3.1. Spatial Distribution of Average Physicochemical Characteristics of the Aerosol above Lake Baikal Based on the Results of Long-Term Studies
According to the data from Arctic and Antarctic Research Institute [
33], from 24 to 27 July 2019, the Baikal region was under the influence of the cyclone that came from the northwest, with low clouds and rainy weather at times; the rains cleared the atmosphere. On 23 July, satellite monitoring recorded a wildfire on the east coast of Lake Baikal near Sosnovka Bay, the smoke of which spread over the water area of Northern Baikal. From 26 July to 1 August, many fire sites were recorded north of Lake Baikal (
Figure 2) [
29]. During the expedition, the total area of wildfires in Siberia reached 3 million ha [
20], the smog from which was recorded in the atmosphere above the entire water area of the lake. It has been shown many times that the emission of combustion products (gases, soot, and aerosol) into the atmosphere negatively affects not only the quality but also its optical and microphysical properties [
34].
For the determination of the influence of different sources on the formation of the atmospheric aerosol above the lake, we constructed back trajectories of air mass transport for each day of the expedition from 24 July to 4 August 2019 (
Figure 3) [
29]. An analysis of back trajectories revealed that the smoke from the wildfire from 25 to 27 July near Sosnovka Bay spread over the lake from the southeast to the northwest. From 28 July to 1 August, air masses came to the lake from the northern areas of the Irkutsk Region and southern areas of Yakutia were engulfed in wildfires, the smoke emission from which spread over Northern and Central Baikal. From 2 to 3 August, the directions of air currents changed; clean air from Mongolia filled the southern basin of Lake Baikal.
Figure 4A–D shows the values of meteorological parameters at altitudes of up to 5000 m in the atmosphere of Central Baikal during the smog above the lake. In the lake basin, we recorded an inverse temperature distribution in the lower one-kilometer layer as well as abrupt changes in wind direction and wind speed, which resulted in a complex pattern of the aerosol distribution [
33]. From the evening on 29 July to the night on 30 July, the smoke aerosol descended to the water surface of the lake and spread over the water area.
3.2. Spatial Distribution of the Mass, MA (μg m−3), and Number, NA (cm−3), Concentrations of Aerosol Particles
Table 1 shows the average mass concentrations of aerosol particles along the RV route in 2019. At the beginning of the route, we observed the elevated mass concentrations of particles during moorings of the RV in the area of Southern Baikal, in the Listvyanka settlement and the Baikalsk town (
Table 1).
In the absence of the influence of wildfires, the anthropogenic factor mainly affected the increase in the mass concentration of particles. Along the eastern coast of the southern basin, there are large settlements with coal-fired boilers and suburban houses with stove heating, and federal highways and railways pass, polluting the atmosphere in this part of the lake. Complex orography of this area and the mountainous margin of the Khamar-Daban range with heights above 2300 m do not contribute to the dispersion of pollutants in the atmosphere. Other areas with the highest mass concentrations of aerosol were located on the east coast of Northern and Central Baikal on 30–31 July 2019 in Chivyrkuy and Barguzin bays. In the atmosphere of this area, we observed a smog from wildfires, which spread to the south along the RV route from Turka to Kharauz (delta of the Selenga River) and to Boyarsk. Background mass concentrations of particles were recorded along the northwestern coast on 25–27 July (1.5–5.9 μg/m
3) and the northeastern coast of the lake on 28–29 July (1.7–6.0 μg/m
3) (see
Figure 1B,
Table 1).
The distribution dynamics of the mass concentration of aerosol above Lake Baikal, depending on the anthropogenic factors, meteorological conditions and wildfires, differs significantly from year to year. In
Table 2, we combined the average mass concentrations of the aerosol above different areas of the lake, which were measured from 2016 to 2019. The highest aerosol concentrations were recorded in July 2016 and July–August 2019 during extensive wildfires. In July 2016, in Southern and Central Baikal, there were local wildfires that engulfed the west coast of the lake; we observed significant air pollution with smoke. This was also observed (see
Figure 2) in July–August 2019 when on the east coast of the lake near Sosnovka Bay, there was a local fire. Moreover, smoke aerosol was recorded from extensive sites of wildfires in the northern areas of the Irkutsk Region, the Krasnoyarsk Territory and the Republic of Sakha (Yakutia). In 2018, there was also some increase in the mass concentration of aerosol, which was associated both with local anthropogenic sources and regional wildfires [
32].
The lowest mass concentration of aerosol was recorded in July 2017 when calm sunny weather without fires prevailed throughout the entire cruise of the RV.
The increase in the mass concentration of particles was correlated well with their number concentration.
