Global and Regional Variations and Main Drivers of Aerosol Loadings over Land during 1980–2018

Abstract

Aerosol particles originated from anthropogenic emissions, volcanic eruptions, biomass burning, and fossil combustion emissions, and their radiative effect is one of the most uncertain factors in climate change. Meanwhile, aerosol particles in fine particle size could also cause irreversible effects on the human respiratory system. This study attempted to analyse the spatial and temporal variations of global aerosol optical depth (AOD, 550 nm) during 1980–2018 using MERRA-2 aerosol reanalysis products and to investigate the effects of natural/anthropogenic emissions of different types of aerosols on AOD values. The results show that the global annual mean AOD values kept high levels with significant fluctuations during 1980–1995 and showed a consistent decreasing and less volatile trend after 1995. Spatially, the AOD values are relatively higher in the Northern Hemisphere than in the Southern Hemisphere, especially in North Africa (0.329), Northern India (0.235), and Eastern China (0.347), because of the intensive natural/anthropogenic aerosol emissions there. The sulphate-based aerosols emitted by biomass burning and anthropogenic emissions are the main types of aerosols worldwide, especially in densely populated and industrialized regions such as East Asia and Europe. Dust aerosols are also the main aerosol type in desert areas. For example, the AOD and AODP values for the Sahara Desert are 0.3178 and 75.32%, respectively. Both black carbon aerosols (BC) and organic carbon aerosols (OC) are primary or secondary from carbon emissions of fossil fuels, biomass burning, and open burning. Thus, the regions with high BC and OC aerosol loadings are mainly located in densely populated or vegetated areas such as East Asia, South Asia, and Central Africa. Sea salt aerosols are mainly found in coastline areas along the warm current pathway. This study could help relevant researchers in the fields of atmospheric science, environmental protection, air pollution, and ecological environment to understand the global spatial–temporal variations and main driving factors of aerosol loadings.

1. Introduction

Atmospheric aerosol particles are stable mixed systems of solid and liquid particles uniformly dispersed in the atmosphere [1,2]. They can act as atmospheric condensation nuclei for water droplets and ice crystals and participate in various chemical cycles [3,4]. Fog, smoke, haze, fine dust, and smog are all atmospheric aerosols from natural or anthropogenic emissions [5]. Aerosol particles can absorb and scatter solar radiation, influencing the energy reaching the ground and reducing the temperature of the lower atmosphere [6]. Meanwhile, fine aerosol particles can be easily accompanied by toxic and harmful substances (heavy metals, microorganisms, etc.) and have a long residence time and long transport distance in the atmosphere, thus having a great impact on human health and air quality [7,8]. Thus, in recent decades, the impact of aerosol particulate matter, especially fine aerosol particulate matter, on the natural environment and human health have become a major environmental concern around the world. Obtaining accurate aerosol optical properties data is the prerequisite for studying the radiative forcing effects of aerosols.

Numerous ground-based observation networks, such as the Aerosol Robotic NETwork (AERONET. Available online: http://aeronet.gsfc.nasa.gov (accessed on 10 October 2021)) [9], AERosol CANada (Aerocan. Available online: http://www.aerocanonline.com/ (accessed on 10 October 2021)) [10], the Sky Radiometer Network (Skynet. Available online: https://www.skynet-isdc.org/aboutSKYNET.php (accessed on 10 October 2021)) [11], and the China Aerosol Remote Sensing Network (CARSNET. Available online: http://data.cma.cn/ (accessed on 10 October 2021)) [12], were established, providing long-term ground-based observations of aerosol optical properties. However, due to the construction cost, the spatial distribution of ground-based observation sites is extremely sparse, which is difficult to be applied to high-resolution aerosol optical properties analysis on a global scale. Satellite remote sensing can provide large-scale and real-time continuous aerosol data with high spatial resolutions [13,14,15]. There are various aerosol remote sensing networks covering the world, such as the MODIS (Moderate Resolution Imaging Spectroradiometer) on the Terra and Aqua satellites [16,17], TOMS (Total Ozone Mapping Spectrometer (TOMS) on the Nimbus satellite [18], and OMI sensor on the Aura satellite [19]. Many studies were previously conducted on the spatial and temporal distributions and the main driving factors of global aerosol optical properties using satellite remote sensing data. For example, Hsu et al. analyzed the global temporal trends of aerosol optical depth (AOD) over land and ocean using SeaWiFS measurements and found an upward trend of 0.00078 ± 0.00019 yr⁻¹ for global average AOD values from 1997 to 2010 [20]. Sudden natural events, especially volcanic eruptions, could cause dramatic fluctuations of AOD values on a global scale [21]. For example, the 1992 eruption of Mount Pinatubo caused a global spike in aerosol loadings, especially in East Asia [22,23]. Meanwhile, anthropogenic emissions are also non-negligible factors for the changes in global or regional AOD values. Sanap et al. found a marked decline in global aerosol loadings during the COVID-19 period, which was inevitably linked to the reduction of human activities during this period [24]. The sharp, downward trend of AOD values during the COVID-19 period were mainly found in India [25], Eastern China [26], Europe [27], and other densely populated areas [26]. In terms of the spatial pattern of global aerosol loadings, it is found to be generally higher in the Northern Hemisphere than for the Southern Hemisphere because it is mostly dominated by the oceans [28]. North Africa is the area with the highest aerosol loadings, owing to the continuous year-round dust particles entrained into the atmosphere [29]. Areas with relatively lower aerosol loadings are mainly distributed in high latitudes where natural and anthropogenic aerosol emissions are relatively low [13]. In addition, there are also large spatial heterogeneities in AOD values in regions that are spatially close areas with the same natural environment. For example, in recent decades, AOD values in western Africa showed a decreasing trend, while those in the Arabian Peninsula showed an increasing trend [20].

Meanwhile, many studies were conducted on the global and regional variations of the emissions and the optical properties of different types of aerosols (dust, black carbon, organic carbon, sea salt aerosols, etc.) in order to trace the main causes of the spatial–temporal variations in global aerosol loadings. Sea salt aerosols are the main type of aerosol particles in the global coastal zone [30]. Vignati et al. estimated the total global emissions of sea salt aerosols to be about 24 Tg fine yr⁻¹ [31]. Dust aerosols are soil particles suspended in the atmosphere in deserts or areas with erodible dry soils and strong wind [32,33]. It is estimated that the Sahara Desert can transport more than 200 and 260 million tons of dust aerosol particles into the atmosphere each year [34], and this could reach North America [35] and the Amazon forest [36] by atmospheric circulation across the Atlantic Ocean. Black carbon aerosols are mainly the products of the incomplete combustion of fossil fuels and biofuels, such as coal, wood, and diesel [37]. Thus, areas with high black carbon aerosol emissions are mainly found in densely populated areas [38]. The annual mean black carbon surface concentrations are 1000–14,000, 500–5000, 100–500, 10–100, and <10 ngm-3 for Asia, Europe, the United States, high northern latitudes, and remote regions, respectively [39]. Sulfur aerosols are secondary inorganic aerosols formed by the oxidation of sulfate in the atmosphere from sulfur dioxide emissions of anthropogenic sources (human activities) and volcano eruptions (natural emissions) [40]. Sulfur aerosols emissions peaked in the early 1970s, decreased until 2000, and increased in recent years because of the rapid increase in anthropogenic emissions in Eastern Asia [41]. Organic carbon (OC) aerosols are the fraction of carbonaceous aerosols which contain compounds of carbon [42]. The biomass burning areas such as Central Africa [43], South America [44], and Indonesia [45] are considered to be the regions with the highest global organic carbon aerosol emissions.

