Introduction

The arid and semi-arid regions in mainland China is considered to be one of the worst emission source areas of atmospheric particulate matter on Earth1,2. Especially in the coal resource-based cities, mining operations can bring economic benefits to the local region but also cause serious environmental pollution3,4. Atmospheric dustfall is the part of atmospheric particulate matter which can settle naturally5,6,7, it is considered to be an important indicator for evaluating atmospheric pollution, atmospheric dustfall not only affects climate and atmospheric visibility but also causes serious harm to human health, leading directly or indirectly to socio-economic losses8,9,10,11,12.

Particle size and morphology are typical microscopic characteristics of atmospheric dustfall, which can directly reflect the source and cause information13,14. Particle size has been studied extensively and is an essential part of microscopic characteristics. Morphology of atmospheric dustfall has been mainly based on the subjective description using the results of observed images15. Shape factor is used to quantitatively characterize the shape of particulate matter, which is mainly used in studies of soil, rock, tailings, and atmospheric fine particles etc16,17,18,19,20, but it is rarely used in the study of atmospheric dustfall. So using the shape factors to statistically analyze the morphological characteristic of atmospheric dustfall, can reflect objectively results. Moreover, these characteristics were normally studied separately. There is a lack of research on the correlation between the morphology and particle size distribution of atmospheric dustfall, so the correlation between these two microscopic characteristics is worth exploring.

The source analysis of atmospheric dustfall mainly utilizes receptor model methods, which are classified into chemical method and microscopic method21,22,23. Chemical method can be quantitatively analyzed pollution sources by analyzing their source spectra and contribution rates23. Microscopic method is used for the qualitative and semi-quantitative analyses of pollutants24,25. Compared with chemical method, microscopic method has certain subjectivity, and it does not allow for exact quantitative analysis. However, using microscopic method in conjunction with the current state of the study area, specific dust sources can be qualitatively analyzed.

Wuhai, an open-pit mining coal resource-based city, is located in the arid desert area of Northwest China, where has an extremely fragile ecosystem26,27. Previous researchs have investigated the temporospatial distribution of dust concentration and the characteristics of heavy metal pollution of atmospheric dustfall, as well as using principal component analysis to analyze the pollution sources and their contribution28,29. However, there have been no studies on microscopic characteristics of atmospheric dustfall and specific dust source analysis in Wuhai. In this study, atmospheric dustfall samples were collected during the windy period in 2018. The sizes, shapes, and elemental compositions of single particles were determined by SEM–EDS, and their mineral compositions were determined by XRD. The characteristics of size, shape, and variation of shape factors with size distribution were analyzed. The types of individual particles were determined based on the elemental and mineral composition, and the specific dust sources of atmospheric dustfall were then analyzed in relation to the characteristics of five functional areas. The results of this study can verify the correlation between microscopic characteristics of atmospheric dustfall and provide new ideas for the study of correlation among atmospheric particulate matter characteristics, and using shape factors can more objectively reflect the morphological characteristic. The identification of specific dust sources can improve the analysis of the source of atmospheric dustfall, provide theoretical support for conducting accurate dust control and a scientific basis for air pollution control measures in open-pit mining coal resource-based cities in the arid desert area of Northwest China.

Materials and methods

Study area

Wuhai City is located in the arid desert area of Northwest China and the southwest of the Inner Mongolian Autonomous Region of China (Fig. 1, 106°52′50′′–106°54′17′′E, 39°42′03′′–39°42′36′′N). Since 2006, the coal mining method in this city has gradually changed from underground mining to open-pit mining. The study area is a continental monsoon climate area. The annual average wind speed is 2.9 m/s, the annual instantaneous maximum wind speed is 33 m/s, the annual average relative humidity is 42%, the annual average precipitation is 159.8 mm, and the annual evaporation is 3289 mm. Annual aeolian sand days are more than 80 days, and the days of strong wind are mainly concentrated in the windy period (March–May). The study period is the windy period (March–May) of 2018 with an average maximum wind speed of 7.15 m/s. Aeolian sand days occurred sixteen times, and the prevailing wind direction was northwest.

Figure 1
figure 1

Image of sampling sites of atmospheric dustfall, and the wind regime.