Figure 5 shows the dispersed composition of the aerosol. The proportion of submicron aerosol fraction is predominant and averages 99.1–99.5% of the total number concentration of particles above the lake.
The number concentration of particles in July 2016 was elevated throughout the entire route of the RV and reached the maximum in Barguzin Bay (
Figure 5A). In August 2016, an increase in the number concentration was also recorded in Barguzin Bay (
Figure 5B). The number concentration, as well as the mass concentration, was the lowest in 2017, but on 20 July, we recorded its increase to 101.7 cm
−3 near the delta of the Selenga River (
Figure 4C), which was associated with the change in meteorological situation manifested in the strengthening of wind and the removal of polluted air masses along the valley of the Selenga River.
The dynamics of the number concentration of aerosol particles in 2019 indicates that, like in 2016–2017, the changes affected a submicron part of the aerosol (up to 1.0 μm in size) (
Figure 5D). Air carrying mainly small combustion particles entered the lake basin. At the same time, an increase in the concentration of submicron particles was significant (two-fivefold) and continued from 30 July to 2 August, until the direction of air currents changed (see
Figure 3).
3.3. Spatiotemporal Distribution of Ionic Concentrations in Aerosol above the Water Area of Lake Baikal
Many researchers have indicated the peculiar Baikal landscape surroundings and climate. The presence of mountain ranges around the Baikal basin leads to the formation of a unique climate both in seasonal and diurnal changes in the main meteorological characteristics, which long-term monitoring observations at the network of weather stations rather densely located on the lake coast confirm [
35]. The most pollutant-free atmosphere above the water surface of Lake Baikal is located above the central abyssal part of the lake. Based on the 2005–2008 data, total ionic concentrations in the composition of aerosol varied from 0.1 to 0.7 µg/m
3. The bulk of dissolved aerosol components was in the submicron size spectrum, accounting for approximately 60% of the total mass. Heterogeneity in the distribution of impurities above the surface of Lake Baikal was mainly observed in the estuarine areas of the large tributaries and bays as well as above Southern Baikal [
30]. In Southern Baikal, substances are removed to the lake along the Angara River valley from the industrial areas of the Southern Baikal region in the form of poorly dispersed plumes during mesoscale jet stream transport in the atmospheric boundary layer [
36]. Averaging the total ionic concentrations in the aerosols above each Baikal basin confirmed the previously revealed [
37] trend of the latitudinal distribution of impurities with the highest accumulation above Southern Baikal and the lowest one above Northern Baikal [
30].
Recent research (from 2010 to 2019) has indicated the significant variability of the revealed trend in the distribution of impurities in the near-water atmosphere above Lake Baikal due to climate change and an increase in wildfires in Siberia (
Table 3).
The frequency of the average ionic concentrations below 1.0 µg/m
3 from 2010 to 2019 was approximately 50%, which corresponds to the natural background. The frequency of the average ionic concentrations above 2.0 µg/m
3 became more than 30% (
Figure 6). Taking into account that the impact of anthropogenic sources during the study period was relatively constant, wildfires in Siberia affected significantly the variability of soluble substances in the near-water atmosphere of the lake. The influence of smoke aerosol increased together with such meteorological events as high temperature, low humidity, and increased wind speed.
Thus, abnormally dry windy weather in June 2012 in conjunction with the smoke aerosol increased the air pollution above the entire water area of the lake (see
Table 3) [
31]. Wildfires in 2015 and 2016 and especially in 2019 also contributed to the increase in pollutants in the near-water atmosphere above the lake (see
Table 2).
Wildfires in July–August 2019, which engulfed large areas of Siberia, significantly affected the distribution of impurities in the atmosphere above the lake (see
Table 1). Form 24 July to 4 August 2019, the total ionic concentrations varied from 1.2 to 11.2 μg/m
3, and their upper limit was the highest over the entire long-term observation period. Based on averaged data, the predominant ions in the aerosol composition were Na
+, NH
4+, NO
3−, and SO
42− in the southern basin and NH
4+, NO
3− and SO
42− in the central and northern basins. In the aerosol of Central Baikal, where a local fire was recorded, the concentrations of K
+ ions were higher than in other areas. Among other ions, the concentrations of Cl
− and Ca
2+ were increased in aerosol (
Table 4).
In previous years of measurements, during the periods of smoke-free atmosphere, the major ions in the aerosol composition were NH
4+, Ca
2+, and SO
42−; under the impact of meteorological factors–Na
+, K
+, Ca
2+, Cl
−, and SO
42−; in the atmosphere with smoke from nearby fires–Na
+, NH
4+, K
+, Ca
2+, Cl
−, and SO
42−; in the atmosphere with smoke from distant fires–NH
4+, K
+, Cl
−, NO
3−, and SO
42− [
30,
31].
