (440–870 nm) also showed great variability, ranging from 0.15 to that the aerosol
typeswith
1.95 (coarse particles,
an average fine±particles
of 1.1 and different
0.38, indicating that the mixtures
aerosol of coarse
types particles
(coarse and fine
particles,
particles)and
particles were different
different under different
mixtures of coarseatmospheric
particles andconditions. Notably,
fine particles) were α (400–870
different nm)
under
was greater
different than 1 in conditions.
atmospheric 66% of the analyzed
Notably, αdays, indicating
(400–870 nm) was thatgreater
the aerosol
than 1population
in 66% of the of
Dongsha days,
analyzed Islandindicating
was dominated
that theby fine particles
aerosol population in most of the analyzed
of Dongsha Island wasdays. This con-
dominated by
clusion
fine was further
particles in mostsupported by the days.
of the analyzed analysisThisofconclusion
fine mode was fraction (FMF),
further which ranged
supported by the
analysis
from 0.18oftofine0.97mode
(meanfraction
0.65 ± (FMF), which
0.19), with 78% ranged
of thefrom
days 0.18 to 0.97having
analyzed (meana0.65 daily± mean
0.19),
with
greater78% of the
than 0.5.days analyzed having a daily mean greater than 0.5.

(a1) 1.8 (a2) 2.4

Dongsha Taiping t (1020 nm)
1.8
t (500 nm)
1.2 t (380 nm)
1.2
t

0.6
0.6

0.0 0.0
(b1) 2.0 (b2) 2.5
a (440-870 nm)
2.0
1.5
1.5
a

1.0
1.0
0.5
0.5

0.0 0.0
2018/1/1 2018/7/1 2019/1/1 2019/7/1 2020/1/1 2020/7/1 2021/1/1 2018/1/1 2018/7/1 2019/1/1 2019/7/1 2020/1/1 2020/7/1 2021/1/1
Day Day

3. (a) Daily average aerosol optical depth at 380, 500 and 1020 nm and (b)
Figure 3. (b) Ångström
Å ngström index
calculated from
calculated from440
440toto870
870nmnmatatDongsha
DongshaIsland
Island(a
(a11,b
,b11)) and
and Taiping
TaipingIsland
Island(a
(a22,b
,b22)) in
in the
the South China
South China
Sea from
Sea from 11 January
January 2018
2018 to
to 31
31 December
December 2020.
2020. The
The error
error bars
bars are
are the
the standard
standard deviations.
deviations.

Table 1. Multi-wavelength (1020, 500 and 380 nm) aerosol optical depth τ, Ångström exponent (α
(440–870)), fine aerosol optical depths at 500 nm (τF (500 nm), coarse aerosol optical depths at τC
(500 nm)) and fraction of optical depth of fine mode particles (FMF) in Dongsha from 1 January
2018 to 31 December 2020. All data were processed using daily averages. The total number of
samples (N total), average coefficient (Mean), standard deviation (SD), Minimum (Min), Median
(Med), Maximum (Max) and the coefficient of fluctuation (FOC) are also counted to assess variations
in the data.

N Total Mean SD Min Med Max FOC

τ (1020 nm) 456 0.11 0.08 0.02 0.1 0.48 69%
τ (500 nm) 456 0.26 0.21 0.04 0.19 1.3 84%
τ (380 nm) 456 0.34 0.28 0.05 0.25 1.69 82%
α 456 1.1 0.38 0.15 1.13 1.952 34%
τF 453 0.18 0.21 0.01 0.111 1.25 111%
τC 453 0.06 0.04 0.01 0.06 0.34 64%
FMF 453 0.65 0.19 0.18 0.66 0.97 30%
Remote Sens. 2023, 15, 501 6 of 16

Eight islands around the world are shown in Figure 4 to compare aerosol loads. These
islands or AERONET stations cover the Pacific (American_Samoa, 14.247S, 170.564W; Oki-
nawa_Hedo, 26.867N, 128.249E), Indian (Amsterdam_Island, 37.800S, 77.572E), Atlantic
(ARM_Graciosa, 39.091N, 28.029W; Ascension_Island, 7.976S, 14.415W), Mediterranean
(Lampedusa, 35.517N, 12.632E) and SCS (Dongsha and Taiping Islands). The mean τ
(500 nm) values over the South China Sea are significantly higher than those over the
open oceanic areas (American Samoa, Amsterdam Island, and ARM_Graciosa) without
the influence of long-distance transport. Additionally, the α (440–870 nm) and FMF values
obtained in this study were smaller than those of open oceanic areas for maritime aerosols.
Later in the article, we compare the Dongsha observations with those of three nearby
AERONET sites during the same period. Pure maritime situations can be generally found
when τ (500 nm) < 0.15 and α (440–870 nm) < 1 according to Smirnov et al. [12]. Based on
this evaluation criterion, 16 percent of the analyzed days on Dongsha Island were pure sea
conditions, the least of all the islands. Clean maritime conditions observed over Taiping
Remote Sens. 2023, 15, x FOR PEER REVIEW 7 of 17
Island were more frequent than those observed over Dongsha Island and were still far
less frequent than in open oceanic areas. In addition, Okanaw_Hebe, Lampedusa and
Ascension_Island show similar
Table 2. Multi-wavelength (1020, 500performance
and 380 nm)inaerosol
marine conditions.
optical depth τ,Marine
Ångström areas that are
exponent (α
mostly
(440–870)), fine aerosol optical depths at 500 nm (τ (500 nm), coarse aerosol optical depths at τCfrom
surrounded by the Baltic, the Mediterranean
F and Sea of Japan were excluded (500
the
nm))pure
and maritime research
fraction of optical [29].ofThe
depth fine SCS
modeisparticles
a marginal
(FMF)seainin southern
Taiping fromChina surrounded
1 January 2018 to 31
by coastal cities and regions in southeast China, the Philippines, the
December 2020; All data were processed using daily averages. The total number of samples Greater Sunda Islands
(N to-
and
tal), the Indochina
average Peninsula.
coefficient (Mean), Dongsha Island and
standard deviation Taiping
(SD), Minimum Island are not
(Min), directly
Median affected
(Med), Maxi-
mum (Max) and the coefficient of fluctuation (FOC) are also counted to assess
by human activities, but air masses reaching the two islands may be influenced by anthro- variations in the data.
pogenic aerosols during their passage over the SCS and continents. Figure 3 (a2 ,b2 ) shows
N Total Mean SD Min Med Max FOC
the aerosol properties in Taiping. The mean τ (500 nm) value of 0.17 and α (440–870 nm)
valueτof (1020
0.96 nm)
over Taiping551 Island were 0.09 0.06 lower0.01
significantly than the mean0.08 τ (5000.7nm) value
65% of
τ (500 nm) 551 0.17 0.13 0.04
0.26 and α (440–870 nm) value of 1.2 over Dongsha Island. Additionally, the average 0.13 1.85 79%FMF
valueτof(380 nm) Island was
Taiping 551 lower 0.21 than that 0.16
of Dongsha 0.04 0.17 2).2.14
Island (Table 78%
The difference
α
in the presence of aerosols 551at these 0.96 0.36be explained
two sites can 0.11 by 0.98 2.03 locations.
their different 37%
τ F 450 0.09 0.13 0.01 0.06
Taiping is farther away from land than Dongsha, and the surrounding air mass will carry 1.8 146%
more sea saltτC aerosol particles,
450 increase 0.06 the proportion
0.03 0.01
of 0.06
coarse particles, 0.24
and reduce 54%the
optical depth.
FMF 450 0.53 0.17 0.16 0.53 0.982 32%

