Analysis of Mineral Aerosol in the Surface Layer over the Caspian Lowland Desert by the Data of 12 Summer Field Campaigns in 2002–2020

: In-situ knowledge on characteristics of mineral aerosols is important for weather and climate prediction models, particularly for modeling such processes as the entrainment, transport and deposition of aerosols. However, ﬁeld measurements of the dust emission ﬂux, dust size distribution and its chemical composition under realistic wind conditions remain rare. In this study, we present experimental data over annual expeditions in the arid and semi-arid zones of the Caspian Lowland Desert (Kalmykia, south of Russia); we evaluate characteristics of mineral aerosol concentration and ﬂuxes, estimate its chemical composition and calculate its long-distance transport characteristics. The mass concentration in different years ranges from several tens to several hundred of µ g m − 3 . The signiﬁcant inﬂuence of wind velocity on the value of mass and counting concentration and on the proposed entrainment mechanisms is conﬁrmed. An increased content of anthropogenic elements (S, Sn, Pb, Bi, Mo, Ag, Cd, Hg, etc.), which is characteristic for all observation points in the south of the European Russia, is found. The trajectory analysis show that long-range air particles transport from the Caspian Lowland Desert to the central regions of European Russia tends to increase in the recent decades. writing—original preparation, N.V.V.; writing—review and editing, and M.S.A.; visualization, O.G.C., A.V.C., N.V.V., D.P.G., and supervision, O.G.C.; administration, funding