Depending on the type of biomass burnt, volatile inorganic elements condense in the form of chlorides and sulfates into a group of particles enriched in potassium [
38,
39,
40]. We calculated enrichment factors for these elements for each sample using the formula from [
41].
where
K–enrichment factor,
-the concentration of
ith element relative to Na
+ in the aerosol (
aer) and Baikal water (
BW) [
42].
Enrichment factors for the K
+ (K = 2 ÷ 14) and SO
42− (K = 5 ÷ 25) ions had similar dynamics and were the highest in the aerosol along the east coast, from Northern (Khakusy Bay) to Southern (the Boyarsk settlement) Baikal (see
Figure 1B), where smog was recorded. We determined the highest enrichment factors for the Cl
− (K = 5 ÷ 26) ions in the aerosol on the west coast during the RV route along the southern margin of the lake. The recalculation of the ionic enrichment factors relative to the concentrations of the elements in the Earth’s crust revealed the same pattern as in the calculation relative to the concentration of ions in Baikal water. At the same time, the enrichment factors of Ca
2+ were close to one above the entire water area of the lake. This indicates a predominantly soil origin of the aerosol affected by the smoke emission.
We carried out a correlation analysis of the ionic composition and the dispersion of the aerosol. The analysis revealed that the NH4+ and K+ ions are absorbed mainly on submicron aerosol particles of 0.3–1.0 μm in size. The correlation coefficients (r) varied from 0.92–0.95 for the particles ranging between 0.3 and 0.5 μm to 0.82–0.84 for the particles with a size of 0.1 μm. The large particles of 2.5–5.0 μm showed a lower correlation of these ions (r = 0.26–0.42). We also determined a high correlation for the concentrations of the NO3− and SO42− with the particles ranging from 0.3 to 0.5 μm (r = 0.55–0.66) and the 1.0 μm particles (r = 0.44–0.54). The concentrations of the Ca2+ (r = 0.25) and NO3− (r = 0.31) showed loose correlation with particles of 5.0 μm in size.
3.4. Polyaromatic Compounds (PAHs) in the Composition of the Near-Water Aerosol
Wildfires are considered one of the PAH sources entering natural environments. PAH compounds resulting from burning natural materials are called pyrogenic. These include compounds with a high molecular weight. Based on the sum concentrations of fourteen identified PAHs (naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[e]pyrene, and benzo[a]pyrene), we investigated in detail the influence of wildfires on the pollution of the near-water atmosphere above Lake Baikal in 2015 and 2016. In August 2015, the fire site was a short distance from on the coast Central Baikal; in 2016, the atmosphere above Lake Baikal was covered with smoke from regional fires that were spread over time. During the so-called “fresh” fire in August 2015, there were the highest concentrations of PAHs, in total varying from 1.6 ng/m
3 to 152.1 ng/m
3. In contrast, in May 2015, in the absence of fires, the sum of concentrations of PAHs in the aerosol ranged from 0.9 to 2.4 ng/m
3. Phenanthrene, the proportion of which varied from 8 to 66% of the sum of concentrations of PAHs, naphthalene (0.4–66%), fluorene (0–46%), fluoranthene (1–38%), and pyrene (1–24%) dominated the identified PAHs. According to the published data, these compounds resulted from burning coniferous and larch forests [
43,
44,
45]. Their total number reached on average 81% of all determined compounds, with a predominance of phenanthrene (on average 27%) and naphthalene (on average 26%). Notably, according to [
46] naphthalene and phenanthrene also resulted from the combustion of fossil fuels. The influence of pyrogenic and petrogenic PAH sources can be determined from the ratio of the fluoranthene/(fluoranthene + pyrene) concentrations. The ratio above 0.4 characterizes the presence of pyrogenic sources, and that below 0.4–petrogenic sources [
47]. In our case, the concentration ratio of these compounds ranged from 0.41 to 0.64 during the entire cruise, which indicates pyrogenic sources.