As can be seen from the above studies, satellite products have made important contributions to the analysis of spatial–temporal variations in aerosol optical properties. However, the spatial–temporal continuity of satellite remote sensing products is easily affected by clouds, surface reflection, and other atmospheric components. Qin et al. counted the number of AOD records derived from MODIS Level 3 Atmosphere Gridded Product over mainland China in 2015 [21]. The results showed that no MODIS AOD records could exceed 330 days for any image element over mainland China. Qin et al. further investigated the AOD records derived from MCD19A2 aerosol products during 2001–2020 with a spatial resolution of 1 km × 1 km [28]. The result indicated that the spatial continuity of MAIAC AOD records is generally poor in high latitudes in the northern hemisphere because the MAIAC algorithm could not retrieve aerosol properties over snow surfaces; the percentage of effective AOD records is higher in summer than in other seasons. Compared with satellite AOD retrievals, the reanalysis products can provide aerosol data with greater spatial–temporal continuity, higher temporal resolution, and acceptable accuracy.

This study attempted to investigate the spatial and temporal variations and temporal trends of global AOD (550 nm) values during 1980–2018 using MERRA-2 aerosol reanalysis products. Meanwhile, the contributions of different types of aerosols on the total AOD values are analyzed. Moreover, the spatial correlation and spatial aggregation patterns for global AOD values and natural/anthropogenic aerosol emissions with different bin sizes are also studied. Below, Section 2 introduces the methodology and data used in this study. Section 3 illustrates the main results of this study.

2. Data and Methods

2.1. MERRA-2 Aerosol Reanalysis Products

MERRA-2 (The Modern-Era Retrospective analysis for Research and Applications, version 2, https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/, accessed on 10 October 2021) aerosol reanalysis products could provide a long-term publicly available gridded dataset of global aerosol optical properties and aerosol diagnostics at hourly, 3-hourly, and monthly time scales [46]. This product is generated using the GOCART aerosol module coupled to the GEOS-5 Data Assimilation System and the Goddard Aerosol Assimilation System. Meanwhile, the AOD values derived from MODIS, MISR AOD, and AERONET AOD observations are also assimilated in the MERRA-2 aerosol reanalysis products. In this study, the aerosol optical depth (550 nm), the AOD for different types of aerosols, the natural aerosol emissions, and the anthropogenic aerosol emissions data during 1980–2018 from MERRA-2 aerosol reanalysis products are used to investigate the global spatial–temporal variation and the main drivers of AOD values. Detailed information about the MERRA-2 aerosol reanalysis products used in this study is shown in Table 1.

In this study, the monthly mean aerosol optical depth data for black carbon aerosols (BCAOD), dust aerosols (DUAOD), organic carbon aerosols (OCAOD), sea salt aerosols (SSAOD), and SO₄ aerosols (SO4AOD) derived from MERRA-2 datasets were used to investigate the main aerosol resources of aerosol particles over the world. The total AOD values are nearly the sum of the BCAOD, DUAOD, OCAOD, SSAOD, and SO₄AOD values with a very small residual. Thus, the contribution of black carbon aerosols, dust, organic carbon aerosols, sea salt aerosols, and SO₄ aerosols on the total AOD values could be expressed as follows:

(1)$AODP=xAOD/AO{D}_{Total}$

(2)$BCAOD+DUAOD+OCAOD+SSAOD+SO4AOD+c=AO{D}_{Total}$

where x could be BC, DU, OC, SS, and SO₄; AODP (%) denotes the contribution of different types of aerosols to $AO{D}_{Total}$, c is a negligible constant (0.000012).

Moreover, the aerosol emissions are divided into natural aerosol emissions (BCEBB, OCEMBB, OCEMBF, OCEMBG, SO2EMBB, SO2EMVE, SO2EMVN) and anthropogenic aerosol emissions (BCEMAN, SO2EMAN, SO4MAN) with different bin sizes (BCEMxxx, DUEMxxx, OCEMxxx, SSEMxxx, SUEMxxx). Detailed information about the MERRA-2 aerosol reanalysis products used in this study is shown in Table 2 and Table 3.

Both black carbon aerosols and organic carbon aerosols are represented with two bins that correspond to hydrophobic and (aged) hydrophilic particles. The dry size and density for the two OC tracers are: 0.35 µm and 1800 kg/m³. Sulfate aerosol Tracers 1, 2, and 4 are gas species and do not have associated sizes with them. The SO4 tracer (#3) corresponds to sulfate particles and has a dry size and density of 0.35 µm and 1700 kg/m³.

2.2. The Empirical Orthogonal Function Analysis

To further identify the coherent spatial and temporal variability of global AOD values, empirical orthogonal function analysis (EOF) is performed on the yearly mean AOD data during 1980–2018. The EOF analysis could decompose multivariate data into independent orthogonal eigenvectors [47]. The EOF analysis makes it convenient to transform spatial–temporal data sets into spatial modalities of physical quantities and the projections (time series) associated with them in time. The specific steps of EOF analysis are as follows:

Firstly, for a spatial–temporal dataset, it could be tabulated as an $m\times n$ matrix A (Equation (3)). Then, we can obtain the anomaly matrix A’ (Equation (4)) of matrix A, which is obtained by subtracting the mean values of matrix A in each row.

(3)$A=\left[\begin{array}{cccc}{x}_{1,1}& x_{1,2}& \dots & x_{1,n}\\ x_{2,1}& x_{2,2}& \dots & x_{2,n}\\ ⋮& ⋮& \ddots & ⋮\\ x_{m,1}& x_{m,1}& \dots & x_{m,n}\end{array}\right]$

(4)$A^{\prime }=\left[\begin{array}{cccc}{a}_{1,1}& a_{1,2}& \dots & a_{1,n}\\ a_{2,1}& a_{2,2}& \dots & a_{2,n}\\ ⋮& ⋮& \ddots & ⋮\\ a_{m,1}& a_{m,1}& \dots & a_{m,n}\end{array}\right]$

(5)$a_{i,j}=x_{i,j}-\widetilde{x_i}$

(6)$C_{m\times m}=\frac{1}{n}A\times A^{\prime }$

where, $x_{i,j}$ is the AOD values in a specific time and place; m (576 $\times$ 361) and n (39) are the spatial and temporal scales of the AOD datasets; $\widetilde{x_i}$ is the mean value of the ith row of matrix A; $C_{m\times m}$ is the covariance matrix. The eigenvalue C (λ1, ..., m) and the eigenvector $V_{m\times m}$ satisfies the following equation:

(7)$C_{m\times m}\times V_{m\times m}=V_{m\times m}\times {\wedge }_{m\times m}$

where, ${\wedge }_{m\times m}$ is the matrix of eigenvalues C. Each non-zero eigenvalue corresponds to a column of eigenvector values, also known as EOF. Finally, the matrix of time coefficients (PCs) can be calculated from the projection of A onto the EOFs using the following Equations (8) and (9):

(8)${\wedge }_{m\times m}=\left[\begin{array}{cccc}{\lambda }_1& 0& \dots & 0\\ 0& {\lambda }_2& \dots & 0\\ \dots & \dots & \dots & \dots \\ 0& 0& \dots & {\lambda }_m\end{array}\right]$

(9)$P{C}_{m\times n}=V_{m\times m}^T\times X_{m\times n}$

where, each row of data in PC is the time coefficient corresponding to each feature vector.