(a) 0.6 2.0 1.0

t (500 nm) South China Sea
0.5 a 1.6 0.8
Fraction
0.4
ARM_Graciosa
Okanawa_Hebe
Amsterdamn_Island

Fraction
1.2 0.6
Dongsha_Island
Ascension_Island

Lampedusa
t

Taiping

0.3
a
American_Samoa

0.8 0.4
0.2
0.1 0.4 0.2

(b) 0.00 0.0 0.0

100
200
Number

300
400
Marine Days
500
Total Days
600
-50 -40 -30 -20 -10 0 10 20 30 40 50
Latitude (o)

Figure
Figure 4.4. Global
Global annual average, meridional distributions of (a) ττ at 500 nm, α
α (440–870 nm), FMF;
(b) the
(b) the total
total number
number of
of samples
samples and
and pure
pure maritime
maritimedays.
days.

As shown in Figure 3(a1,b1), Dongsha Island is strongly affected by aerosols for sev-
eral days each year, with τ (500 nm) values exceeding 0.35. High aerosol loads (τ (500 nm)
> 0.35) [28] over Dongsha Island were observed for 90 out of 449 days of analysis. These
events were observed on 66 analysis days from March to April. In these cases, the 60-day
α (440–870 nm) value was higher than 1.1, and the FMF value was higher than 0.7, indi-
Remote Sens. 2023, 15, 501 7 of 16

Table 2. Multi-wavelength (1020, 500 and 380 nm) aerosol optical depth τ, Ångström exponent
(α (440–870)), fine aerosol optical depths at 500 nm (τF (500 nm), coarse aerosol optical depths at
τC (500 nm)) and fraction of optical depth of fine mode particles (FMF) in Taiping from 1 January
2018 to 31 December 2020; All data were processed using daily averages. The total number of
samples (N total), average coefficient (Mean), standard deviation (SD), Minimum (Min), Median
(Med), Maximum (Max) and the coefficient of fluctuation (FOC) are also counted to assess variations
in the data.

N Total Mean SD Min Med Max FOC

τ (1020 nm) 551 0.09 0.06 0.01 0.08 0.7 65%
τ (500 nm) 551 0.17 0.13 0.04 0.13 1.85 79%
τ (380 nm) 551 0.21 0.16 0.04 0.17 2.14 78%
α 551 0.96 0.36 0.11 0.98 2.03 37%
τF 450 0.09 0.13 0.01 0.06 1.8 146%
τC 450 0.06 0.03 0.01 0.06 0.24 54%
FMF 450 0.53 0.17 0.16 0.53 0.982 32%

As shown in Figure 3(a1 ,b1 ), Dongsha Island is strongly affected by aerosols for several
days each year, with τ (500 nm) values exceeding 0.35. High aerosol loads (τ (500 nm) >
0.35) [28] over Dongsha Island were observed for 90 out of 449 days of analysis. These
events were observed on 66 analysis days from March to April. In these cases, the 60-day α
(440–870 nm) value was higher than 1.1, and the FMF value was higher than 0.7, indicating
the dominant role of fine particles. In the remaining high aerosol loading events, high
aerosol loading was associated with relatively low α (440–870 nm) values, reaching the
lowest α (440–870 nm) values (about 0.83). During these days, the FMF values were also
low and reached the lowest mean daily value of 0.52. This behavior indicates that coarse
particles mainly transported from other regions, except for sea salt aerosol, as there is no
local coarse particle activity in Dongsha. According to the analyses of back trajectories
during high aerosol loading days (Figure 5), all of these cases were linked to anthropogenic
and dust intrusion from the southeastern coast of China and Mongolia. The important
thing to note about these events is that τF (500 nm) was also relatively high and ranged
from 0.17 to 1.25 with a mean value of 0.56 ± 0.26. These results highlight the significant
contribution of fine model particles, with FMF ranging from 52% to 97% during these high
aerosol loading events. A backward trajectory analysis of high aerosol loads on these days
shows that the air mass that originated over the Taiwan Strait (57%) and Mongolia (17.12%)
in the study area was at a lower altitude (500 m). However, compared with Dongsha Island,
high aerosol loads (τ (500 nm) > 0.35) over Taiping Island can still be observed on 19 of
the 450 analysed days, and occurred from August to September, especially in 2018 and
2019. From 10 August to 30 September, the back trajectories in Taiping Island were almost
related to high aerosol loads from the region between Singapore and Indonesia. On these
days, high aerosol loads were associated with relatively high α (440–870 nm) values, which
peaked at about (2.14).
The values of τF (500 nm) were also high (>0.19), and the highest value was 1.8. Since
there is no significant anthropogenic activity in Taiping, it is likely that the fine particulate
matter in these cases originated mainly from mainland industrial/urban areas. Similarly,
the α (440–870 nm) value of Taiping Island was less than 0.9 on two days of high aerosol
loading days. However, α (440–870 nm) values were relatively low (<0.88) with high τF
(500 nm) values (approximately 0.22). These behaviors suggest that coarse particles were
transported from other locations, as there was no local coarse particle activity in Taiping
in addition to sea salt aerosols. In this event, the high τ values observed were associated
with a sustained monsoon climate centered in the SCS, which favours the transport of
anthropogenic particles emitted from Southeast Asia to the SCS. According to the EAC,
the aerosol optical depths of black carbon aerosol (BC), dust aerosol (DU), organic matter
aerosol (OM), sulfate aerosol (SU) and sea salt aerosol (SS) in Dongsha and Taiping are