Introduction
Mineral aerosol is the most common one among various aerosol species. It contributes more than half of the global aerosol amount [1][2][3]. In arid zones, the degree of impact of surface aerosol on geochemical, ecological, and radiation-climatic processes increases. The mineral particles, emitted particularly due to dust storms, comprise the organic matter and nutrients and may be transported horizontally over a long distance. This leads to land degradation because of structure and compositional changes of the soils both at dust sources and at dust receptors and to changes in biogeochemical cycles over land and ocean [4][5][6][7][8][9], to loss to public utilities because of reduced visibility [10][11][12] and air pollution [4,5,8,9,13,14], and even to global environmental [15] and climate changes because of aerosol direct influence on radiation fluxes [4,8,13,16], cloud formation [12,17], vegetation productivity [7,9], glacier and snow cover albedo [17], and atmospheric electricity [11]. berg et al. [44] and airborne number concentration measurements of Gorchakov et al. [45] showed a larger fraction of fine particles than if it was generated by saltation only. Measurements in deserts clearly confirm the presence of mineral dust in the ABL under windless and low-wind conditions not sufficient enough for initiation of saltation mechanism [46,47]. Khalfallah et al. [43] in their study found an empirical correlation between this PSD and the thermal structure of the surface ABL. The mass concentrations of fine aerosol obtained for fluid dynamic parameters in the viscous thermal boundary layer near the ground surface are related to a temperature drop in the thermal layer at the surface and from the values of friction velocity-so called non-saltation emission [48]. For sufficiently coarse particles, the Convective Turbulent Dust Emission model [35,36,49] is widespread. For fine particles (PM 1.5 /PM 2.0 ), authors were successful in using non-saltation model [48,50]. The extensive investigation of PSD in this particular area was performed in [51].
Long-term series of dust mainly comes from the ground visibility observation at meteorological stations and used for research on dust storms and spatio-temporal distribution characteristics of entrained aerosols [52][53][54]. However, data of fixed meteorological station alone is often not enough for mathematical modelling of dust emission; moreover, there are many areas without meteorological stations at all. In this regard, the annual expeditions to arid and semi-arid zones of the Caspian Lowland Desert may be useful for progress in that field. Dust storms are a serious environmental problem in arid and semi-arid regions of the Caspian Lowland. This region is easily eroded due to its underlying surface (common in arid zones) and meteorological features with typical wind and temperature gradient strong enough to entrainment the particles. Such experimental data sets are also essential for validation against various numerical simulation findings performed in the frame of this particular region.
The number of large-scale field experiments have been recently carried out (in [55] for Russia; in [35] the examples are ADEC [56], and ACE-Asia [57,58]), and the networks for obtaining aerosol (including dust) characteristics over large areas have been constructed. For example, among the latter is the AErosol RObotic NETwork (AERONET)-a federation of ground-based remote sensing aerosol networks (including over 150 stations), that evaluate aerosol optical properties (spectral aerosol optical depth, single scattering albedo, Angstrom parameter, aerosol scattering function, particle size distribution and others) in atmospheric column and validates satellite data retrievals of similar aerosol parameters.
In this paper, the direct observations of mineral dust aerosol within the range from 0.15 to 15 µm were carried out on arid and semi-arid zones of Kalmykia (Caspian Lowland, South of Russia) during summer 12 expeditions in 2002-2020 under different wind conditions with and without saltation process. The campaigns were carried out mostly during the period of maximum temperature and humidity. The description of the experimental site and the details of the instrumentation and methods are given in the following 'Materials and methods' section.
The aerosol optical depth (AOD) time series retrieved from the data of the EUMETSAT-CMSAF product based on the SEVIRI instrument on the METEOSAT satellite [59] (see Section 2.2.1) are analyzed. We use field campaigns' data in the context of long-term changes from satellite data and from meteorological data to determine the campaigns' date and duration, and to investigate whether aerosol characteristics are typical or exceptional for the region of interest. Accordingly, the campaigns were carried out mostly during the period of maximum temperature and humidity.
The complex and detailed results of expeditional data are discussed in the "Results and Discussion" Section. The study of dust emission including the analysis of the aerosol mass concentration and the aerosol flux, its chemical composition under realistic wind conditions is described in Sections 3.1 and 3.2.
The fine aerosol contributes greatly to the total aerosol content of the atmospheric surface layer. The surface air layer over the arid region of Kalmykia is characterized an increased aerosol emission during the observation period that may contribute to the longrange aerosol transport from the ABL over the arid territories to remoted regions. Fine particles suspended in air with the chemical composition and concentration related to those described in Sections 3.1 and 3.2 can be transported over long paths and form the aerosol layers which greatly influenced to the radiation and energy balance. Consequently, it is of the utmost importance to study the probability of such a transport of air particles (potentially polluted with dust), that reflects the potential transport of atmospheric pollutants. It is discussed in Section 3.3.
The conclusions are given in the Section 4.

Month June July August
The site, described in more details in [48,60], is located in the arid region. A map of the experimental field observation stations in Kalmykia is shown in Figure 1 (see also [61] for detailed map).
Description of the underlying surface (landscapes of Kalmykia and the Black Lands): the mosaic of the landscape cover includes dry-steppe landscapes (vegetation: Artemisia alba, Artemisia austriaca, a projective cover of 60-80%) with Solonchaks and very dynamic anthropogenic sand massifs, which in dry years occupy up to a quarter of the territory of Kalmykia.

Measurements
It is known, that there are difficulties in measuring emission-dust PSD, and it approximated using airborne-dust PSD measured at some, often different, heights [39,62]. In our study, we use the "gradient method" [43], which was designed by Gillette et al. [63] with the use of an Optical Particle Counter for measuring aerosol flux.
The required instrumentation deployed during the field experiments and methods are described in this paragraph. Description of the underlying surface (landscapes of Kalmykia and the Black Lands): the mosaic of the landscape cover includes dry-steppe landscapes (vegetation: Artemisia alba, Artemisia austriaca, a projective cover of 60-80%) with Solonchaks and very dynamic anthropogenic sand massifs, which in dry years occupy up to a quarter of the territory of Kalmykia.