Full size image

In this area, soil types can be mainly divided into six types, gray desert soil, brown soil, chestnut soil, sandy soil, meadow soil, and saline soil, among which gray desert soil and brown calcium soil are the main components. The vegetation type is mostly desert steppe vegetation, mainly herbs, shrubs, and small trees. The vegetation coverage of Wuhai City is 30.26% (including crops on agricultural land and green plants on construction land) during the study period.

The study area was divided into five functional areas, open-pit mining area, off-mine coal road, industrial area, commercial and residential area, and desert area (Table 1).

Table 1 Introduction of the five functional areas.
Full size table

Sample collection

Samples were collected using deposition cylinders (cylindrical plexiglass cylinders) with reference to the Chinese national standard “GB/T15265-1994”, stating the following layout principles. To avoid damage to or replacement of the dust deposition cylinders by humans, the cylinders are usually placed on the top floor of low-level buildings or on electric poles; they should be located far away from tall buildings to avoid disturbance by unnatural factors; the cylinders should be placed at the same height, which was set at a height of 2 m above ground, to avoid the impact of dust from platforms.

In this study, the dust deposition cylinders were fixed on electric poles in open and flat areas without tall buildings. After the samples were collected, debris such as branches and leaves were removed, and the samples were stored in dry and sealed hermetic bags. Samples were collected from March 1, 2018, to May 31, 2018.

A total of 37 sampling points were set up in five functional areas (Fig. 1), including eight sampling points in the open-pit mining area, eight sampling points on the off-mine coal road, six sampling points in the industrial area, eleven sampling points in the commercial and residential area, and four sampling points in the desert area.

Samples analysis

Scanning electron microscopy coupled energy dispersive spectrometer

Samples of each functional area were evenly mixed, and the resulting five mixed samples were analyzed as representatives of atmospheric dustfall in the five functional areas. Scanning electron microscopy coupled with an energy dispersive spectrometer (SEM–EDS) (Hitachi-SU8010, Japan) was used for the determination of particle size, shape observations, and element analyses of single particles.

For each functional area, taking an appropriate amount of each sample and evenly dispersing it onto 5 conductive adhesives (conductive adhesive size: 2 × 2 mm) pasted on the aluminum mount, and a very thin film of gold was deposited on the sample surface before testing. The FESEM images of particles were acquired at magnifications of 200, 500, 800, 1200, and 2000 by using the five-point sampling method. Ten images were acquired for each sample with a total number of particles between 900 and 1300 per sample. The SEM is used to select the single particles, and the equipped EDS is used to scan them, so that the energy spectra of the main elements and the proportion of the content of individual elements can be obtained.

The SEM–EDS system was equipped with a cold field-emission electron gun, operated under vacuum with a voltage of 3 kV and a current of 10 μA; the EDS was set at an accelerating voltage of 15 kV, a fixed angle of 35°, and an elemental analysis range of 4Be–92U.

Image-Pro Plus 6.0 software can determine the number of particles and various particle-parameters in SEM images, applicable to the research of correlation between particle size distribution and shape factors of atmospheric dustfall in this study. So after collecting the FESEM images, importing them into this software for particle size analysis and shape factor parameter acquisition.

X-ray diffraction

XRD is a non-destructive technique used to identify mineral phases and is well established in the determination of the mineral compositions of powders. This method is the most commonly applied technique for determining the mineral content of dust. It can be used for qualitative and semi-quantitative analysis of mineral compositions, but it is not possible to accurately quantify them, so in this study it was only used to qualitatively analyze.

In this study, atmospheric dustfall was examined using the XRD instrument (MiniFlex 600, Rigaku, Japan). The samples were thoroughly ground using a mortar and pestle, and then were pasted onto slides and placed on an XRD sample stage to measure with Cu Kα radiation (40 kV and 15 mA) in the 2θ range of 3° to 50° with a scanning rate of 0.05°/min and a step size of 0.01° in order to get the XRD spectra. Then compared the XRD spectra with the standard card data by Jade 6.5 to determine the mineral phases.

Calculation methods

Graphical method

In this study, a graphical method was chosen to analyze the particle size of atmospheric dustfall. The characteristic parameters of the grain size include the mean diameter (MZ), standard deviation (S), skewness (Sk) and kurtosis (KG), which were calculated by using the Folk-Ward formula30 (Table 2), where D5, D16, D25, D50, D75, D84, and D95 indicate the particle size corresponding to 5%, 16%, 25%, 50%, 75%, 84%, and 95% of the cumulative percentage, respectively.