Remote Sens. 2023, 15, 501
loading days. However, α (440–870 nm) values were relatively low (<0.88) with high τF
(500 nm) values (approximately 0.22). These behaviors suggest that coarse particles were
transported from other locations, as there was no local coarse particle activity in Taiping
in addition to sea salt aerosols. In this event, the high τ values observed were associated
with a sustained monsoon climate centered in the SCS, which favours the transport of
anthropogenic particles emitted from Southeast Asia to the SCS. According to the EAC,8 of 16
the aerosol optical depths of black carbon aerosol (BC), dust aerosol (DU), organic matter
aerosol (OM), sulfate aerosol (SU) and sea salt aerosol (SS) in Dongsha and Taiping are
shown
shownininFigure
Figure6.6.It Itreveals that
reveals high
that aerosol
high loading
aerosol in Dongsha
loading and and
in Dongsha Taiping Island
Taiping was was
Island
caused by the sharp increase in black carbon aerosols, dust aerosols and organic aerosols
caused by the sharp increase in black carbon aerosols, dust aerosols and organic aerosols in
in addition to sea salt aerosols. Figure 7 shows the distribution of BC in Dongsha and OM
addition to sea salt aerosols. Figure 7 shows the distribution of BC in Dongsha and OM in
in Taiping and its surrounding areas before and after the high aerosol loading in 2019. BC
Taiping and its surrounding areas before and after the high aerosol loading in 2019. BC
and OM are typical urban aerosols. The appearance of these two aerosols confirms the
and OM are typical urban aerosols. The appearance of these two aerosols confirms the
intrusion from foreign air masses in the process of high aerosol loading. Therefore, the
intrusion from foreign air masses in the process of high aerosol loading. Therefore, the
high loading of fine particles in Dongsha mainly originated from BC and OM brought
high loading of fine particles in Dongsha mainly originated from BC and OM brought from
from the southeastern coast of China. A small amount of coarse DU particles were loaded
the southeastern coast of China. A small amount of coarse DU particles were loaded from
from Mongolia. In addition, the high fine particle loading in Taiping was mainly produced
Mongolia.
by OM andIn BC,addition,
which came the high
fromfine particleAsia,
Southeast loading in Taiping
including was mainly
Singapore and theproduced
region by
OM and BC, which came
between Singapore and Indonesia.from Southeast Asia, including Singapore and the region between
Singapore and Indonesia.

Dongsha

Taiping

Figure
Remote Sens. 2023, Figure
15, Five-day
5.5.Five-day
x FOR PEER trajectories
trajectories
REVIEW ending
ending at Dongsha
at Dongsha (15 (15 March
March to 20toApril
20 April every
every year)year) and Taiping
and Taiping (10 (10 9 o
August
Augustto
to30
30September
Septemberevery
everyyear) at at
year) 500 mm
500 during high
during aerosol
high loading.
aerosol loading.

BC
0.2 Dongsha Taiping DU
High
0.1 Aerosol
Loads
0.0
AOD (550nm)

1.5 OM
SU
1.0

0.5

0.0
SS
0.2

0.1

0.0
/1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1
8/1 018/7 019/1 019/7 020/1 020/7 021/1 018/1 018/7 019/1 019/7 020/1 020/7 021/1
201 2 2 2 2 2 2 2 2 2 2 2 2 2
Day Day

Figure 6. Time evolution of AOD

Figure 6. Time at 550 nm
evolution of BC,
of AOD at DU, OM,
550 nm ofSU
BC,and
DU,SSOM,
obtained
SU andfrom EAC. from EAC.
SS obtained

3.2. Back Trajectories

(a) and Aerosol Optical Properties over Dongsha
(b) Island and Taiping Island
The Hysplit Dispersion Model was used to calculate the 120-h backward trajectory
ending at Dongsha and Taiping to determine the source of the air mass (Figure 8). Each
trajectory was linked to the corresponding aerosol optical depths. Previously, this method
 0.5

A
0.0
SS
0.2

0.1
Remote Sens. 2023, 15, 501 9 of 16
0.0
/1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1 /1
8/1 018/7 019/1 019/7 020/1 020/7 021/1 018/1 018/7 019/1 019/7 020/1 020/7 021/1
201 2 2 2 2 2 2 2 2 2 2 2 2 2
has been Day
Day widely used to evaluate the distribution characteristics of aerosol optical depths
of different
Figure 6. Timeair massesof[30,31].
evolution AOD at 550 nm of BC, DU, OM, SU and SS obtained from EAC.

(a) (b)

(c) (d)

Remote Sens. 2023, 15, x FOR PEER REVIEW 10 of 17

Figure
Figure 7.7.Before
has been Before(a)(a)
widely and
used
andafter
to (b)
after thethe
sharp
evaluate
(b) theincrease in black
distribution
sharp increase carbon aerosol
in characteristics
black carbon ofloading
aerosol inoptical
aerosol theinDongsha
loading thedepths
Dongsha
area; before (c)
of different
area; and
before (c)air
and after
masses(d) the sharp
[30,31].
after (d) increase in organic matter aerosol loading in the Taiping
the sharp increase in organic matter aerosol loading in the Taiping area. area.

3.2. Back Trajectories and Aerosol Optical Properties over Dongsha Island and Taiping Island
The Hysplit Dispersion Model was used to calculate the 120-h backward trajectory
ending at Dongsha and Taiping to determine the source of the air mass (Figure 8). Each
trajectory was linked to the corresponding aerosol optical depths. Previously, this method

Figure8.8.Five-day
Figure Five-daytrajectories
trajectoriesending
endingatat500
500mmfrom
fromJanuary
January2018
2018totoJanuary
January2021
2021atat(a)
(a)Dongsha
Dongshaand
and (b) Taiping; The color of the trajectory represents the value of aerosol optical depth. One trajec-
(b) Taiping; The color of the trajectory represents the value of aerosol optical depth. One trajectory
tory per day was displayed; The different source areas are separated into four sections (A, B, C and
per
D),day was displayed;
as shown The different source areas are separated into four sections (A, B, C and D), as
in the panel.
shown in the panel.
Broad geographic areas are divided based on the possibility of determining the major
sources of aerosol type and optical depth. In this way, the definitions of the A, B, C and D
sectors are constrained by optical depth in the different areas. The large degree of arbi-
trariness may stem from the choice of borders. By considering τ and the different sectors
Remote Sens. 2023, 15, 501 10 of 16

Broad geographic areas are divided based on the possibility of determining the major
sources of aerosol type and optical depth. In this way, the definitions of the A, B, C
and D sectors are constrained by optical depth in the different areas. The large degree
of arbitrariness may stem from the choice of borders. By considering τ and the different
sectors along the trajectory, the sector of origin of the aerosol can be determined. We defined
four broad geographical sectors, displayed in Figure 8, in relation to different aerosol optical
depths; four broad geographical sectors are defined in Figure 8. The identified sectors
at Dongsha were: (A) North of Dongsha Island, where aerosols mainly come from the
southeast coast of China (46%); (B) The western sector (36%), which includes the China-
Indochina Peninsula and the SCS; in this region, the prevalent sources produce continental
and, to a lesser extent, marine aerosols; (C) the southeast sector (17%), coinciding with the
Luzon Strait and the SCS; (D) other parts had no obvious boundary (1%), and the proportion
was negligible. The identified sectors at Taiping were: (A) the southwestern portion of
Taiping Island (21%), which is largely bounded by Singapore and Indonesia, together which
provides an important source of urban aerosols; (B) an east-north sector (55%), coinciding
with the Philippines; in this region, the prevalent sources produce continental aerosols and,
to a lesser extent, marine aerosols; (C) an East-South sector (10%), which included the Sulu
Sea; (D) Other parts (14%), which accounted for less and have unclear boundaries, were
not analyzed in this study.
The source distribution of the air mass in the Dongsha area will be described first. The
high aerosol loads were well identified from the trajectory analysis and correspond to class
A. As we mentioned earlier, aerosols from sector A had the largest average aerosol optical
depth (0.35 ± 0.23) (Figure 9(a1 )). In addition, 80% of α values in class A were higher than 1
(Figure 9(b1 )). Sector A mainly originates from the southeast coast of China. As mentioned
above, this part of the air mass mainly contains polluting fine aerosol loading. Sector C
particles had the smallest τ (0.14 ± 0.11) and α ranging from 0.15 and 1.95. This means
that part C mainly included clean aerosols, with a relatively uniform distribution of coarse
and fine particles. Sector B includes the China-Indochina Peninsula and the SCS. In this
region, the possible sources produce continental and, to a lesser extent, marine aerosols.
The optical depth value of part B was lower than that of the polluting source of part A
and higher than that of the cleaning source of part C. In part B, the central value (xc ) of α
was smaller than 1, indicating that marine aerosols bring about the significant growth of
coarse particles. It must be emphasized that different aerosol types may be present in the
air column simultaneously, influencing the observed optical parameters.
Interestingly, the relationship between the aerosol optical depth and wavelength index
in different parts of the Taiping area has a similar distribution to that in the Dongsha area
(Figure 9(a2,b2)). Similar to sector A of the Dongsha area discussed above, some of these
cases are due to the intrusion of anthropogenic aerosols in the southwest, which produced
high aerosol loading. In Part B, due to the addition of marine aerosols, the loading of coarse
particles increases, and in Part C, there are mainly clean aerosols.
In summary, (1) the pollution aerosol in the Dongsha area mainly originates from
the southeast coast of China, which mainly contains some fine particles; (2) the mixing of
marine aerosols in the southwest of the Dongsha area increased the proportion of coarse
particles; (3) the pollution aerosol in the Taiping area mainly originates from the area
between Indonesia and Singapore; (4) The source of marine aerosol on Taiping Island may
be more from the northeast direction; and (5) a small number of clean aerosol sources enter
from the southeast of the two places, and the coarse and fine particles of this part of the
aerosol were evenly distributed.