Measurements
It is known, that there are difficulties in measuring emission-dust PSD, and it approximated using airborne-dust PSD measured at some, often different, heights [39,62]. In our study, we use the "gradient method" [43], which was designed by Gillette et al. [63] with the use of an Optical Particle Counter for measuring aerosol flux.
The required instrumentation deployed during the field experiments and methods are described in this paragraph.

Weather Conditions
Wind velocity, air temperature, pressure and humidity were measured continuously and simultaneously at levels of 0.2, 0.5, 1.0, 2.0, 3.0, and 5.0 m. Additionally, the five sensors of humidity and temperature of the surface were placed in the vicinity of the base of aerosol particle counters within the distance of 1-2 m [48]. Wind direction was measured with a sonic anemometer, fluctuations in three wind-velocity components and air temperature were measured with an USA-1 (METEK). The measuring instrument was constructed on the basis of AANDERAA Data Instruments' sensors. The east wind orientation in amplitude is predominant over the observation period ( Figure 2).

Weather Conditions
Wind velocity, air temperature, pressure and humidity were measured continuously and simultaneously at levels of 0.2, 0.5, 1.0, 2.0, 3.0, and 5.0 m. Additionally, the five sensors of humidity and temperature of the surface were placed in the vicinity of the base of aerosol particle counters within the distance of 1-2 m [48]. Wind direction was measured with a sonic anemometer, fluctuations in three wind-velocity components and air temperature were measured with an USA-1 (METEK). The measuring instrument was constructed on the basis of AANDERAA Data Instruments' sensors. The east wind orientation in amplitude is predominant over the observation period ( Figure 2).
With a general low variability of the monthly amount of precipitation, the simultaneous extension of both rainy and rainless periods, with a general increase in temperature ( Figure 3), can contribute to the formation of droughts (in rainless periods), leading to more probable dust storms. However, according to the station data [64], this increase is insignificant. Only the linear trend of air temperature is well pronounced: the growth in temperature linked to global warming points the increasing trend of about 0.5 • C/decade since 1960 to 2020 ( Figure 3).
The campaigns were carried out mostly during the summer period when the maximum in the temperature and humidity occur ( Figure 4).
We also used satellite data to estimate whether date of field campaigns and hence aerosol characteristics were typical or exceptional for the region of interest. In particular, we analyzed the daily variability of the aerosol optical thickness (depth) (AOD) in the arid region of the south of the European territory of Russia (45-48 • N, 43-47 • E) using the CMSAF product for AOD (at a wavelength of 550 nm). These products are based on the SEVIRI instrument installed on the METEOSAT geostationary satellite (till 2012 year). With a general low variability of the monthly amount of precipitation, the simultaneous extension of both rainy and rainless periods, with a general increase in temperature ( Figure 3), can contribute to the formation of droughts (in rainless periods), leading to more probable dust storms. However, according to the station data [64], this increase is insignificant. Only the linear trend of air temperature is well pronounced: the growth in temperature linked to global warming points the increasing trend of about 0.5 °C/decade since 1960 to 2020 ( Figure 3).
The campaigns were carried out mostly during the summer period when the maximum in the temperature and humidity occur ( Figure 4).   We also used satellite data to estimate whether date of field campaigns and hence aerosol characteristics were typical or exceptional for the region of interest. In particular, we analyzed the daily variability of the aerosol optical thickness (depth) (AOD) in the arid region of the south of the European territory of Russia (45-48° N, 43-47° E) using the CMSAF product for AOD (at a wavelength of 550 nm). These products are based on the    Interannual variability of AOD ( Figure 5) shows that periods of field campaigns-at least those that were conducted before 2012-in general, were typical for this time of the year, except for the 2009 campaign that was characterized by decreased AOD values. It is of note, that high AOD values are typical for May-July for the analyzed region ( Figure 6) with the absolute maximum AOD observed in the second half of June. The average AOD for the region is 0.25-0.3, and the values over 0.95 quantile reach 0.5.  The black thick line is the regional daily averaged AOD. The yellow line is the regional monthly averaged AOD.