Table 2 Formulas of grain-size parameters.
Full size table

Shape factors

FESEM images were analyzed with Image-Pro Plus 6.0 software to obtain primary and secondary parameters of the particles. Roundness was directly derived from the software, and Aspect ratio, Convexity and roughness were calculated by the respective formulas. The shape factors chosen in this study were two-dimensional parameters.

Roundness (R) is defined as the similarity of a particle to a circle,

$$ {\text{R}} = {\text{P}}^{2} /4{\pi A} $$
(1)

where P is the projected perimeter, and A is the projected area. The range of R is [1,∞]. A perfectly spherical particle shape has an R value of one.

Aspect ratio (AR) is defined as the length of a particle shape17,31,

$$ {\text{AR}} = {\text{F}}_{{{\text{min}}}} /{\text{F}}_{{{\text{max}}}} $$
(2)

where Fmax and Fmin are the maximum and minimum Feret diameters. The range of AR is [0,1]. When AR is close to one, the particle shape is close to a sphere or square. When A approaches zero, the particle shape becomes column- or flake-like.

Convexity (Con) is defined as the angularity of a particle shape17,31,

$$ {\text{Con}} = {\pi D}_{{{\text{Aeq}}}} /{\text{P}} $$
(3)

where DAeq is the particle equivalent diameter. The range of Con is [0,1]. When Con is close to one, the angles of the particles are sharp; otherwise, the angles are blunt.

Roughness (r) is defined as the undulation of a particle shape32,

$$ {\text{r}} = ({\text{P}}/{\text{P}}_{{\text{C}}} )^{2} $$
(4)

where PC is the circumscribed polygon perimeter of the particle. The standard value of r is one, and the closer r is to one, the smoother is the particle surface, and conversely, the rougher is the surface.

Results and discussion

Particle size characteristics of atmospheric dustfall

In this study, the widely used Udden-Wentworth classification method was chosen for particle size classification33,34, and the particle size was converted to the Φ value by using the equation Φ = − log2D proposed by Krumbein35. The particles were classified as clay (Φ > 9), very fine silt (8–9Φ), fine silt (7–8Φ), medium silt (6–7Φ), coarse silt (5–6Φ), very coarse silt (4–5Φ), very fine sand (3–4Φ), fine sand (2–3Φ), medium sand (1–2Φ), and coarse sand (Φ < 1).

The results showed that the particle size distribution of atmospheric dustfall was in the Φ range of 0–8 in the open-pit mining area, the industrial area and on the off-mine coal road, was in the Φ range of 0–9 in the commercial and residential area, and was only in the Φ range of 0–5 in the desert area. Figure 2 presents the cumulative percentage of particle size of atmospheric dustfall in different functional areas. Considering a fraction of more than 10% the main distribution size, the particle sizes of atmospheric dustfall in the open-pit mining area and industrial area were mainly distributed in the ranges of 1–6Φ. For the off-mine coal road, the commercial and residential area, and the desert area, the main particle size distribution ranges were 1–5Φ.

Figure 2
figure 2

The cumulative percentage of particle size of atmospheric dustfall in each area (A—off-mine coal road; B—industrial area; C—commercial and residential area; D—desert area; E—open-pit mining area).

Full size image

The grain-size characteristic parameters of atmospheric dustfall determined in each functional area are presented in Fig. 3. The maximum MZ was found in the open-pit mining area, and the minimum MZ was in the desert area. The values of S ranged from 1.22 to 1.76 and can be classified as poorly sorted in each functional area, as highlighted in Table 3. The values of Sk ranged from − 0.04 to 0.17, and can be classified as positive skewed on the off-mine coal road and in the commercial and residential area, and symmetrical in the industrial area, the desert area, and the open-pit mining area. The values of KG ranged from 0.78 to 0.99, and can be classified as mesokurtic on the off-mine coal road, in the commercial and residential area, and the open-pit mining area, and platykurtic in the industrial area and the desert area.