3.3. Seasonal Properties of Aerosol over Dongsha Island and Taiping Island
Monthly mean values of τ (500 nm), τF (500 nm), τC (500 nm), α (440–870 nm), FMF
and their corresponding standard deviations in Dongsha and Taiping from January 2018
to December 2020 are shown in Figure 10a,b. The monthly average value is derived from
the daily average data. The large values of τ (500 nm), indicating the high aerosol load,
Remote Sens. 2023, 15, 501 11 of 16

Remote Sens. 2023, 15, x FOR PEER REVIEW 11 of 17

were observed in March-April, while the minimum (0.093–0.096) was observed in June-July
(Figure 10(a1 )). The monthly value of α (440–870 nm) was always approximately 1, and
all mean values of FMF were higher than 0.5, reflecting a relatively high fine particle
of coarse and fine particles. Sector B includes the China-Indochina Peninsula and the SCS.
loading
In in each
this region, themonth,
possibleespecially in spring
sources produce and autumn
continental and,(Figure 10(bextent,
to a lesser 1 )). For summer
marine
(May-August), the monthly mean values of mean α (440–870 nm) gradually
aerosols. The optical depth value of part B was lower than that of the polluting source of decreased
fromA1.17
part andtohigher
0.87, and
thanthe FMF
that decreased
of the cleaningfrom 0.62oftopart
source 0.52,C.indicating
In part B, an
theincrease in coarse
central value
particle
(xc) of αcontribution
was smaller during
than 1, these months
indicating that(Figure
marine10(b 1 )). It is
aerosols alsoabout
bring worththenoting that all τF
significant
(500
growthnm) ofshowed a pronounced
coarse particles. It must decrease in summer,
be emphasized suggesting
that different decreased
aerosol types may loads of fine
be pre-
particles (Figure 10(a )).
sent in the air column simultaneously, influencing the observed optical parameters.
1

(a1) Dongsha (b1) Dongsha A

50
60 B
A：xc= 0.25,w = 0.2 A：xc= 1.14,w = 0.5
Frequency (%)

Frequency (%)
40
B：xc= 0.94,w = 0.88
C
B：xc= 0.17,w = 0.11
40 C：xc= 0.09,w = 0.13 C：xc= 1.3,w = 0.77
30

20
20
10

0 0
80 (a2) Taiping (b2) Taiping
50
A：xc= 0.17,w = 0.34 A：xc= 1.07,w = 0.56
Frequency (%)

Frequency (%)

60 B：xc= 0.14,w = 0.12 40 B：x = 0.93,w = 1.01

c
C：xc= 0.07,w = 0.11 C：xc= 1.25,w = 0.62
30
40
20

20
10

0 0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.0 0.5 1.0 1.5 2.0

t a
Figure 9.
Figure Distribution
9.Distribution of of
thethe frequency
frequency of occurrences
of occurrences of (a1of
,a2)(aτ,1 ,a2 ) τ,
and (b1and (bin
,b2) α 1 ,b2 ) α sectors:
three in threeA,sectors:
B,
C. These
A, data are
B, C. These shown
data are in black in
shown (sector
blackA), blue (sector
(sector B) (sector
A), blue and redB) (sector C). The
and red distribution
(sector C). The ofdistri-
different
bution of different sectors (colored lines) was identified by Gaussian fitting (fitted line, y =+y0 +
sectors (colored lines) was identified by Gaussian fitting (fitted line, y = y0
(A/(w*sqrt(pi/2)))*exp(−2*((x
(A/(w*sqrt(pi/2)))*exp(−2*((x − xc)/w)^2)). The center
− xc)/w)ˆ2)). valuevalue
The center (xc) and (xc)waist widthwidth
and waist (w) of (w)
eachofpart
eachofpart
the Gaussian function are shown.
of the Gaussian function are shown.