2.3. Temporal Trend Analysis

In this study, the Sen’s Slope index and Mann–Kendall trend test are introduced to investigate the temporal trend of global AOD values during 1980–2018. The Sen’s Slope index is basically used to identify the magnitude of increasing/decreasing trend of a time series [48], which was expressed as the median of the slopes of each pair of data ($Q_{med}$) as follows:

(10)$Q_i=\frac{x_j-x_k}{j-k} for i=1,\dots ,N$

(11)$Q_{med}=\left\{\begin{array}{cc}{Q}_{\left[\left(i+1\right)/2\right]}& if i is an odd numer\\ \frac{Q_{\left[i/2\right]}+Q_{\left[\left(i+2\right)/2\right]}}{2}& if i is an even numer \end{array}\right.$

where $x_j$ and $x_k$ are the AOD values at the jth and kth time (j > k), i is the index of the ith data in the time series.

The Mann–Kendall trend test is a non-parametric test method recommended by the World Meteorological Organization to analyze the significance of increasing or decreasing temporal trends for a time series [21]. The Mann–Kendall trend test could be conducted using following steps. Firstly, a test statistical indicator S and its variance are calculated as follows:

(12)$S={\displaystyle \sum }_{i=2}^n{\displaystyle \sum }_{j=1}^{i-1}\mathrm{sign}\left(x_j-x_i\right)$

(13)$\mathrm{VAR}\left(S\right)=\left[n\left(n-1\right)\left(2n+5\right)-{\displaystyle \sum }_{k=1}^p{q}_k\left(q_k-1\right)\left(2q_k+5\right)\right]/18$

where p is the tie group of x; $q_k$ represent the sample number of the kth tie group; n is the total sample number; $\mathrm{sign}\left(x_j-x_i\right)$ is a symbolic function, which is equal to −1, 0, or 1 when $x_j-x_i$ is less than zero, greater than zero, or equal to zero, respectively. Finally, the $MK$ values representing the significance of increasing or decreasing trends are calculated using the following equations.

(14)$MK=\left\{\begin{array}{cc}\left(S+1\right)/\sqrt{\mathrm{VAR}\left(S\right)}& S>0\\ 0& S=0\\ \left(S-1\right)/\sqrt{\mathrm{VAR}\left(S\right)}& S<0\end{array}\right.$

when the absolute value of $MK$ is greater than or equal to 1.96 and 2.58, the increasing/decreasing trend of AOD values is significant at the levels of 95% and 99%, respectively.

2.4. Spatial Correlation and Aggregation Analysis Method

In this study, the Local Bivariate Moran’s I is introduced to quantify the spatial correlations and aggregation between natural/anthropogenic aerosol emissions and aerosol optical depth. In essence, it captures the relationship between the value for one variable at location i, x_i, and the average of the neighboring values for another variable [49]. Apart from a constant scaling factor (that can be ignored), the statistic is the product of x_i with the spatial lag of y_i.

(15)$I=\frac{N}{{{\displaystyle \sum }}_i{{\displaystyle \sum }}_j{w}_{ij}}\frac{{{\displaystyle \sum }}_i{{\displaystyle \sum }}_j{w}_{ij}\left(X_i-\overline{X}\right)\left(Y_j-\overline{Y}\right)}{{{\displaystyle \sum }}_i{\left(Y_i-\overline{Y}\right)}^2}$

where $w_{ij}$ is the spatial weight matrix for the input data matrix. All the above steps for calculating the Local Bivariate Moran’s I index could be implemented on free and open-source software GeoDa. A detailed description of the Local Bivariate Moran’s I could be found in ref [49].

In this study, the spatial correlation between AOD values and aerosol emissions were conducted on GeoDa software. The Z-value test was performed according to the LISA value. The results of spatial autocorrelation were divided into four spatial distribution patterns based on the Z-values: high–high correlation (HH), high–low correlation (HL), low–high correlation (LH), and low–low correlation (LL).

3. Result and Discussion

3.1. Global Spatial and Temporal Variation of AOD

Figure 1 illustrate the annual global variations of AOD values from 1980 to 2018. The natural volcano eruptions show considerable effects on the global variation and mutation of aerosol loadings. The ALCH Joan volcanic eruption (1982) and Pinatubo volcanic eruption (1991) directly led to two significant global upward trends of AOD values during 1982–1984 and 1991–1994, respectively [50]. The global mean AOD values rose rapidly to 0.2423 in 1992, which caused a dramatically decrease of surface downward solar irradiances, especially in eastern Asia. After 1992, the AOD values dropped to relatively low points with slight fluctuations, with the implementation of emission control policies in most countries of the world.

The spatial distributions and monthly variations of the global AOD values are shown in Figure 2, Figure 3 and Figure 4. The areas with high aerosol loadings are identified and marked with red boxes. North Africa is the area with the highest AOD value (0.329) owing to the continuous emission of dust aerosols into the atmosphere from the Saharan region of Africa throughout the year there. The AOD values in North Africa are relatively higher in spring and summer than in other seasons because of the frequent sandstorms in spring and summer. Northern India (0.235) and Eastern China (0.347) are also areas with high AOD values, which are highly dominated by anthropogenic emissions. In Northern India, the AOD values are higher in summer than in other seasons because of the frequent sand and dust weather caused by the strong heating atmosphere there. In Eastern China, the AOD values are higher in spring and summer than in other seasons due to the intensive anthropogenic emissions caused by straw burning in spring and the sticky weather in summer. Areas covered by dense forests such as Central Africa (0.219) and Central South America (0.133) are also high, which may be due to the non-negligible emissions of organic carbon. The AOD values in areas with highly developed industries, such as the Western United States and most of Europe, are relatively lower than those in Northern India and Eastern China due to the control of anthropogenic aerosol emissions in recent decades. The lowest AOD values are mainly found in polar areas with scarcely natural and anthropogenic aerosol emissions.