Figure 3
figure 3

Grain-size parameters for atmospheric dustfall in each area (A—off-mine coal road; B—industrial area; C—commercial and residential area; D—desert area; E—open-pit mining area).

Full size image
Table 3 Grain-size parameter ranges and classifications.
Full size table

Combining the size distribution characteristics and particle size characteristics, the proportion of atmospheric dustfall with sizes larger than 31 μm was greater than 90% in each functional area. The atmospheric dustfall in all functional areas was dominated by coarse-grained particles, which the highest content was very coarse silt in the industrial area and the open-pit mining area and fine sand on other three functional areas. It has been previously reported that when the wind speed is greater than or equal to the threshold wind velocity, the transport distance was determined by the particle size, and different particle sizes can only be transported over long distances under different dynamic conditions. Particles greater than 10 μm less than 100 km, and particles exceeding 31 μm can generally be transported less than 1000 m36. Consequently, the atmospheric dustfall in each area was mainly near-source material, which was highly influenced by local sources.

Previous reports showed that the atmospheric dustfall is mainly composed of fine silt, coarse silt and fine sand in underground coal mining resource-based city in the arid desert area of Northwest China37, which is similar to the results of this study. In China’s economically developed cities, it is mainly composed of the particles smaller than very coarse silt38,39. It can be seen that the main particle size distribution range of atmospheric dustfall in coal resource-based cities is larger than that in economically developed cities. The industrial activities, coal mining activities, and spring wind activities have a significant effect on the particle size distribution of atmospheric dustfall in coal resource-based cities in the arid desert area of northwest China.

Shape factors of atmospheric dustfall

Shape factors are generally used to describe the degree of deviation of particulate matter from spherical particles. Since particle shapes are difficult to quantify, software-based image processing and morphological analysis is an effective approach for characterizing the shape and size distribution of particles. In this study, the shape information of individual particles was extracted from two-dimensional SEM images using Image-Pro Plus 6.0 software, and the shape characteristics of atmospheric dustfall in each functional area were obtained according to the shape factors.

Table 4 provides the mean values and standard deviations of the different shape factors of atmospheric dustfall in the selected five functional areas of Wuhai. The Roundness value in the desert area was closest to one with minimal variations, while being most deviated from one in the industrial area with maximal variations, the variation in each functional area was in the range of 0.04–0.10. Aspect ratio was closest to one in the desert area and deviated from one to the largest extent in the commercial and residential area, the variation in each functional area was in the range of 0.02–0.03, and the overall variation was small. Convexity was closest to one in the desert area and exhibited their greatest deviation from one in the industrial area. The variation in each functional area was between 0.01 to 0.03, and the overall variation was small. Regarding roughness, the mean value and overall distribution of atmospheric dustfall were similar in each functional area with no significant differences.

Table 4 Shape factor values of each area (Values = mean ± standard deviation).
Full size table

There was no significant difference in roughness of atmospheric dustfall among the five functional areas, indicating that this factor was less affected by the different functional areas. Except for the desert area, the differences in Roundness, Aspect ratio and Convexity of atmospheric dustfall were small in each functional area, which were found to be mainly influenced by the dominant wind direction and the proximity of each functional area. The particle shape of atmospheric dustfall was closest to spherical shape in the desert area and deviated most from spherical shape in the industrial area and the commercial and residential area, which indicated that the angles of particles were sharper in the areas influenced by human factors. In general, the shape of atmospheric dustfall deviated from spherical shape in each functional area, which indicated that the impact of wear by atmospheric transportation was small, and the dustfall was not transported over long distances but was mainly influenced by local sources.

The correlation between particle size distribution and shapes of atmospheric dustfall

Figure 4 depicts the variation of shape factors with size distribution in the five functional areas. Except for the desert area, the values of Roundness and roughness showed a W-shaped variation with increasing particle size (Φ), while the values of Aspect ratio and Convexity exhibited an M-shaped trend. With increasing Roundness values, roughness increased and Aspect ratio and Convexity decreased, indicating that a particle shape was closer to a circle and its surface was smoother. Except for the off-mine coal road, the maximum values of Roundness and roughness and the minimum values of Aspect ratio and Convexity of atmospheric dustfall in each functional area were at the maximum Φ value.

Figure 4
figure 4

Shape factor values of atmospheric dustfall for different size distribution ranges in each functional area.