Interestingly,
Figure 11 shows the relationship
the monthlybetween the aerosol
wind direction optical depth
probability obtainedand at wavelength
Dongsha in- Island
dex
and in different
Taiping parts
Island of the
from 2018Taiping
to 2020.area
In has
summera similar distributionthe
(May–August), to that
south in wind
the Dongsha
prevails in
area (Figure 9(a2,b2)).
the Dongsha area, while Similar to sector wind
the northeast A of the Dongsha
prevails areaseasons.
in other discussed Inabove, some of
fact, northeasterly
these cases
winds are due
will bring to the
more intrusion of anthropogenic
anthropogenic pollution sources aerosols
(OC, inBC,the southwest,
etc.) and dustwhichaerosols,
produced high aerosol loading. In Part B, due to the addition of marine
while in summer, there were no significant pollution intrusions. Additionally, the summer aerosols, the load-
ing of coarse
monsoon andparticles
southwestincreases,
oceanand in Part
current C, there
over are (Figure
the SCS mainly clean aerosols.
12) during this season may
favour the accumulation of marine aerosols that can explain the coarseoriginates
In summary, (1) the pollution aerosol in the Dongsha area mainly from during
particle loads the
southeast coast of China, which mainly contains some fine particles;
summer in comparison with spring and autumn. In addition, the aerosols show lower fine (2) the mixing of ma-
rine aerosols
particle in the
loading andsouthwest of the Dongsha
higher coarse area increased
particle loading in winterthe (Figure
proportion 10(aof coarse par-
1 )). At the same
ticles; (3) the pollution aerosol in the Taiping area mainly originates
time, the monthly mean α (440–870 nm) of aerosols in winter was not more than from the area between
1, and the
Indonesia and Singapore; (4) The source of marine aerosol on Taiping
mean FMF value was approximately 0.5, indicating that the Dongsha area in winter Island may be more has a
from the northeast direction; and (5) a small number of clean aerosol sources enter from
higher loading of coarse particles than that in other seasons. In this sense, the high mass of
the southeast of the two places, and the coarse and fine particles of this part of the aerosol
sea salt aerosols observed in winter was associated with the increase in the significant height
were evenly distributed.
of combined wind waves and swell (SWH) (see, for example, Figure 10(c1 )). The low aerosol
loads registered in winter can be explained by the (Figure 10(c )) high sea salt aerosol loads
3.3. Seasonal Properties of Aerosol over Dongsha Island and Taiping 1Island
and the absence of pollution intrusions in this period. Figure 10(a2 ,b2 ,c2 ) shows the monthly
meanMonthly
values of mean values
τ (500 nm), ofτF τ(500
(500nm)
nm), τF τ(500
and nm),
C (500 nm)τC as
(500 nm),
well as αα (440–870
(440–870 nm) nm),andFMFFMF
and their corresponding standard deviations in Dongsha and Taiping
with the corresponding standard deviations at Taiping from January 2018 to December from January 20182020.
to December 2020 are shown in Figure 10a,b. The monthly average value
The largest values of τ (500 nm) were observed during August-September, while the lowest is derived from
the daily average data. The large values of τ (500 nm), indicating the high aerosol load,
values (0.06–0.08) were measured from October to January (Figure 10(a2 )). According to the
were observed in March-April, while the minimum (0.093–0.096) was observed in June-
Remote Sens. 2023, 15, 501 12 of 16

Remote Sens. 2023, 15, x FOR PEER REVIEW 12 of 17

conclusion in Section 3.1, the high aerosol loading in August and September was mainly
due to the high polluting aerosol brought by the southwest air mass. In addition, the
July (Figure 10(a1)). The monthly value of α (440–870 nm) was always approximately 1,
pollution transport from the middle sea area surrounded by Singapore and Indonesia may
and all mean values of FMF were higher than 0.5, reflecting a relatively high fine particle
be brought with the southwest wind at Taiping Island (Figure 11). There is no doubt that in
loading in each month, especially in spring and autumn (Figure 10(b1)). For summer (May-
winter, strong waves bring more sea salt aerosols in the Taiping area and the Dongsha area,
August), the monthly mean values of mean α (440–870 nm) gradually decreased from 1.17
making the aerosol optical depth at low values. Therefore, we can conclude that the unique
to 0.87, and the FMF decreased from 0.62 to 0.52, indicating an increase in coarse particle
meteorological characteristics and large-scale circulation in the South China Sea bring about
contribution during these months (Figure 10(b1)). It is also worth noting that all τF (500
different aerosol sources in the surrounding waters and then drive the aerosol properties of
nm) showed a pronounced decrease in summer, suggesting decreased loads of fine parti-
Dongsha Island and Taiping Island to show obvious monthly variation characteristics.
cles (Figure 10 (a1)).

-ter
-ter Spring Summer Autumn Win -ter Spring Summer Autumn Win
(a1) (a2) t (500nm)
Dongsha Taiping tF(500nm)
0.6
0.4 tC(500nm)
τ

0.3

0.0 0.0
2 2 1.0
(b1) (b2) a
α (400-870nm)

FMF

FMF
1 1 0.5

Remote Sens. 2023, 15, x FOR PEER REVIEW

0 13 of 17
0 0.0
3 3
4 (c1) (c2) SWH

SS (10-4kg/m2)
SS
SWH (m)

2 2 2
from2 October to January (Figure 10(a2)). According to the conclusion in Section 3.1, the
high aerosol loading in August and September 1 was mainly due to the high polluting 1 aer-
osol brought by the southwest air mass. In addition, the pollution transport from the mid-
0 0 0 0
dle sea Febarea surrounded
Apr Jun by
AugSingapore
Oct Decand Indonesia Feb may
Apr beJun brought Aug with Oct the Dec
southwest
wind at Taiping Island Month(Figure 11). There is no doubt that inMonth winter, strong waves bring
more sea salt aerosols in the Taiping area and the Dongsha F area, making
C (500 the aerosol opti-
Figure10.
Figure Monthly
10.Monthly mean
mean values
values of1,a(a2)1 ,a
of (a ) τ (500
τ 2(500 nm), nm),
τ (500
F (500τnm),
τ nm), C (500τnm), (bnm),
1,b2) α(b 1 ,b2 ) α (440–
(440–870
cal
nm),
depth
(c1,c2(c
at
) the
low values.
significant
Therefore,
height of
we
combined
can conclude
wind waves
that
and
the
swell
unique
(SWH)
meteorological
and sea salt
char-
aerosol
870 nm), ,c
1 2 ) the significant height of combined wind waves and swell (SWH) and sea salt aerosol
acteristics
mass and large-scale circulation in the South China Sea bring about different aerosol
massconcentration
concentrationFMF FMF ininDongsha
Dongsha and andTaiping
Taiping from January
from January20182018to December
to December 2020.2020.
The error
The error
sources
bars in the surrounding
are standard deviations. waters and then drive the aerosol properties of Dongsha Island
bars are standard deviations.
and Taiping Island to show obvious monthly variation characteristics.
Figure 11 shows the monthly wind direction probability obtained at Dongsha Island
Dongsha Taiping
and Taiping Island from 2018 to 2020. In summer (May–August), the south wind prevails
N N N
Mar N
in the Jau
DongshaNEarea, while
60%
the northeast
45%Feb wind prevails in other 45%
seasons. In fact,
45% Apr
northeast-
NW NW NE NW NE NW NE
40% 30% 30% 30%
erly winds will bring more anthropogenic pollution sources (OC, BC, etc.) and dust aero-
20% 15% 15% 15%

sols,
W while in summer, E 0% W there were no 0% E significant
W pollution
0%E intrusions.
W Additionally,
0% E the
summer monsoon and southwest ocean current over the SCS (Figure 12) during this sea-
sonSWmay favour SE the accumulation
SW of marine
SE SW aerosols thatSE can explain
SW the coarse
SE particle
loads during S summer in comparison S with springS and autumn. In addition, S the aerosols
N N N N
show
NW
lower fine NE
30%particle
May
NW
loading NEand45%Jun higher coarse particle
NW
45%
NE
Jul loading
NW
in winter
45%
NE
Aug (Figure
20% 30% 30% 30%
10(a1)). At the same time, the monthly mean α (440–870 nm) of aerosols in winter was not
10% 15% 15% 15%

more
W than 1, andEthe mean 0% W FMF valueE wasWapproximatelyE0.5, indicating
0% 0% W that the
0% E Dongsha
area in winter has a higher loading of coarse particles than that in other seasons. In this
sense,
SW the highSE mass ofSWsea salt aerosols SE observed
SW in winter
SE was SW associated SE with the in-

crease in Sthe significant height S of combined wind Swaves and swell (SWH) S (see, for exam-
N N N N
ple,NWFigure 10(cNE1)).Sep
30% The low
NW
aerosol NE loads
60% Oct registered in winter
NW
60%
NE
Nov can be explainedDec
NW
60%
NE
by the (Fig-
ure 10(c1)) high sea salt aerosol loads and the absence of pollution intrusions in this period.
20% 40% 40% 40%