In July 2016, the sum of concentrations of PAHs in the aerosol was lower and varied within a narrower range from 6.3 to 42.4 ng/m3. The concentrations of individual PAHs, as well as their share, were lower. In 2016, the set of predominant PAHs in the aerosol differed from its composition in 2015. Perhaps, the long presence of the smoke aerosol in the atmosphere contributed to the deposition of heavy molecules. Naphthalene (0.3–82%), chrysene (0–34%), phenanthrene (0.1–27%), pyrene (0–25%), fluorene (0.0–10.4%), and fluoranthene (0.3–5.6%) dominated PAHs. During almost the entire cruise, the ratio of the fluoranthene/(fluoranthene + pyrene) concentrations was below 0.4. Only by the end of the cruise, this ratio changed to 0.47–0.57 near the delta of the Selenga River. An analysis of the variability in the concentrations of PAHs has revealed that the composition of PAHs in the aerosol coming from distant (regional) wildfires differs significantly from the PAH composition resulting from the sites of local fires in the immediate vicinity of the lake.
In 2019, the atmosphere experienced the influence of distant and local fires.
Table 5 shows the variability of PAH concentrations in the aerosol in July 2019.
In 2019, the sum of PAHs ranged from 0.11 to 4.62 ng/m
3 and were significantly lower than in 2015 and 2016, despite the wider range of various compounds (18 PAH compounds) (
Table 6). Among individual PAHs, acenaphthylene (1.8–10.5%), phenanthrene (12.6–17.9%), fluoranthene (9.4–14.7%), benzo[b]fluoranthene (5.8–9.8%), benzo[a]pyrene (6.2–7.7%), indeno[1,2,3-c,d]pyrene (2.7–12.2%), and benzo[g,h,i]perylene (2.4–13.8%) prevailed in the aerosol of Southern Baikal and along the west coast. Most of these compounds are indicators of the anthropogenic pollution [
46,
48]. The benzo[a]pyrene/benzo[g,h,i]perylene ratio below 0.6 indicates emissions from vehicles near the Baikalsk town (Southern Baikal), and that above 0.6–from the stationary sources in the aerosol sampled along the west coast from Southern to Central Baikal. The indeno[1,2,3-c,d]pyrene/indeno[1,2,3-c,d]pyrene+ benzo[g,h,i]perylene ratio below 0.5 in the aerosol composition of these areas indicates the combustion of liquid fuel [
47]. The PAH composition changed in the aerosol under the influence of the smog in Northern and Central Baikal (see
Table 6). Phenanthrene (3.4–14.3%), fluoranthene (11.4–24.5%), benzo[b]fluoranthene (10.7–20.7%), indeno[1,2,3-c,d]pyrene (4.2–11.8%), benzo[k]fluoranthene (3.8–11.7%), pyrene (9.9–15.4%), and benzo[g,h,i]perylene (3.9–10.4%) were dominant compounds. The diagnostic ratios of the PAH concentrations indicate mainly pyrogenic sources of these compounds.
The comparison of the 2019 data with similar studies in 2015 and 2016 indicated that the PAH composition in the aerosol of this period was more identical to the PAH composition in 2016. The low contribution of the predominant phenanthrene, fluoranthene, and pyrene compounds in the aerosol indicates the influence of smoke emission from wildfires occurring in distant areas.
3.5. Elemental Composition of the Near-Water Aerosol above Lake Baikal
A detailed analysis of the variability of the concentrations for 26 elements in the aerosol above the lake revealed several periods during the cruise, in which their concentrations increased relative to the average values calculated for all the samples obtained (
Figure 7,
Table 7).
Table 7 shows the average concentrations of the elements for each period defined, from 1 to 8 (see
Figure 7), in comparison with the overall average value. The Background column shows the averaged data for the periods with the minimum concentrations of elements, and the values in bold type those exceeding the average values (see
Table 7). Based on the data averaged for the entire period, with the concentrations above 1.0 ng/m
3, we identified the elements (Fe, Al, Cu, and Zn), whose contributions varied up to 80.0% of their total amount. The contribution of the elements with the concentrations ranging within 0.1 ÷ 1.0 ng/m
3 was approximately 15.0%; Mn, V, Cr, Ni, Sr, Ti, As, Se, Ba, Pb, and Cd were most common in this concentration range.
The concentrations of elements in the aerosol during the defined sampling periods (No.1 ÷ No.8 see
Table 7,
Figure 7) differed in absolute values of the concentrations, which varied upward and downward. The highest concentrations of elements were recorded in the aerosol collected during the RV mooring in the Baikalsk town on 25 July 2019 (No. 1, see
Table 7). As noted above, the anthropogenic factor affects more the southern margin of the lake. The area limited by high-mountain ranges and the northwestern transport of air masses prevailing here contribute to the accumulation of impurities in the atmosphere.