Full size image

The variations of shape factors with size distribution in each functional area were similar. With increasing Φ, the difference in the shape factor values among the functional areas increased, indicating a greater influence of the regional difference with decreasing particle size. Previous studies did not report on the relationship between the shape variation of atmospheric dustfall and particle size in different functional areas. The results of this study showed that the difference in the shape of atmospheric dustfall in different functional areas was significantly correlated with the particle size distribution. So the microscopic characteristics do not exist independently but affect each other, the correlation between them needs to be further investigated.

Mineral composition of atmospheric dustfall

XRD data of the atmospheric dustfall from the five functional areas were acquired, and the minerals present in the samples were determined by comparing the characteristic XRD peaks with the RRUFF database of the reference standards. The compositions in each functional area were found to be similar with only partial differences shown in Fig. 5. The most obvious strong peaks in the XRD patterns of the atmospheric dustfall of each functional area in this study were attributed to quartz (a). Feldspar (b), calcite (c) and muscovite (e) were also detected in all functional areas. Dolomite (d) was detected in all functional areas except the desert area, gypsum (f) was detected in the commercial and residential area and the open-pit mining area, and kaolinite (g) was only detected in the open-pit mining area. In addition, non-crystalline particulate matter was not detected by XRD2. The low content of particulate matter was also one of the reasons why they could not be detected by this technique.

Figure 5
figure 5

XRD determination of the mineral composition of atmospheric dustfall from different areas (a—quartz; b—feldspars; c—calcite/CaCO3; d—dolomite; e—muscovite; f—gypsum; g—kaolinite).

Full size image

XRD is capable of analyzing crystalline mineral particles in the physical phase, effectively resolving the mineral composition of atmospheric particles, but is unable to identify non-mineral particles, whereas EDS is used to identify particle types by determining the elemental composition of particulate matter, and can identify non-mineral particles in atmospheric dustfall40. Therefore, the determination of the compositions of atmospheric dustfall by XRD could be complemented by EDS to identify other particle types that were not detected by XRD.

Elemental composition of atmospheric dustfall

Different types of single particles retrieved from the five functional areas were analyzed. Based on the EDS results, a total of 13 particle types were identified, including quartz, feldspar (plagioclase and potassium feldspar), calcite, kaolinite, dolomite, muscovite, gypsum, Fe-rich mineral, Fe-rich particles, soot aggregate, fly ash, and biogenic impurities (Fig. 6).

Figure 6
figure 6

EDS spectra of different types of particulate matter.

Full size image

Quartz is displayed in Fig. 6a and constituted a bulk particle with sharp edges, an elemental [Si + O] content greater than 99%, and small amounts of other crustal elements. Figure 6b and c present feldspar. The columnar particle in Fig. 6b was identified as plagioclase, which surface was covered with crystalline impurities, and its main elemental content decreased in the order O > Si > Fe > Al > Ca > C > Mg. Figure 6c features a granular particle of potassium feldspar with a main elemental content decreasing in the order O > Si > Al > K > Mg. The elements O, Si, and Al are the main elements of feldspar with an Al:Si ratio of 1:3, and the most common feldspars are included in the KAlSi3O8–NaAlSi3O8–CaAl2Si2O8 system. Figure 6d shows a square particle of calcite, and its main elemental content followed the order O > Ca > C > Si > Al. A granular particle of kaolinite with porous surface is displayed in Fig. 6e. The content of the main elements of this particle decreased in the order O > C > Si > Al > Fe > Ca > Mg. Figure 6f presents dolomite as an irregular particle with a main elemental content of O > C > Ca > Mg > S > Si > Al and a Ca:Mg mass ratio of 5:3, corresponding to the chemical formula of dolomite, CaMg(CO3)2. An irregular particle of muscovite, which surface was covered with crystalline impurities, is displayed in Fig. 6g, the main elemental content was O > Si > Al > C > K > Fe, and the chemical formula of muscovite, KAl2(AlSi3O10)(OH)2, corresponded to the found elemental mass ratio. Figure 6h shows an irregular and sharp-edged particle of gypsum, and the main elemental content decreased in the order O > Ca > S > C > Si > Al, gypsum has a Ca:S mass ratio of 5:4, and the established elemental mass ratio of the particle agreed well with the theoretical value. Figure 6i and j show Fe-rich particles, an irregular particle in (i) with a main elemental content of Fe > O > C and an Fe content of 63.3% and a subspherical particle in (j) with a main elemental content of Fe > O > Si > Al > C and an Fe content of 51.1%. The particle in Fig. 6k was soot aggregate, usually appearing in the form of chains and fluff41,42, having the main elemental content of O > Si > C > Al > Ca > Fe > K in decreasing order. A square particle of fly ash is displayed in Fig. 6l. The main elemental content in this particle followed the order O > Si > Al > C > K. This type of particle is mainly formed at high temperatures43, and the surface is smooth and mostly aggregated with other particles6. Figure 6m shows biogenic impurity as an irregular particle with porous surface with a main elemental content of C > O > Si > Al. In previous studies, particulate matter with a [C + O] content greater than 75% was defined as biogenic particles; the [C + O] content detected in this study exceeded 80%.