10% 20% 20% 20%

Figure
W 10(a2,b2,c2E) shows
0% W the monthlyE mean
0% W values of τ (500 0%E nm),
W τF (500 nm)Eand τC (500
0%

nm) as well as α (440–870 nm) and FMF with the corresponding standard deviations at
Taiping
SW from January
SE 2018
SW to December
SE 2020.
SW The largest
SE values SW of τ (500 SE nm) were ob-
served duringS August-September, S while the lowest S values (0.06–0.08) S were measured
Figure Monthlywind
Figure 11. Monthly windrose
rosegraphics
graphicsobtained
obtainedat at Dongsha
Dongsha andand Taiping
Taiping island
island for for three
three years.
years.

(a) (b)
 N N N N
30% Sep 60% Oct 60% Nov 60% Dec
NW NE NW NE NW NE NW NE
20% 40% 40% 40%

10% 20% 20% 20%

W 0% E W 0% E W 0% E W 0% E

SW SE SW SE SW SE SW SE

Remote Sens. 2023, 15, 501 S S S S 13 of 16

Figure 11. Monthly wind rose graphics obtained at Dongsha and Taiping island for three years.

(a) (b)

Remote Sens. 2023, 15, x FOR PEER REVIEW

Figure 12. Distribution
Figure 12. of wind field,
Distribution monsoon
of wind field, and oceanand
monsoon current
ocean in the SCS.
current in the(a) Summer
SCS. (b)14Winter.
(a) Summer
of 17
(b) Win-
ter.
3.4. Regional Properties of Aerosol over the South China Sea
3.4. Regional
In this study, Properties of Aerosol
the AERONET data over the South China
of Dongsha Sea Taiping Island and their sur-
Island,
In this study,
rounding four AERONET the AERONET
stations data 1)
(see Figure of Dongsha Island, Taiping
from 1 January 2018 toIsland and their sur-
31 December 2020
were used torounding four AERONET
study regional stations
properties of (see Figure
aerosol 1) from
over SCS1 (Figure
January 2018
13). toTo31analyze
December 2020
aerosol
were used to study regional properties of aerosol over SCS (Figure 13). To analyze aerosol
regional variability, we compared nearby sites using only time-consistent aerosol data
regional variability, we compared nearby sites using only time-consistent aerosol data
(Table 3). Figure
(Table13a
3). shows comparisons
Figure 13a of the daily
shows comparisons average
of the daily τ (500
average nm)
τ (500 nm)atatDongsha Island
Dongsha Island
and Hong Kong Island
and Hong from
Kong 1 January
Island 20182018
from 1 January to 31 to December
31 December2020. Hong
2020. Hong Kong
Kong is about
is about 315
315 km fromkm Dongsha Island.Island.
from Dongsha The temporal
The temporalvariation
variation in in
thethedaily
dailymean
mean of τ (500
(500nm)nm)was was
similar
similar for both sitesforonboth sites
most on most
days days
during theduring the analysis
analysis period,period, indicating
indicating that processes
that the the pro-
cesses controlling
controlling aerosol loading ataerosol
both loading at both
sites were sites were
similar. The similar. The correlation
correlation coefficient coefficient
R between R
between these two sites is 0.87. However, there are also large differences on some days
these two sites is 0.87. However, there are also large differences on some days (for example,
(for example, τ (500 nm) at Dongsha Island was 0.05 on 1 July 2019, compared to 0.22 in
τ (500 nm) atHong
Dongsha
Kong).Island was 0.05 on 1 July 2019, compared to 0.22 in Hong Kong).

2.4
(a) Dongsha (d) Taiping
Hongkong Nha Trang
1.8

1.2

0.6

0.0
2.4
(b) Dongsha (e) Taiping
1.8 NSPO Singapore
t

1.2

0.6

0.0
1

2.4
1/

7/

1/

7/

1/

7/

1/

(c) Dongsha
Taiping Date
1.8

1.2

0.6

0.0
1

1
1/

7/

1/

7/

1/

7/

1/

Date
Figure 13. Comparisons of τ (500 nm):
Figure 13. Comparisons (a) nm):
of τ (500 Dongsha IslandIsland
(a) Dongsha and NSPO, (b)(b)
and NSPO, Dongsha
DongshaIsland andHong
Island and Hong
Kong, (c) Dongsha Island and Taiping Island, (d) Taiping Island and Nha Trang, and (e) Taiping
Kong, (c) Dongsha Island and Taiping Island, (d) Taiping Island and Nha Trang, and (e) Taiping
Island and Singapore.
Island and Singapore.
Table 3. Mean value of τ, α (440–870 nm), τF (500 nm), τC (500 nm) and FMF obtained over Dongsha
(DS), NSPO (NS), Hong Kong (HK), Taiping (TP), Nha Trang (NT) and Singapore (SP). Here, the
same time series of stations are compared with Dongsha Island and Taiping Island from 1 January
2018 to 31 December 2020.

DS HK DS NS DS TP TP NT TP SP
Remote Sens. 2023, 15, 501 14 of 16

Table 3. Mean value of τ, α (440–870 nm), τF (500 nm), τC (500 nm) and FMF obtained over Dongsha
(DS), NSPO (NS), Hong Kong (HK), Taiping (TP), Nha Trang (NT) and Singapore (SP). Here, the
same time series of stations are compared with Dongsha Island and Taiping Island from 1 January
2018 to 31 December 2020.

DS HK DS NS DS TP TP NT TP SP
τ (1020 nm) 0.09 0.17 0.11 0.13 0.13 0.08 0.08 0.12 0.09 0.14
τ (500 nm) 0.21 0.44 0.25 0.35 0.31 0.16 0.14 0.27 0.17 0.37
τ (380 nm) 0.29 0.56 0.33 0.46 0.41 0.19 0.18 0.38 0.21 0.51
α 1.17 1.27 1.02 1.25 1.14 0.99 0.90 1.20 0.94 1.31
τF 0.15 0.37 0.18 0.31 0.24 0.10 0.08 0.21 0.10 0.30
τC 0.06 0.06 0.06 0.04 0.06 0.05 0.06 0.06 0.06 0.06
FMF 0.66 0.77 0.63 0.8 0.68 0.57 0.52 0.69 0.55 0.74
Coincident day 212 212 197 197 210 210 154 154 310 310

Differences in the timing and intensity of anthropogenic aerosol intrusion into these
sites are most likely responsible for the differences in aerosol loading in these regions. In
fact, the correlation in τF (500 nm) between Dongsha Island and Hongkong, R, is 0.82. The
aerosol properties in the Dongsha area and NSPO area were also compared, and similar
high correlation results were obtained (R = 0.73 for τF (500 nm)). From long wavelength
(1020 nm) to short wavelength (380 nm), the difference between Dongsha Island and the
other two places was becoming more and more obvious (Table 3). This is consistent with
the previous conclusion that aerosol loading in the Dongsha area is mainly affected by
fine particles in the southeast coast of China. On the other hand, for Taiping Island, the
variation of optical depth in Taiping island was similar to those in Nha Trang (Figure 13d)
and Singapore (Figure 13e). There is no doubt that the South China Sea region shows a
similar coarse particle loading, and the coarse particle optical depth of the whole study
area was maintained at about 0.06. In addition, the aerosol loading in Taiping area is
significantly weaker than that in Dongsha Island, but the proportion of coarse particles
(FMF = 0.57) was significantly higher than that in Dongsha Island (FMF = 0.68). Thus, we
can conclude that the aerosols in the South China Sea were greatly affected by the aerosols
in the surrounding urban areas, and the variation was very consistent with surrounding
urban areas, but the aerosol loading was far lower than that in the coastal cities.