Along the west coast of the lake, from the Listvyanka settlement to the Khuzhir settlement (No. 2, No. 3,
Table 7), elevated concentrations of elements in the aerosol were mainly of lithogenic origin. The Southwest coast of Southern Baikal and the area of the Maloye More Strait have weak easily decaying soil and vegetation cover with steppe and forest-steppe zones. The northernmost margin of the lake (No. 4,
Table 7), as well as the southern one, experiences the anthropogenic impact due to the large settlements (the Severobaikalsk town and the Nizhneangarsk settlement) located on the coast.
Of greatest interest was the elemental composition of the aerosol in the influence zone of wildfires (No. 5). Owing to the studies conducted in different areas of Siberia and the East Kazakhstan Region, the chemical elements were divided into two groups in terms of the nature of their behavior during wildfires: the group of active air migrants, including Cd, Pb, As, Sb, Se, Mn, Zn, U, and Sr, and the group of elements that accumulated in the burned area were Cr, Ni, Co, V, Th, Mg, K, Na, Ca, and Al [
49]. During our studies, when the atmosphere was affected by smoke emissions, the following elements had elevated concentrations: Ti, V, Mn, Fe, Ni, Co, Cu, Zn, As, Se, Sr, Cd, Ba, and Pb, from macroelements allocated by K. It is noteworthy that the emissions from industrial enterprises and soil dust are carried along with plumes from wildfires [
50]. The effect of soil dust, emissions from the industrial enterprises, together with smoke emissions from wildfires was most manifested in the elemental composition of the aerosol near the delta of the Selenga River (No. 6) and the Boyarsk settlement (No. 7). With a change in the direction of wind currents at the final sampling site (No. 8), the elemental composition of the aerosol again showed an increase in the concentrations of almost all the elements.
The enrichment factors calculated according to the formula 1 relative to Al revealed a significant enrichment of the aerosol (K > 1000) with Be, B, Zn, Mo, U, Sn, As, W, Tl, Sb, Cd, Ag, and Se, elements of both lithophilic and anthropogenic origin. Factors in the range 100 < K < 1000 were determined for Ni, Co, Li, Th, Cu, and Pb. High enrichment factors for the elements were determined during the passage of the RV at the sites 3–5, 7, and 8 (see
Figure 7). At site 6, the most exposed to the smoke emission, as well as during the passage of the RV from Listvyanka to Buguldeika (site 2), the enrichment factors for the elements were lower.
Factor analysis was used to identify groups of elements correlating between themselves. The concentration of each element in an individual sample is considered a result of the joint impact of some sources (factors) that must be identified during analysis. In other words, such groups are formed from the elements that have significant loadings on the given factor. Factor loadings are calculated from the correlation matrix of the elements. To interpret the analysis results, the factors that add up to the maximum variance in this series of observations are selected in the aerosol (
Table 8,
Figure 8A). The first four factors explain 88% of the total variance of the variables. We determined that the soil in the region is the main source of the aerosol particles in the near-water atmosphere of the lake, but in each factor, there are additional sources of variability in the concentrations of the elements.
During our measurements, the invasion of the air masses from the northern areas of the Irkutsk Region and the Krasnoyarsk Territory, which contained smoke from wildfires, was an additional source of the formation of aerosol particles in the atmosphere above the lake (see
Figure 2 and
Figure 3). This can be seen in
Figure 8B where Factor I is responsible for the soil aerosol and reflects the contribution of the sources in the aerosol above Southern and Central Baikal. Factor II marked in red in
Figure 8B shows the contribution of the emission from wildfires in the composition of the aerosol along the east coast of Lake Baikal.
Using
Table 8, we can describe Factor II as the joint contribution of the group of the following elements in the formation of the elemental composition of aerosol particles: B, Mn, Zn, As, Sr, Cd, and Pb. This allowed us to use them as markers during the entry of air masses from the areas prone to wildfires. Aerosol samples are rather well separated in elements in the plane of Factors I and II (see
Figure 8B). Aerosol samples exposed to smoke emission from wildfires are marked in red; those collected near settlements (Baikalsk, Severobaikalsk, and Nizhneangarsk), are marked in green, and aerosol collected in clean areas are marked in blue.
Similarly, we conducted a factor analysis for the ionic (soluble) part of aerosols (
Table 9,
Figure 9). Accumulation and dissipation of ions in the soluble part of aerosol occurs differently than in the solid part due to the constant watering of the particles in the air. Therefore, the distribution of samples on the plane of the first two factors has shown in this case (
Figure 9B) that the first factor is the effect of the forest fire emissions (the incidence is marked in red), and only the second factor is responsible for the soil aerosol (marked in blue). Based on
Table 9, we can predict that during the influx of smoke air masses, the soluble part of aerosols is enriched with the NH
4+, K
+, SO
42−, and Li
+.