Dust Number Concentration
The number (counting) concentration and size distribution of aerosol particles were measured in the daytime (usually from 09:00 to 19:00) at two levels: 0.   The black thick line is the regional daily averaged AOD. The yellow line is the regional monthly averaged AOD.

Dust Number Concentration
The number (counting) concentration and size distribution of aerosol particles were measured in the daytime (usually from 09:00 to 19:00) at two levels: 0.  The black thick line is the regional daily averaged AOD. The yellow line is the regional monthly averaged AOD.

Dust Number Concentration
The number (counting) concentration and size distribution of aerosol particles were measured in the daytime (usually from 09:00 to 19:00) at two levels: 0. up to 2.103 particles per cm 3 . The 10-channels optical-electric aerosol spectrometer OEAS-05 measured dispersed composition, number concentration in the aerosol particle size (diameter) range 0.3-6.5 µm at a number concentration of up to 103 particles per cm 3 . Based on the gravimetric method using filters installed at a certain particle size (cascade in 6-stage impactor), we can estimate the mass concentration of particles of such a size; and subsequently, after making a chemical analysis of the filter, evaluate aerosol composition of this size.
An aspiration sampler (air flow rate of 12 m 3 h −1 ) with a set of filters APhA-KhA-20 was used for evaluating chemical composition and mass concentration of aerosols.
Determination of the disperse composition and number concentration of aerosol particles has been performed by laser spectroscopy. Determination of the mass concentration of particles has been received by the gravimetric method.
To determine the aerosol mass concentration, samples were deposited on analytical filters of the KhA-20 type at the height of 1.5-2.0 m from the underlying surface using an aspiration sampler with a sampling time of at least 12 h (day or night). The mass concentration was determined discretely by the weight method. This method include capturing aerosol on the filter from a certain volume of air, determining the filter weight gain and calculating the mass concentration value in µgm −3 . (6-stage cascade impactor was used simultaneously with sampling for filters to determine the mass concentration of aerosol particles).
The measurement of meteorological parameters and turbulence parameters was carried out simultaneously with the determination of aerosol parameters.
A more detailed description of the instruments and the instrument's working mechanisms is given in the [48,51].
The chemical composition of aerosols and soils was determined by the method of atomic emission spectrometry with inductively coupled plasma (ICP-AES).

Dust Flux Computation
The calculation of the vertical dust flux based upon the gradient method requires to measure and evaluate the dynamical parameters characterizing the structure of the surface boundary layer (the wind friction velocity, surface roughness length, Monin-Obukhov's length [66]) and the dust concentrations at two heights (instrumentation is described at Section 2.2.2).
Experimental data and theoretical estimates show that particle detachment from the underlying surface can be associated with turbulent stresses created by wind shear in the surface boundary layer-saltation mechanism. However, a large amount of fine particles exists in the ABL under weak wind conditions. Such particles can be lifted in the absence of saltation, when the wind force is insufficient to form strong shear turbulence over the underlying surface. The aerosol entrainment is related to the strong convection of the air over the sand layer during hot weather, where temperature drops in the thin surface air layer cause the dust resuspension-non-saltation mechanism. Detailed description of the physics of non-saltation emission processes is given in [48,60]. The emission occurs when the friction velocity reaches a corresponding for each mechanisms critical value.
The friction velocity was calculated according to the Monin-Obukhov theory [66]  The wind friction velocity, surface roughness length, and Monin-Obukhov's length have been determined using the measured wind velocity and temperature profiles. Aerosol flux has been calculated using [33,63], assuming neutral conditions and that the vertical transport of particles with a diameter ≤20 µm is similar to that of momentum: where k is the von Karman's constant (k = 0.4), u * -friction velocity and C l and C h are the dust mass concentrations measured in the surface boundary layer at a low (z l ) and high (z h ) altitudes (0.5 and 1.5 (or 2.0, it is dependent on the year of measurements) m). Heat fluxes were calculated, see, for example, [67] using: where k is the von Karman's constant (k = 0.4), L MO -Monin-Obukhov's length, gacceleration of gravity, θ-air temperature, ρ-air density, and C p -specific heat at constant pressure.
For the PM 20 emission three mechanisms are considered as being responsible: direct aerodynamic entrainment [68], the combination of saltation and sandblasting, and nonsaltation emission [48]. These particles, especially fine mode of PM20, have a residence time in the troposphere long enough to be transported over great distances and to form the aerosol layers [35,38,69,70]. During their stay in the ABL, they can impact the Earth's climate [71][72][73][74] and human health [15,[75][76][77]. The detailed description of the investigating methods for PSD's study in Kalmykia was given in [51].