4. Conclusions
With unbalanced socio-economic development and a complex meteorological system,
the aerosol pollutants in the South China Sea are of various types. To investigate the change
of columnar aerosol properties over time and space in this little-studied region, aerosol
optical depths obtained over Dongsha Island, Taiping Island, and four other nearby sites in
the SCS were analysed. Within the analysed period, the mean of observed τ (500 nm) values
over Dongsha Island and Taiping Island were significantly higher than those reported in
the open ocean unaffected by long-range aerosol transport. The high aerosol loads over
Dongsha were mainly linked to anthropogenic fine particle transport from the southeast
coast of China and the occasional advection of desert dust from Mongolian areas. The high
fine aerosol loading in the Taiping area originates from the southwestern region between
Singapore and Indonesia.
The intrusion of aerosol particles from the southeastern coast of China, Singapore and
Indonesia has brought high loading to the aerosols in the SCS. In particular, in the Dongsha
area, 46% of the air masses come from the southeast coast of China, 36% from the China-
Indochina Peninsula and the South China Sea, and 17% from the Luzon Strait and the SCS.
Similarly, the Taiping area was also disturbed by various sources of air masses. Twenty-one
percent of the air masses come from the area enclosed by Singapore and Indonesia, 55%
from the Philippines, and 10% from the Sulu sea and the SCS.
The obvious seasonal cycles of aerosol optical depth and Ångström exponent over
the SCS are caused by the seasonal distribution of meteorological regime over the SCS and
Remote Sens. 2023, 15, 501 15 of 16

the mechanism of aerosol sources. In the Dongsha area, particles originating from spring
show a maximum in aerosol optical depth, while in the Taiping area, aerosol particles from
autumn show a higher optical depth. The seasonal increase of aerosol optical depth was
related to the seasonal ocean current and wind direction in the SCS. In addition, the high
wave height in winter in the two places produces the most sea salt aerosols, resulting in
the smallest wavelength index and the lowest proportion of fine particles in Dongsha and
Taiping. Aerosols in the SCS were greatly affected by the aerosols in the surrounding urban
areas, and the variation was very consistent with surrounding urban areas, but the aerosol
loading was far weaker than that in the coastal city.

Author Contributions: J.C.: Methodology, Formal analysis, Validation, Writing. W.Z.: Conceptu-
alization, Resources, Supervision, Formal analysis, Funding acquisition. Q.L.: Conceptualization,
Resources, Supervision, Formal analysis, Funding acquisition. X.Q.: Formal analysis, Writing—
review & editing. X.C.: Funding acquisition, resources. J.Z.: Validation, Writing—review & editing.
T.Y. (Tao Yang): Validation, Software. Q.X.: Validation, Software. T.Y. (Tengfei Yang): Validation,
Software. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by National Natural Science Foundation of China (Grant No.
41805014), Foundation of Advanced Laser Technology Laboratory of Anhui Province (Grant No.
20191002), Key Program in the Youth Talent Support Plan in Universities of Anhui Province (Grant
No. gxyqZD2020032), Strategic Priority Research Program of Chinese Academy of Sciences (Grant
No. XDA17010104), Key Laboratory of Science and Technology Innovation of Chinese Academy of
Sciences (Grant No. CXJJ-21S028), Open Research Fund of Key Laboratory of Atmospheric Optics,
Chinese Academy of Sciences (Grant No. JJ-19-01) and the President’s Fund of Hefei Institutes of
Physical Science (Grant No. YZJJ2021QN08).
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Conflicts of Interest: The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported in this paper.

References
1. Aerosol Optical Depth. Available online: https://aeronet.gsfc.nasa.gov/ (accessed on 27 May 2022).
2. Atwood, S.A.; Reid, J.S.; Kreidenweis, S.M.; Cliff, S.S.; Zhao, Y.; Lin, N.-H.; Tsay, S.-C.; Chu, Y.-C.; Westphal, D.L. Size resolved
measurements of springtime aerosol particles over the northern South China Sea. Atmos. Environ. 2013, 78, 134–143. [CrossRef]
3. Benedetti, A.; Morcrette, J.-J.; Boucher, O.; Dethof, A.; Engelen, R.; Fisher, M.; Flentje, H.; Huneeus, N.; Jones, L.; Kaiser, J.; et al.
Aerosol analysis and forecast in the European Centre for Medium-Range Weather Forecasts Integrated Forecast System: 2. Data
assimilation. J. Geophys. Res. 2009, 114, D13205. [CrossRef]
4. Chuang, M.-T.; Chang, S.-C.; Lin, N.-H.; Wang, J.-L.; Sheu, G.-R.; Chang, Y.-J.; Lee, C.-T. Aerosol chemical properties and related
pollutants measured in Dongsha Island in the northern South China Sea during 7-SEAS/Dongsha Experiment. Atmos. Environ.
2013, 78, 82–92. [CrossRef]
5. Dubovik, O.; Holben, B.; Eck, T.F.; Smirnov, A.; Kaufman, Y.J.; King, M.D.; Tanré, D.; Slutsker, I. Variability of absorption and
optical properties of key aerosol types observed in worldwide locations. J. Atmos. Sci. 2002, 59, 590–608. [CrossRef]
6. EAC4. Available online: https://ads.atmosphere.copernicus.eu/cdsapp#!/dataset/cams-global-reanalysis-eac4?tab=form (ac-
cessed on 28 May 2022).
7. ERA5. Available online: https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-single-levels?tab=form (accessed
on 1 May 2022).
8. Formenti, P.; Andreae, M.; Andreae, T.W.; Galani, E.; Vasaras, A.; Zerefos, C.; Amiridis, V.; Orlovsky, L.; Karnieli, A.; Wendisch,
M.; et al. Aerosol optical properties and large-scale transport of air masses: Observations at a coastal and a semiarid site in the
eastern Mediterranean during summer 1998. J. Geophys. Res. Atmos. 2001, 106, 9807–9826. [CrossRef]
9. Heintzenberg, J.; Covert, D.C.; Van Dingenen, R. Size distribution and chemical composition of marine aerosols: A compilation
and review. Tellus B Chem. Phys. Meteorol. 2016, 52, 1104–1122. [CrossRef]
10. Holben, B.N.; Eck, T.F.; Slutsker, I.; Tanré, D.; Buis, J.P.; Setzer, A.; Vermote, E.; Reagan, J.A.; Kaufman, Y.J.; Nakajima, T.; et al.
AERONET—A Federated Instrument Network and Data Archive for Aerosol Characterization. Remote Sens. Environ. 1998, 66,
1–16. [CrossRef]
11. Huang, S.-J.; Lin, C.-C. Distribution of Atmospheric Aerosol over the South China Sea. Adv. Meteorol. 2015, 2015, 692762.
[CrossRef]
12. Hyplist. Available online: https://www.arl.noaa.gov/hysplit/ (accessed on 28 May 2022).
Remote Sens. 2023, 15, 501 16 of 16