Figure 5 depict the result of the first three leading modes of the empirical orthogonal function analysis. Variance explained by Mode1, Mode2, and Mode3 are 55.70%, 18.71%, and 9.40%, which reveal three distinct spatial patterns, respectively. The spatial pattern of Mode1 indicate that the maximum aerosol loading is in Northern India and Eastern China with serious anthropogenic aerosol emissions. Thus, Mode1 is clearly dominated by anthropogenic aerosol emissions owing to intensive human activities. The spatial pattern of Mode2 is restricted in most areas of Europe. As the key areas of every industrial revolution in history, although Europe has taken measures to save energy and reduce emissions in recent years, the AOD values are still higher than most developing countries due to relatively high anthropogenic emissions. Mode 3 indicate high aerosol loadings in Indonesia because of the episodic and extensive biomass burning in recent decades there. The right part of Figure 5 show the temporal variation of PCs for aerosol optical depth corresponding to Mode1, Mode2, and Mode3. Combined with the results of Figure 1, it can be concluded that the ALCH Joan volcanic eruption (1982) and Pinatubo volcanic eruption (1991) had great interference on the temporal variations of the principal components of Mode1, Mode2, and Mode3. The anthropogenic emissions were not the dominant factors for AOD changes during 1982–1984 and 1991–1993. Except for the years before and after these two volcanic eruptions, the PC values of Mode1 were always positive, which indicated that the aerosol loadings in Northern India and Eastern China remained high with slight fluctuations from 1980–2018. The PC values for Mode2 show a downward trend from 1980 to 2018, which reflects that the air pollutant emission-reducing policies in recent decades in Europe achieved certain effects on controlling the aerosol loadings in the atmosphere in Europe. The PC values for Mode3 were dominated by biomass burning before 1991, then the values sharply decreased in 1992 owing to the Pinatubo volcanic eruption (1991) and then fluctuated with slight variation after 1995.

The spatial distributions of the MK and Sen slope values of the global annual mean AOD values during 1980–2018 are illustrated in Figure 6 and Figure 7. Great spatial heterogeneity of the temporal trend of AOD values were found all over the world. A significantly increasing trend of AOD values was noted over Eastern China and India due to the rising anthropogenic emissions with growing populations and rapidly developing industries in recent decades. A growing trend of AOD was also found in the Amazon Rainforest, which may be due to the excessive logging and biomass burning there. Reductions in AOD values were identified in Europe and Eastern America, owing to the anthropogenic emission reduction policies implemented in recent years. An interesting result found that the AOD values show inconspicuous temporal variations in areas less affected by human activities, even in North Africa and the Taklimakan Desert in Northwestern China as areas with high aerosol loadings. Therefore, it is understood from the analysis that human activities may be the main driver for the temporal variations of global annual mean AOD values in the last 40 years.

3.2. The Main Drivers of the Global Distributions of AOD

The spatial pattern of the global AOD and AODP values for black carbon aerosols, dust aerosols, organic caron aerosols, sea salt aerosols, and SO4 aerosols during 1980–2018 is plotted in Figure 8. Meanwhile, the spatial distribution of monthly mean AODP values for different types of aerosols is also plotted in Figures S1–S6 in the Supplementary Materials. It should be stated that the spring, summer, autumn, and winter for the northern hemisphere are March to May, June to August, September to November, and December to February of the next year, respectively; while for the southern hemisphere, they are September to November, December to February of the next year, March to May, and June to August.

The black carbon aerosol is the dominant form of absorbing aerosol in the atmosphere emitted by incomplete combustion processes from human activities and natural wildfire. However, the AODP for black carbon aerosols (BCAODP) is relatively lower than that of dust aerosols, organic carbon aerosols, and SO₄ aerosols because it can only remain in the atmosphere in 1–2 weeks [51]. The ranges of the annual mean BCAOD and BCAODP are 0.0009–0.0611 and 0.79%–13.68%, respectively. Eastern China is the area with high annual mean AOD (0.0228) and AODP (6.27%) values for black carbon aerosols because of the relatively high black carbon aerosol emissions caused by fossil fuel combustion. Central Africa and South America are also areas with high black carbon aerosol emissions caused by frequent wildfire events. In terms of the monthly variations, BCAODP is generally higher in summer than in other seasons for South America and Central Africa, owing to the frequent wildfire events there. In the northern hemisphere, BCAODP is higher in summer and winter because of the intensive human activities in summer and coal combustion for winter heating.

Dust aerosol is a complex mixture of dust suspended in the atmosphere and is usually considered of natural origin or human activities through desertification and land use/land cover change [32]. The ranges of the annual mean DUAOD and DUAODP are 0.0005–0.5761 and 0.39–88.45%, respectively. The areas with high DUAOD and DUAODP values are generally more widely distributed in summer than that in other seasons owing to the extremely heating atmosphere and strong trade winds in desert zones in summer. The areas with high DUAOD and DUADOP values are highly concentrated in deserts, such as the Gobi Desert, Sahara, Karakum, and Taklimakan desert. Among them, the Sahara is the major source on Earth of dust aerosols, spanning 9 million square kilometers from the Atlantic Ocean to the Red Sea. The annual mean AOD and AODP values for the Sahara Desert are 0.3178 and 75.32%, respectively.

Organic carbon aerosols are the fraction of carbonaceous aerosols containing carbon, which may be primary or secondary from carbon emissions of fossil fuels, biofuels, and open burning. The ranges of the global annual mean OCAOD and OCAODP are 0.0032–0.3523 and 2.05–68.63%, respectively. The areas with high OCAOD and OCAODP values are mainly located in areas with dense tropical forest/grassland and year-round high temperatures such as Central Africa, South America, and Indonesia. For example, the annual mean OCAOD and OCAODP values for Central Africa are 0.1508 and 46.00%, respectively; the annual mean OCAOD and OCAODP values for South America are 0.0579 and 41.13%, respectively. Central Africa has a savannah climate with a rainy season from April to May and a dry season in other months. Thus, the OCAOD and OCAODP values are generally lower in April and May than in other seasons because of the more frequent biomass burning in the dry season. Northern North America and Siberia are also high-value areas for OCAOD and OCAODP, but with large seasonal variation.

Sea salt aerosol originally comes from sea spray, which is only distributed in the atmosphere above the sea or coastal areas. The ranges of the annual mean SSAOD and SSAODP over land are 0.0006–0.1265 and 0.29–68.86%, respectively. Meanwhile, large spatial variability exists in SSAOD and SSAODP along the global coastline. High SSAOD and SSAODP values are mainly located along the coastline where warm currents flow through because of the relatively stronger atmospheric convections and sea–air interactions. For example, the sea salt aerosol is the main composition of aerosol in the Caribbean Periphery owing to a steady stream of sea salt aerosol particulate matter emissions affected by the North Equatorial Warm Current and the Gulf of Mexico Warm Current. The South Pacific Island States are also areas with high sea salt aerosol loadings affected by the East Australian Warm Current and South Equatorial Warm Current.