Trajectory Analysis
The backward trajectories of air masses were calculated using the NOAA HYSPLIT model [78]. The evaluating of transport probability fields in the context of its long-term variability was through the technique explained in [79].
To study the transport probability the statistics of direct air particle trajectories was used. In [55], estimates of the probability of air transport from Kalmykia to the Moscow region from the surface layer of 0.1 and at an altitude of 1 km in 1986-2018 were calculated. In this work, the change in the probability of transport in recent decades are estimated.

Forward Trajectory Modelling
Five-day forward trajectories of air particles were calculated using the NOAA HYS-PLIT_4 trajectory model [78] and the gridded (horizontal resolution 2.5 • × 2.5 • , 17 levels in height, time resolution 3 h) meteorological achieve NCEP/NCAR Reanalysis [80,81] covering the period from 1949 to the present. The trajectories were calculated for 6 times (UTC 01:00, 07:00 and so on) of every day started from the ABL above unnamed point (45.3 N, 46.1 E; south of Kalmykia) simultaneously at three heights H/3, 2 × H/3 and H, where H is the mixed layer depth that is practically equal to ABL height.

Air Transport Probability Reconstruction
Probability of air transport was calculated for periods of 1991-2018, 1991-2004 and 2005-2018 using the methodic described in [55]. Then fields of change of air transport probabilities in 2005-2018 relative to 1991-2004 (i.e., for the last 15-30 years) were calculated. The fields of the probability and its change (spatial resolution of 2 • × 2 • ) were drown using GIS-Softwate MeteoInfo [82].        In 2009, 2013, the convective regime of emission prevailed, in 2011 the wind regime dominated. This is seen from the magnitude of fluxes and from the dependence of the aerosol mass concentration on the dynamic friction velocity in Figures 9 and 10. The variations in aerosol mass concentration due to variation in friction velocity indicate the mechanism of emission. The power dependence between these two characteristics represents the strong influence of wind to the entrainment process and, respectively, saltation mechanism of aerosol emission.

Mass Concentration and Aerosol Flux
sents the strong influence of wind to the entrainment process and, respectively, saltation mechanism of aerosol emission.
The combination of regimes was noted in all years. Nevertheless, wind entrainment of particles, saltation mechanism prevailed for 2010 and 2011, and convectional entrainment of particles, non-saltation mechanism prevailed for 2009, 2013 and 2014. The characteristic feature of the non-saltation regime is its connection with heat fluxes. Consequently, the dependence of the aerosol flux on sensible heat flux is almost linear for the non-saltation regime. The heat flux depends on saltation, but, in general, the wind velocity is weakly related to the heat flux, so the aerosol flux mostly grows with the increase of sensible heat flux. Figure 11 represent the variance of the aerosol flux depending on the sensible heat flux. 8

Aerosol Composition
Apart from mass concentrations and fluxes, aerosol composition is also of great importance. The combination of regimes was noted in all years. Nevertheless, wind entrainment of particles, saltation mechanism prevailed for 2010 and 2011, and convectional entrainment of particles, non-saltation mechanism prevailed for 2009, 2013 and 2014. The characteristic feature of the non-saltation regime is its connection with heat fluxes. Consequently, the dependence of the aerosol flux on sensible heat flux is almost linear for the non-saltation regime. The heat flux depends on saltation, but, in general, the wind velocity is weakly related to the heat flux, so the aerosol flux mostly grows with the increase of sensible heat flux. Figure 11 represent the variance of the aerosol flux depending on the sensible heat flux.