13. Itahashi, S.; Sakurai, T.; Shimadera, H.; Araki, S.; Hayami, H. Long-term trends of satellite-based fine-mode aerosol optical depth
over the Seto Inland Sea, Japan, over two decades (2001–2020). Environ. Res. Lett. 2021, 16, 064062. [CrossRef]
14. Kim, T.-W.; Lee, K.; Duce, R.; Liss, P. Impact of atmospheric nitrogen deposition on phytoplankton productivity in the South
China Sea. Geophys. Res. Lett. 2014, 41, 3156–3162. [CrossRef]
15. Kubilay, N.; Koçak, M.; Çokacar, T.; Oguz, T.; Kouvarakis, G.; Mihalopoulos, N. Influence of Black Sea and local biogenic activity
on the seasonal variation of aerosol sulfur species in the eastern Mediterranean atmosphere. Glob. Biogeochem. Cycles 2002, 16,
27-1–27-15. [CrossRef]
16. Liu, K.-K.; Chao, S.-Y.; Shaw, P.-T.; Gong, G.-C.; Chen, C.-C.; Tang, T. Monsoon-forced chlorophyll distribution and primary
production in the South China Sea: Observations and a numerical study. Deep. Sea Res. Part I Oceanogr. Res. Pap. 2002, 49, 5.
[CrossRef]
17. Lyamani, H.; Valenzuela, A.; Perez-Ramirez, D.; Toledano, C.; Granados-Muñoz, M.J.; Olmo, F.J.; Alados-Arboledas, L. Aerosol
properties over the western Mediterranean basin: Temporal and spatial variability. Atmos. Chem. Phys. 2015, 15, 2473–2486.
[CrossRef]
18. Mangold, A.; De Backer, H.; De Paepe, B.; Dewitte, S.; Chiapello, I.; Derimian, Y.; Kacenelenbogen, M.; Léon, J.-F.; Huneeus, N.;
Schulz, M.; et al. Aerosol analysis and forecast in the European Centre for Medium-Range Weather Forecasts Integrated Forecast
System: 3. Evaluation by means of case studies. J. Geophys. Res. 2011, 116, D03302. [CrossRef]
19. Morcrette, J.-J.; Beljaars, A.; Benedetti, A.; Jones, L.; Boucher, O. Sea-salt and dust aerosols in the ECMWF IFS model. Geophys. Res.
Lett. 2008, 35, L24813. [CrossRef]
20. O’Neill, N.T. Spectral discrimination of coarse and fine mode optical depth. J. Geophys. Res. Atmos. 2003, 108, 4559. [CrossRef]
21. Reid, J.S.; Hyer, E.J.; Johnson, R.S.; Holben, B.N.; Yokelson, R.J.; Zhang, J.; Campbell, J.R.; Christopher, S.A.; Di Girolamo, L.;
Giglio, L.; et al. Observing and understanding the Southeast Asian aerosol system by remote sensing: An initial review and
analysis for the Seven Southeast Asian Studies (7SEAS) program. Atmos. Res. 2013, 122, 403–468. [CrossRef]
22. Reid, J.S.; Lagrosas, N.D.; Jonsson, H.H.; Reid, E.A.; Atwood, S.A.; Boyd, T.J.; Ghate, V.P.; Lynch, P.; Posselt, D.J.; Simpas, J.B.;
et al. Aerosol meteorology and Philippine receptor observations of Maritime Continent aerosol emissions for the 2012 7SEAS
southwest monsoon intensive study. Atmos. Chem. Phys. 2016, 16, 22. [CrossRef]
23. Smirnov, A.; Holben, B.N.; Kaufman, Y.J.; Dubovik, O.; Eck, T.F.; Slutsker, I.; Pietras, C.; Halthore, R.N. Optical Properties of
Atmospheric Aerosol in Maritime Environments. J. Atmos. Sci. 2002, 59, 501. [CrossRef]
24. Smirnov, A.; Holben, B.N.; Dubovik, O.; Frouin, R.; Eck, T.F.; Slutsker, I. Maritime component in aerosol optical model derived
from Aerosol Robotic Network data. J. Geophys. Res. 2003, 108, 4033. [CrossRef]
25. Tseng, C.-M. A unique seasonal pattern in phytoplankton biomass in low-latitude waters in the South China Sea. Geophys. Res.
Lett. 2005, 32, L08608. [CrossRef]
26. Wang, S.-H.; Tsay, S.-C.; Lin, N.-H.; Chang, S.-C.; Li, C.; Welton, E.J.; Holben, B.N.; Hsu, N.C.; Lau, W.K.; Lolli, S.; et al. Origin,
transport, and vertical distribution of atmospheric pollutants over the northern South China Sea during the 7-SEAS/Dongsha
Experiment. Atmos. Environ. 2013, 78, 124–133. [CrossRef]
27. Xiao, H.-W.; Xiao, H.-Y.; Luo, L.; Shen, C.-Y.; Long, A.-M.; Chen, L.; Long, Z.-H.; Li, D.-H. Atmospheric aerosol compositions over
the South China Sea: Temporal variability and source apportionment. Atmos. Chem. Phys. 2017, 17, 3199–3214. [CrossRef]
28. Xiao, H.-W.; Xie, L.-H.; Long, A.-H.; Ye, F.; Pan, Y.-P.; Li, D.-N.; Long, Z.-H.; Chen, L.; Xiao, H.-Y.; Liu, C.-Q. Use of isotopic
compositions of nitrate in TSP to identify sources and chemistry in South China Sea. Atmos. Environ. 2015, 109, 70–78. [CrossRef]
29. Zhang, C.; Xu, H.; Li, Z.; Xie, Y.; Li, D. Maritime Aerosol Optical and Microphysical Properties in the South China Sea Under
Multi-source Influence. Sci. Rep. 2019, 9, 17796. [CrossRef] [PubMed]
30. Zhang, X.; Zhuang, G.; Guo, J.; Yin, K.; Zhang, P. Characterization of aerosol over the Northern South China Sea during two
cruises in 2003. Atmos. Environ. 2007, 41, 7821–7836. [CrossRef]
31. Zheng, X.-S.; Zhang, Y.-N.; Pan, Y. Spatial and Temporal Variations of Aerosol Optical Depth and Influence Factors over the East
China Sea. Int. J. Environ. Monit. Anal. 2016, 4, 94. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.