SO₄ aerosol plays an indispensable role in the radiation energy balance between the atmosphere and the Earth’s surface [52]. The ranges of the annual mean SO₄AOD and SO₄AODP are 0.0196–0.4825 and 7.62–78.80%, respectively. Although volcanic eruptions greatly impact global SO₄ aerosol changes, anthropogenic aerosol emission is the main factor of global and regional aerosol changes due to the low frequency of volcanic eruptions. Thus, the areas with high SO4AOD and SO4AODP values are mainly distributed in densely populated areas, such as Eastern China, Eastern America, and most of Europe. For example, the annual mean SO4AOD and SO4AODP values for Eastern China are 0.2077 and 58.32%, respectively. The SO4AODP values are also high in high latitude areas, but the SO4AOD values are not as high as those in the other areas of the world because other types of AOD values are relatively lower than those of SO4AOD in high latitude areas, making the SO4AODP values there appear relatively high. The AODP values in the northern hemisphere are relatively higher in winter than in other seasons, which may be related to the relatively higher sulfur dioxide emissions caused by winter heating in these areas.

3.3. The Effect of Aerosol Emissions on Global AOD Variations

To further investigate the effects of natural and anthropogenic aerosol emissions in different bin sizes on the spatial distributions of global AOD values, we analyzed their spatial correlation and spatial aggregation characteristics. Figure 9 show the global spatial aggregation of AOD and aerosol emissions. The spatial aggregation patterns of aerosol emissions and AOD are characterized by a more pronounced regional. The less populated areas such as Greenland, the Qinghai–Tibet Plateau, and the Antarctic region are characterized by low aerosol emissions and low AOD values with low–low aggregation. Dust aerosol emissions are the main component of atmospheric aerosol particulate matter in North Africa, West Asia, Central Asia, and Northwest China, and thus dust aerosol emissions of different particle sizes and AOD values in these regions are characterized by a significant ‘high–high’ concentration pattern. Both organic carbon and black carbon are mainly sourced from biomass burning emissions, biofuel emissions, and anthropogenic emissions, and thus both have similar spatial aggregation patterns. In densely populated regions or vegetated regions, such as India, Eastern China, Eastern USA, Central Europe, and Central Africa, the organic carbon and black carbon aerosol emissions show an obvious ‘high–high’ clustering with AOD values. However, it could also be observed that organic carbon and black carbon emissions in the central regions of South America and Central Africa are mainly caused by biomass burning.

4. Conclusions

This paper systematically analyses the spatial and temporal distributions of global aerosol optical depth (550 nm) during 1980–2018, using MERRA-2 aerosol reanalysis products. Furthermore, we decompose the spatial–temporal dataset into spatial modes and their associated temporal projections using EOF analysis to deconstruct the intrinsic spatial–temporal variability. At the same time, we explore the main types of aerosols in various regions of the world and analyse the spatial correlation between natural and anthropogenic aerosol emissions and AOD at different particle sizes.

During 1980–1995, the annual mean global AOD values fluctuated considerably due to the ALCH Joan volcanic eruption (1982) and Pinatubo volcanic eruption (1991), which reached a maximum value of 0.2423 in 1992. After 1995, the spatial and temporal variability of global AOD values dropped to a relatively low point with slight fluctuations. Further, the results of the first mode of EOF analysis indicate that the impact of anthropogenic emissions on AOD values before 1995 is much smaller than that of natural aerosol emissions, especially during the two volcanic eruption events. Spatially, North Africa, Northern India, Eastern China, Central Africa, and South America are always the areas with the highest annual mean AOD values. Relatively lower annual mean AOD values are mainly found in polar areas with scarcely natural and anthropogenic aerosol emissions. A significantly increasing trend of AOD values is noted over Eastern China and India, while there are downward trends of AOD values in Europe and Western America because of the anthropogenic emission reduction policies implemented in recent years.

In terms of aerosol particle compositions, dust aerosol is one of the main aerosol types over the world, but its spatial influence is limited to desert areas such as North Africa, West Asia, Central Asia, and Northwestern China. For example, the annual mean DUAOD and DUAODP values for the Sahara Desert are 0.3178 and 75.32%, respectively. The sulphate-based aerosols have a larger spatial influence than other types of aerosols, particularly in densely populated and more industrialized areas such as East Asia, the East Central United States, and Europe. For example, the annual mean SO4AOD and SO4AODP values for Eastern China are 0.2077 and 58.32%, respectively. The SO2EMVE and SO2EMVN show less spatially correlated relationships with AOD values than those of other aerosol particles because volcanic eruptions are less frequent events than anthropogenic emissions and biomass burning. Organic carbon aerosols are the dominant aerosol type in densely vegetated regions such as Central Africa and Central South America, where natural and anthropogenic biomass burning is more frequent. For example, the annual mean OCAOD and OCAODP values for Central Africa are 0.1508 and 46.00%, respectively. Black carbon aerosols are mainly derived from anthropogenic aerosol emissions and biomass burning, so their impact is mainly in densely populated or vegetated areas such as East Asia, South Asia, Central Africa, and Central South America. However, the AODP values for black carbon aerosols are relatively lower than those of aerosol types, owing to the relatively short retention time in the atmosphere. For example, the annual mean BCAOD and BCAODP values for Eastern China are only 0.0228 and 6.27%, respectively. Sea salt aerosols are the dominant aerosol type in coastal areas and are mainly found in areas along the warm current pathway due to the relatively stronger atmospheric convections and Sea–Air interactions there.

Certainly, further studies need to be conducted to explore the global spatial and temporal distribution of aerosol loadings, their chemical composition, and their correlation with particulate matter emissions of different aerosol types using satellite remote sensing data and reanalysis products at finer spatial and temporal resolutions. In addition, there is still a long way to go to investigate the main drivers of regional spatial and temporal variability in aerosol particulate matter in conjunction with numerical weather forecasting models for areas with high aerosol loadings.

Author Contributions

Z.L. and K.D. designed the research. Z.L. and J.S. performed the experiments and analyzed the data. Z.L. and K.D. wrote the manuscript. Z.L. and H.H. revised the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This work was financially supported by the National Natural Science Foundation of China (No. 42174012).

Data Availability Statement

The MERRA-2 (The Modern-Era Retrospective analysis for Research and Applications, version 2, aerosol reanalysis products are openly available in https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/ (accessed on 10 October 2021).

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. The annual variation of AOD values over the world during 1980–2018. The black line with black makers represents the annual mean AOD values, while the histogram represents the anomaly annual mean AOD values.

Figure 2. The global spatial distributions of annual mean AOD values during 1980–2018.

Figure 3. The global spatial and monthly variations of AOD values from January to June.

Figure 4. The global spatial and monthly variations of AOD values from July to December.

Figure 5. First three leading modes of the empirical orthogonal function analysis (EOF).

Figure 6. The MK values for the global variation of AOD values during 1980–2018.

Figure 7. The Sen slope values for the global variation of AOD values during 1980–2018.

Figure 8. The spatial distributions of AOD and AODP values for different types of aerosols. The subfigures on the left indicate the AOD values, while the subfigures on the right indicate the AODP values.