Aerosol Composition
Apart from mass concentrations and fluxes, aerosol composition is also of great importance.  Table 4 [61]. The elemental composition of aerosols is closest to the composition of fine and medium dust (particles 1-10 μm in size (diameter)) of sandy soil (correlation coefficient R = 0.73), differing from the composition of soils of dry steppe (R = 0.69) and Solonchaks (R = 0.50).
Having studied the chemical composition of sandy soils and aerosol particles in the  Table 4 [61].  Table 4 [61]. The elemental composition of aerosols is closest to the composition of fine and medium dust (particles 1-10 μm in size (diameter)) of sandy soil (correlation coefficient R = 0.73), differing from the composition of soils of dry steppe (R = 0.69) and Solonchaks (R = 0.50). The mass content in the surface arid aerosol of most of the elements of global distribution, which have a mineral nature (Al, Si, Na, Mg, Ca), strongly depends on meteorological conditions. This content increases in windy conditions, during dust storms due to saltation emission of particles from the underlying surface [61]. The saltation is the main factor that determines the revealed interannual variability of the concentration of these elements in the surface aerosol during the period under consideration.
Phosphorus and sulfur, with its mixed (anthropogenic-lithogenic) nature, are classified as elements of global importance due to their wide distribution and constantly increased concentration. The interannual variability of their concentration in aerosol particles of Kalmykia can be associated with the particularity of local anthropogenic sources, such as activities of enterprises for the development of oil-and-gas and oil-and-gas-condensate fields. In addition, the increased sulfur content in the deserted areas of Kalmykia may be associated with the entrainment of particles from the surface of the Solonchaks, typical soil cover of the deserted regions of the Caspian Lowland and Central Asia.
Most of the considered elements of regional and local significance are of a mixed nature: either predominantly lithogenic or predominantly anthropogenic. More detailed For example, local soils are enriched with arsenic and heavy metals such as Hg, Co, Cr, Cu, Zn, Mo and Cd [61]. These trace elements enter the atmosphere due to the emission from the underlying surface. However, oil-and-gas enterprises and motor transport also contribute to the total content of these elements in aerosols of the observation region. The main source of elements such as Sn, Pb, Ni, Hf, W is anthropogenic activity and regional transport from adjacent territories. Therefore, the interannual variability of the concentration of elements of regional and local significance is strongly dependent on the specifics of local and regional anthropogenic sources.
The groups of elements according to their variability from 2007 to 2016 were described in [61]. Chemical composition of soils of major landscapes in the Chernozemel'skii district of Kalmykia were given in Table 3 in [61]. The highlighted elements are characterized by significant values of the percentage mass content and linked to industrial activities. Table 5 in [61] show the coefficient of aerosol concentration and the coefficient of local aerosol concentration in soil and surface atmospheric particles in background landscapes of Kalmykia related to the enrichment factor by mass of the element in aerosol (aerosol). The calculated aerosol-concentration coefficients (CC a = aerosol/earth crust) and local aerosol-concentration coefficients (CC as = aerosol/soil) are the degree of enrichment of soil landscapes and surface aerosols in semiarid regions of observed area. Accumulation of elements in surface aerosol particles classified according to the natural and naturalanthropogenic landscapes of Kalmykia is presented in Table 6 in [61].
The elemental composition of aerosols is closest to the composition of fine and medium dust (particles 1-10 µm in size (diameter)) of sandy soil (correlation coefficient R = 0.73), differing from the composition of soils of dry steppe (R = 0.69) and Solonchaks (R = 0.50).
Having studied the chemical composition of sandy soils and aerosol particles in the surface layer of arid zones of Kalmykia, we have revealed that the predominant fraction in the soil is the fraction of fine sand (86.6%), significantly exceeding the percentage of fractions of silt, medium and coarse sand. The chemical composition of the sand sample is aluminum-silicon. The amount of silicon oxide exceeds 80%, aluminum-10%. The amount of iron oxide is only 2.5%, calcium-2.2%. Alkaline earth elements are −1.5% calcium oxide and 1.8% magnesium oxide. The predominant fraction for the main microelements of aerosol particles is the fraction of fine sand (1-5 µm), the fraction of sludge by percentage is several times less, and large particles (5-10 µm) are the smallest values, with the exception of lead. The results of elemental and mineralogical analysis of soil and aerosol particles confirmed the identity of the chemical composition of the highly dispersed fraction of soil and aerosol.