Figure 9. Global spatial aggregation of AOD and aerosol emissions. Sulphur acid-based aerosols are mainly caused by anthropogenic emissions, biomass burning emissions, and volcanic eruption emissions. Thus, SO2EMVE and SO2EMVN do not have high–high aggregation spatial distribution with AOD values. Like OCEMBB and BCEMBB, SO2EMBB is also mainly concentrated in areas with dense vegetation or areas where biomass fuel is the main fuel for daily use. However, unlike OCEMAN and BCEMAN, anthropogenic emissions of sulphur acid-based aerosols are mainly associated with industrial productions, and therefore high SO2EMAN, SO4EMAN, and SUEM003 value areas are mainly located in more densely populated and industrially developed regions such as East Asia, South Asia, Central Europe, and the eastern United States. The areas with high concentrations of marine aerosol emissions and SSAOD of different particle sizes are mainly found in coastal areas.

Table 1

Basic information about the aerosol optical depth that was used in this study.

Short Name	Full Name
AODANA	aerosol optical depth analysis
SO4SO4AOD	so4 aerosol optical depth [550 nm]
SSAOD	sea aerosol optical depth [550 nm]
OCAOD	organic carbon aerosol optical depth [550 nm]
DUAOD	dust aerosol optical depth [550 nm]
BCAOD	black carbon aerosol optical depth [550 nm]

Table 2

Information about aerosol emission parameters.

Aerosol Types	Short Name	Long Name
Black Carbon	BCEM001	black carbon emission bin001
	BCEM002	black carbon emission bin002
	BCEMAN	black carbon anthropogenic missions
	BCEMBB	black carbon biomass burning emissions
Dust Aerosols	DUEM001	dust emission bin001
	DUEM002	dust emission bin002
	DUEM003	dust emission bin003
	DUEM004	dust emission bin004
	DUEM005	dust emission bin005
Sulfur Aerosols	SO2EMAN	so2 anthropogenic emissions
	SO2EMBB	so2 biomass burning emissions
	SO2EMVE	so2 volcanic (explosive) emissions
	SO2EMVN	so2 volcanic (non-explosive) emissions
	SO4EMAN	so4 anthropogenic emissions
	SUEM003	sulfate emission bin003
Sea Salt Aerosols	SSEM001	sea salt emission bin001
	SSEM002	sea salt emission bin002
	SSEM003	sea salt emission bin003
	SSEM004	sea salt emission bin004
	SSEM005	sea salt emission bin005
Organic Carbon Aerosols	OCEM001	organic carbon emission bin001
	OCEM002	organic carbon emission bin002
	OCEMAN	organic carbon anthropogenic emissions
	OCEMBB	organic carbon biomass burning emissions
	OCEMBG	organic carbon biogenic emissions

Table 3

The bin sizes information for dust and sea salt aerosols.

Aerosol Types	Bin	1	2	3	4	5
Dust aerosols	radius	0.73	1.4	2.4	4.5	8
	radius lower	0.1	1	1.8	3	6
	radius upper	1	1.8	3	6	10
	density	2500	2650	2650	2650	2650
Sea salt aerosols	radius	0.079	0.316	1.119	2.818	7.772
	radius lower	0.03	0.1	0.5	1.5	5
	radius upper	0.1	0.5	1.5	5	10
	density	2200	2200	2200	2200	2200

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/rs14040859/s1, Figure S1 The spatial and monthly variations of AODP values from January to February; Figure S2 The spatial and monthly variations of AODP values from March to April; Figure S3 The spatial and monthly variations of AODP values from May to June; Figure S4 The spatial and monthly variations of AODP values from July to August; Figure S5 The spatial and monthly variations of AODP values from September to October; Figure S6 The spatial and monthly variations of AODP values from November to December.

References

1. Hinds, W.C. Aerosol Technology: Properties, Behavior, and Measurement of Airborne Particles; John Wiley & Sons: Hoboken, NJ, USA, 1999.

2. George, I.J.; Abbatt, J. Heterogeneous oxidation of atmospheric aerosol particles by gas-phase radicals. Nat. Chem.; 2010; 2, pp. 713-722. [DOI: https://dx.doi.org/10.1038/nchem.806]

3. Riipinen, I.; Pierce, J.R.; Yli-Juuti, T.; Nieminen, T.; Häkkinen, S.; Ehn, M.; Junninen, H.; Lehtipalo, K.; Petäjä, T.; Slowik, J. Organic condensation: A vital link connecting aerosol formation to cloud condensation nuclei (CCN) concentrations. Atmos. Chem. Phys.; 2011; 11, pp. 3865-3878. [DOI: https://dx.doi.org/10.5194/acp-11-3865-2011]

4. Ayers, G.P.; Gras, J.L. Seasonal relationship between cloud condensation nuclei and aerosol methanesulphonate in marine air. Nature; 1991; 353, pp. 834-835. [DOI: https://dx.doi.org/10.1038/353834a0]

5. Pöschl, U. Atmospheric aerosols: Composition, transformation, climate and health effects. Angew. Chem. Int. Ed.; 2005; 44, pp. 7520-7540. [DOI: https://dx.doi.org/10.1002/anie.200501122]

6. Haywood, J.; Boucher, O. Estimates of the direct and indirect radiative forcing due to tropospheric aerosols: A review. Rev. Geophys.; 2000; 38, pp. 513-543. [DOI: https://dx.doi.org/10.1029/1999RG000078]

7. Wei, J.; Huang, W.; Li, Z.; Xue, W.; Peng, Y.; Sun, L.; Cribb, M. Estimating 1-km-resolution PM2. 5 concentrations across China using the space-time random forest approach. Remote Sens. Environ.; 2019; 231, 111221. [DOI: https://dx.doi.org/10.1016/j.rse.2019.111221]

8. Wei, J.; Li, Z.; Cribb, M.; Huang, W.; Xue, W.; Sun, L.; Guo, J.; Peng, Y.; Li, J.; Lyapustin, A. Improved 1 km resolution PM 2.5 estimates across China using enhanced space-time extremely randomized trees. Atmos. Chem. Phys.; 2020; 20, pp. 3273-3289. [DOI: https://dx.doi.org/10.5194/acp-20-3273-2020]

9. Holben, B.N.; Eck, T.F.; Slutsker, I.A.; Tanre, D.; Buis, J.P.; Setzer, A.; Vermote, E.; Reagan, J.A.; Kaufman, Y.J.; Nakajima, T. AERONET-A federated instrument network and data archive for aerosol characterization. Remote Sens. Environ.; 1998; 66, pp. 1-16. [DOI: https://dx.doi.org/10.1016/S0034-4257(98)00031-5]

10. Sioris, C.E.; Abboud, I.; Fioletov, V.E.; McLinden, C.A. AEROCAN, the Canadian sub-network of AERONET: Aerosol monitoring and air quality applications. Atmos. Environ.; 2017; 167, pp. 444-457. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2017.08.044]

11. Sano, I.; Mukai, S.; Yamano, M.; Takamura, T.; Nakajima, T.; Holben, B. Calibration and validation of retrieved aerosol properties based on AERONET and SKYNET. Adv. Space Res.; 2003; 32, pp. 2159-2164. [DOI: https://dx.doi.org/10.1016/S0273-1177(03)90538-2]

12. Che, H.; Zhang, X.; Xia, X.; Goloub, P.; Holben, B.; Zhao, H.; Wang, Y.; Zhang, X.; Wang, H.; Blarel, L. Ground-based aerosol climatology of China: Aerosol optical depths from the China Aerosol Remote Sensing Network (CARSNET) 2002–2013. Atmos. Chem. Phys.; 2015; 15, pp. 7619-7652. [DOI: https://dx.doi.org/10.5194/acp-15-7619-2015]

13. Subba, T.; Gogoi, M.M.; Pathak, B.; Bhuyan, P.K.; Babu, S.S. Recent trend in the global distribution of aerosol direct radiative forcing from satellite measurements. Atmos. Sci. Lett.; 2020; 21, e975. [DOI: https://dx.doi.org/10.1002/asl.975]

14. Lyapustin, A.; Wang, Y.; Laszlo, I.; Kahn, R.; Korkin, S.; Remer, L.; Levy, R.; Reid, J.S. Multiangle implementation of atmospheric correction (MAIAC): 2. Aerosol algorithm. J. Geophys. Res.; 2011; 116, D03211.

15. Lyapustin, A.I.; Wang, Y.; Laszlo, I.; Hilker, T.; Hall, F.G.; Sellers, P.J.; Tucker, C.J.; Korkin, S.V. Multi-angle implementation of atmospheric correction for MODIS (MAIAC): 3. Atmospheric correction. Remote Sens. Environ.; 2012; 127, pp. 385-393. [DOI: https://dx.doi.org/10.1016/j.rse.2012.09.002]

16. Lee, H.J.; Liu, Y.; Coull, B.A.; Schwartz, J.; Koutrakis, P. A novel calibration approach of MODIS AOD data to predict PM 2.5 concentrations. Atmos. Chem. Phys.; 2011; 11, pp. 7991-8002. [DOI: https://dx.doi.org/10.5194/acp-11-7991-2011]

17. Lyapustin, A.; Wang, Y.; Korkin, S.; Huang, D. MODIS Collection 6 MAIAC algorithm. Atmos. Meas. Tech.; 2018; 11, pp. 5741-5765. [DOI: https://dx.doi.org/10.5194/amt-11-5741-2018]

18. Guo, J.; Zhang, X.; Wu, Y.; Zhaxi, Y.; Che, H.; La, B.; Wang, W.; Li, X. Spatio-temporal variation trends of satellite-based aerosol optical depth in China during 1980–2008. Atmos. Environ.; 2011; 45, pp. 6802-6811. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2011.03.068]

19. Bibi, H.; Alam, K.; Chishtie, F.; Bibi, S.; Shahid, I.; Blaschke, T. Intercomparison of MODIS, MISR, OMI, and CALIPSO aerosol optical depth retrievals for four locations on the Indo-Gangetic plains and validation against AERONET data. Atmos. Environ.; 2015; 111, pp. 113-126. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2015.04.013]

20. Hsu, N.C.; Gautam, R.; Sayer, A.M.; Bettenhausen, C.; Li, C.; Jeong, M.J.; Tsay, S.C.; Holben, B.N. Global and regional trends of aerosol optical depth over land and ocean using SeaWiFS measurements from 1997 to 2010. Atmos. Chem. Phys.; 2012; 12, pp. 8037-8053. [DOI: https://dx.doi.org/10.5194/acp-12-8037-2012]

21. Qin, W.; Liu, Y.; Wang, L.; Lin, A.; Xia, X.; Che, H.; Bilal, M.; Zhang, M. Characteristic and Driving Factors of Aerosol Optical Depth over Mainland China during 1980–2017. Remote Sens.; 2018; 10, 1064. [DOI: https://dx.doi.org/10.3390/rs10071064]

22. Lambert, A.; Grainger, R.G.; Remedios, J.J.; Rodgers, C.D.; Corney, M.; Taylor, F.W. Measurements of the evolution of the Mt. Pinatubo aerosol cloud by ISAMS. Geophys. Res. Lett.; 1993; 20, pp. 1287-1290. [DOI: https://dx.doi.org/10.1029/93GL00827]

23. Minnis, P.; Harrison, E.F.; Stowe, L.L.; Gibson, G.G.; Denn, F.M.; Doelling, D.R.; Smith, W.L. Radiative climate forcing by the Mount Pinatubo eruption. Science; 1993; 259, pp. 1411-1415. [DOI: https://dx.doi.org/10.1126/science.259.5100.1411]

24. Sanap, S.D. Global and regional variations in aerosol loading during COVID-19 imposed lockdown. Atmos. Environ.; 2021; 246, 118132. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2020.118132]

25. Ranjan, A.K.; Patra, A.K.; Gorai, A.K. Effect of lockdown due to SARS COVID-19 on aerosol optical depth (AOD) over urban and mining regions in India. Sci. Total Environ.; 2020; 745, 141024. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2020.141024]

26. Filonchyk, M.; Hurynovich, V.; Yan, H.; Gusev, A.; Shpilevskaya, N. Impact assessment of COVID-19 on variations of SO2, NO2, CO and AOD over East China. Aerosol Air Qual. Res.; 2020; 20, pp. 1530-1540. [DOI: https://dx.doi.org/10.4209/aaqr.2020.05.0226]

27. Filonchyk, M.; Hurynovich, V.; Yan, H. Impact of Covid-19 lockdown on air quality in the Poland, Eastern Europe. Environ. Res.; 2021; 198, 110454. [DOI: https://dx.doi.org/10.1016/j.envres.2020.110454]

28. Qin, W.; Fang, H.; Wang, L.; Wei, J.; Zhang, M.; Su, X.; Bilal, M.; Liang, X. MODIS high-resolution MAIAC aerosol product: Global validation and analysis. Atmos. Environ.; 2021; 264, 118684. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2021.118684]

29. Mehta, M.; Singh, R.; Singh, A.; Singh, N.; Anshumali.,. Recent global aerosol optical depth variations and trends—A comparative study using MODIS and MISR level 3 datasets. Remote Sens. Environ.; 2016; 181, pp. 137-150. [DOI: https://dx.doi.org/10.1016/j.rse.2016.04.004]

30. Jaeglé, L.; Quinn, P.K.; Bates, T.S.; Alexander, B.; Lin, J.T. Global distribution of sea salt aerosols: New constraints from in situ and remote sensing observations. Atmos. Chem. Phys.; 2011; 11, pp. 3137-3157. [DOI: https://dx.doi.org/10.5194/acp-11-3137-2011]

31. Vignati, E.; Facchini, M.C.; Rinaldi, M.; Scannell, C.; Ceburnis, D.; Sciare, J.; Kanakidou, M.; Myriokefalitakis, S.; Dentener, F.; O’Dowd, C.D. Global scale emission and distribution of sea-spray aerosol: Sea-salt and organic enrichment. Atmos. Environ.; 2010; 44, pp. 670-677. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2009.11.013]