Aerosol Transport
The surface air layer, as well as the ABL, over the arid region of Kalmykia is characterized in the warm season by an increased aerosol load [79]; therefore, it is of interest to study the probability of long-range transport of air particles (potentially polluted with dust) from the ABL over the arid territories to remoted regions. All other things being equal, this probability reflects the potential transport of atmospheric pollutants (including dust aerosol) from the arid region.
Fine particles suspended in air can be transported over long paths and form the aerosol layers which greatly contributed to the radiation and energy balance. Characteristic aerosol trajectories of different regions requires thorough analysis.
The 3-day backward trajectories of air masses, calculated using the NOAA HYSPLIT model [78] for the observation period in Kalmykia from 21 July to 31 July 2020, are shown in the Figure 14.
To study mean climatic characteristics in the context of long-term variability, the transport probability fields were reconstructed using array of about 300,000 modeled forward trajectories. The fields characterize the potential aerosol transport in the lower troposphere from the observation site to Russian regions and to neighbor countries.
Only those cells are shown that crossed with at least 150 trajectories. Seasonal average probability of air particle transport may illustrate the probability of transport in different seasons of the year, as well as the seasonal areas of the transport. Air transport from the boundary layer over the deserted areas of Kalmykia in all seasons is most likely in the western, southwestern and, slightly less probably, in the southeastern directions from the areas. Southwestern regions of Russia (Stavropol, Rostov, Astrakhan, Krasnodar, Volgograd oblasts and the Republic of Crimea) are most susceptible to potential aerosol transport from the arid region of Kalmykia. Figure 15 reflects the changes in potential dust transport probability in the modern period (last 15-25 years). Over the period in all seasons, the transport probability to the central regions of the European part of Russia (Tatarstan and Samara and Volgograd Oblasts and others), and in the summer also to the Rostov Oblast and Crimea, has increased.
Variations in the average seasonal probability of air particles transport from the ABL over the deserted lands of Kalmykia to Moscow are given at Figure 16. It is seen the probability is decreased or stable in all seasons excluding winter when the probability is doubled for the last 25-30 years. The linear trend values (and their statistical significance) from Figure 16 are given in Table 3. In winter, spring and summer over the past 50-60 years there has been a decrease (statistically significant only in summer and spring) of the average seasonal air transport probability; in autumn, there are no statistically significant trends.
Until 1990 there was a decrease in the probability in all seasons, the largest and statistically significant one in summer. After 1990 there was a statistically significant positive trend in winter; in other seasons, there is no significant transport trend in the lower troposphere.
osol layers which greatly contributed to the radiation and energy balance. Characteristic aerosol trajectories of different regions requires thorough analysis.
The 3-day backward trajectories of air masses, calculated using the NOAA HYSPLIT model [78] for the observation period in Kalmykia from July 21 to July 31, 2020, are shown in the Figure 14.  To study mean climatic characteristics in the context of long-term variability, the transport probability fields were reconstructed using array of about 300,000 modeled forward trajectories. The fields characterize the potential aerosol transport in the lower troposphere from the observation site to Russian regions and to neighbor countries.
Only those cells are shown that crossed with at least 150 trajectories. Seasonal average probability of air particle transport may illustrate the probability of transport in different seasons of the year, as well as the seasonal areas of the transport. Air transport from the

Conclusions
As it mentioned in [35], the recent achievements in dust emission research are substantial, but to improve quantitative emission modelling and prediction we should focus on accurate determination of the key parameters which affect emission. Some fundamental aspects of the physics of emission are not completely understood. Consequently, dustemission schemes contain parameters which must be determined empirically. However, there is a lack of high-quality observational data sets for the thorough validations of dustemission schemes; therefore, large uncertainties may exist in scheme parameters and new field experiments need to be carried.
In spite of their importance for the prediction and simulation of the entrainment, transport and deposition of mineral dust, "in-situ" measurements of the dust emission flux and PSD under realistic wind conditions remain rare, especially for the former territory of the USSR.
In this paper the spatial and temporal trends of aerosol emissions, its chemical composition and transport characteristics for arid and semi-arid areas in the Caspian Lowland Desert in the south of Russia (Kalmykia) were performed based on field sample collections, satellite images and direct measurements in the site of investigation.
Long-term series of data on the mass and number concentration, dispersed and elemental composition of aerosol in the surface layer of the atmosphere of deserted areas of the steppe zone of Kalmykia were obtained. The auxiliary data of meteorological parameters (air temperature and humidity, wind velocity and direction, pressure, etc.) were provided.
The characteristic values of the mass and number concentration of aerosol particles in the semi-desert zones of Kalmykia in the summer period are determined for the observation period 2002-2020. In particular, the mass concentration in different years of observation ranged from several tens to several hundred µg m −3 . It was confirmed that for these field campaigns the aerosol emission arose from two different mechanisms: saltation (associated with turbulent stresses created by wind shear in the surface boundary layer) and nonsaltation (related to the strong convection of the air over the sand layer during hot weather, where temperature drops in the thin surface air layer cause the dust entrainment). The latter mechanism was predominant in Kalmykia in summer period. It is noted from the data received that a large number of fine particles exists in the ABL under weak wind conditions. An identification of possible aerosol sources of natural and anthropogenic origin as well as trajectory analysis of air mass transport from and to studied arid region were performed. An increased content of anthropogenic elements (S, Sn, Pb, Bi, Mo, Ag, Cd, Hg, etc.), which is characteristic of all observation points in the south of the European part of Russia, was found. The elements were revealed mainly in submicron and coarse aerosol and enter the studied regions with air masses from industrial and/or highly populated territories due to long-range air transport.
Analysis of 5-day forward trajectories of air particles for 1949-2018 showed that air transport from the boundary layer over the deserted areas of Kalmykia in all seasons is most likely in the western, southwestern and, slightly less probably, in the southeastern directions from the regions. Southwestern regions of Russia are most probable recipients of the dust emitted from deserts of Kalmykia. At the same time, over the past three decades, compared to three decades before the start of modern global warming, the probability of the air transport to these regions of the Russia has decreased, while for a number of central regions of the European territory of Russia (ETR) and some regions of the Asian territory of Russia (ATR) has increased. Thus, the dust emitted from the deserted lands of Kalmykia in recent decades has become an increasingly significant climatic and ecological factor both for most of the ETR and for a number of western Siberia regions of Russia.
A preliminary analysis of submicron aerosol fluxes does not clearly reveal an increase or decrease trend; however, there is a relationship with general synoptic factors of a general annual scale. Despite the trends towards an increase in temperature and humidity, there are no stable emission trends towards an enhance; but the situation may worsen: for example, the hot dry summer of 2020 led to an intensification of dust emissions and the formation of a severe dust storm event in autumn 2020 (that reached even the Atlantic) [83].