32. Mahowald, N.; Albani, S.; Kok, J.F.; Engelstaeder, S.; Scanza, R.; Ward, D.S.; Flanner, M.G. The size distribution of desert dust aerosols and its impact on the Earth system. Aeolian Res.; 2014; 15, pp. 53-71. [DOI: https://dx.doi.org/10.1016/j.aeolia.2013.09.002]

33. Goudie, A.S.; Middleton, N.J. Saharan dust storms: Nature and consequences. Earth-Sci. Rev.; 2001; 56, pp. 179-204. [DOI: https://dx.doi.org/10.1016/S0012-8252(01)00067-8]

34. Ganor, E.; Mamane, Y. Transport of Saharan dust across the eastern Mediterranean. Atmos. Environ. (1967); 1982; 16, pp. 581-587. [DOI: https://dx.doi.org/10.1016/0004-6981(82)90167-6]

35. McKendry, I.G.; Strawbridge, K.B.; O’Neill, N.T.; Macdonald, A.M.; Liu, P.S.; Leaitch, W.R.; Anlauf, K.G.; Jaegle, L.; Fairlie, T.D.; Westphal, D.L. Trans-Pacific transport of Saharan dust to western North America: A case study. J. Geophys. Res. Atmos.; 2007; 112. [DOI: https://dx.doi.org/10.1029/2006JD007129]

36. Swap, R.; Garstang, M.; Greco, S.; Talbot, R.; Kallberg, P. Saharan dust in the Amazon Basin. Tellus B.; 1992; 44, pp. 133-149. [DOI: https://dx.doi.org/10.3402/tellusb.v44i2.15434]

37. Cooke, W.F.; Wilson, J.J. A global black carbon aerosol model. J. Geophys. Res. Atmos.; 1996; 101, pp. 19395-19409. [DOI: https://dx.doi.org/10.1029/96JD00671]

38. Reddington, C.L.; McMeeking, G.; Mann, G.W.; Coe, H.; Frontoso, M.G.; Liu, D.; Flynn, M.; Spracklen, D.V.; Carslaw, K.S. The mass and number size distributions of black carbon aerosol over Europe. Atmos. Chem. Phys.; 2013; 13, pp. 4917-4939. [DOI: https://dx.doi.org/10.5194/acp-13-4917-2013]

39. Koch, D.; Schulz, M.; Kinne, S.; McNaughton, C.; Spackman, J.R.; Balkanski, Y.; Bauer, S.; Berntsen, T.; Bond, T.C.; Boucher, O. Evaluation of black carbon estimations in global aerosol models. Atmos. Chem. Phys.; 2009; 9, pp. 9001-9026. [DOI: https://dx.doi.org/10.5194/acp-9-9001-2009]

40. Eliseev, A.V. Impact of tropospheric sulphate aerosols on the terrestrial carbon cycle. Global Planet. Change; 2015; 124, pp. 30-40. [DOI: https://dx.doi.org/10.1016/j.gloplacha.2014.11.005]

41. Smith, S.J.; van Aardenne, J.; Klimont, Z.; Andres, R.J.; Volke, A.; Delgado Arias, S. Anthropogenic sulfur dioxide emissions: 1850–2005. Atmos. Chem. Phys.; 2011; 11, pp. 1101-1116. [DOI: https://dx.doi.org/10.5194/acp-11-1101-2011]

42. Penner, J.E. Carbonaceous Aerosols Influencing Atmospheric Radiation: Black and Organic Carbon; Lawrence Livermore National Lab.: Livermore, CA, USA, 1994.

43. Kirchstetter, T.W.; Novakov, T.; Hobbs, P.V. Evidence that the spectral dependence of light absorption by aerosols is affected by organic carbon. J. Geophys. Res. Atmos.; 2004; 109, D20208. [DOI: https://dx.doi.org/10.1029/2004JD004999]

44. Lack, D.A.; Tie, X.X.; Bofinger, N.D.; Wiegand, A.N.; Madronich, S. Seasonal variability of secondary organic aerosol: A global modeling study. J. Geophys. Res. Atmos.; 2004; 109, D03203. [DOI: https://dx.doi.org/10.1029/2003JD003418]

45. Fujii, Y.; Iriana, W.; Oda, M.; Puriwigati, A.; Tohno, S.; Lestari, P.; Mizohata, A.; Huboyo, H.S. Characteristics of carbonaceous aerosols emitted from peatland fire in Riau, Sumatra, Indonesia. Atmos. Environ.; 2014; 87, pp. 164-169. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2014.01.037]

46. Randles, C.A.; Da Silva, A.M.; Buchard, V.; Colarco, P.R.; Darmenov, A.; Govindaraju, R.; Smirnov, A.; Holben, B.; Ferrare, R.; Hair, J. et al. The MERRA-2 Aerosol Reanalysis, 1980 Onward. Part I: System Description and Data Assimilation Evaluation. J. Climate.; 2017; 30, pp. 6823-6850. [DOI: https://dx.doi.org/10.1175/JCLI-D-16-0609.1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29576684]

47. Naik, P.; D’Sa, E.J. Empirical Orthogonal Function (EOF) Analysis of Sea-Surface Temperature and Chlorophyll in the Eastern Bering Sea. SPIE: Bellingham, DC, USA, 2010; Volume 7825, 782503-8.

48. Gocic, M.; Trajkovic, S. Analysis of changes in meteorological variables using Mann-Kendall and Sen’s slope estimator statistical tests in Serbia. Global Planet. Change; 2013; 100, pp. 172-182. [DOI: https://dx.doi.org/10.1016/j.gloplacha.2012.10.014]

49. Anselin, L. A local indicator of multivariate spatial association: Extending Geary’s C. Geogr. Anal.; 2019; 51, pp. 133-150. [DOI: https://dx.doi.org/10.1111/gean.12164]

50. Gu, L.; Baldocchi, D.D.; Wofsy, S.C.; Munger, J.W.; Michalsky, J.J.; Urbanski, S.P.; Boden, T.A. Response of a deciduous forest to the Mount Pinatubo eruption: Enhanced photosynthesis. Science; 2003; 299, pp. 2035-2038. [DOI: https://dx.doi.org/10.1126/science.1078366]

51. Schwarz, J.P.; Spackman, J.R.; Gao, R.S.; Watts, L.A.; Stier, P.; Schulz, M.; Davis, S.M.; Wofsy, S.C.; Fahey, D.W. Global-scale black carbon profiles observed in the remote atmosphere and compared to models. Geophys. Res. Lett.; 2010; 37, L11812. [DOI: https://dx.doi.org/10.1029/2010GL044372]

52. Myhre, G.; Myhre, C.; Samset, B.H.; Storelvmo, T. Aerosols and their relation to global climate and climate sensitivity. Nat. Educ. Knowl.; 2013; 4, 7.