Main sources of atmospheric dustfall

Combining the results of mineral composition and elemental composition analysis, 13 types of particles were identified and classified into five categories: silicate particles, carbonate particles, sulfate particles, Fe-rich particles, and other particles.

Silicate particles

Silicate particles are mainly composed of O, Si, and Al and contain small amounts of K, Ca, Mg, Na, or other elements. In this study, quartz, feldspar, kaolinite and muscovite belonged to the category of silicate particles, and all were found in the five functional areas. Silicate particles are crustal components widely distributed in nature44. Those observed in this study had significantly sharp edges, indicating that they had not been transported over long distances and originated mainly from local soil dust7,45, of which kaolinite was mainly produced by weathering of silicate particles. Except for the desert area, silicate particles in each functional area might also stem from industrial dust and construction dust and might be affected by human factors14. A large amount of coal dust is generated during the coal mining process in the open-pit mining area, and the main constituent minerals of coal dust include kaolinite, quartz, and gypsum46. Therefore, coal dust was one of the sources of kaolinite, quartz, and gypsum in this study. Kaolinite from coal-associated ore was also a source of kaolinite in atmospheric dustfall47.

Carbonate particles

Carbonate particles are mainly composed of carbonate anions (CO3)2− and metal cations such as Ca2+, K+, Mg2+ and Na+. In this study, calcite and dolomite belonged to the category of carbonate particles, which were found in the five functional areas. Carbonate particles are crustal components and are widespread in nature. The observed features were similar to those of the silicate particles, indicating that they had not been transported over long distances and mainly originated from local soil dust. Except for the desert area, carbonate particles in each functional area might also stem from industrial dust and construction dust. Large coal transportation vehicles using the concrete road caused the abrasion of the road surface and the production of calcite. These particles were affected by the wind and driven into the atmosphere by the movement of vehicles to be one source of calcite on the off-mine coal road.

Sulfate particles

Sulfate particles are mainly composed of sulfate anions (SO4)2− and metal cations. In this study, gypsum belonged to the category of sulfate particles and was found in the industrial area, the commercial and residential area and the open-pit mining area. The presence of sharp edges in the gypsum indicated that it mainly originated from a local emission source. Gypsum is a widely used industrial and construction material. In this study, gypsum mainly originated from industrial dust and construction dust caused by human factors. In addition to the coal dust generated during the coal mining operations, the chemical reaction of calcite (CaCO3) present in the atmosphere and harmful sulfur dioxide, generated from the spontaneous combustion of coal gangue in the open-pit mining area and from the coal coking plant in the industrial area, would also generate gypsum, indicating that a part of gypsum was the secondary crystallization product of the chemical reaction between atmospheric particulate matter and acid gases48,49.

Fe-rich particles

Fe-rich particles were of subspherical and irregular shapes with Fe contents greater than 50%. Subspherical Fe-rich particles were mainly stemmed from wear of Fe-containing parts of motor vehicles25,50 in all functional areas except the desert area and were also emitted from steel plants in the industrial area. Irregular Fe-rich particles were present in all functional areas, mainly originating from local soil dust51, but the coal gangue emission in the open-pit mining area was also a source of these particles52.

Other particles

Soot aggregate and fly ash were present in all functional areas except the desert area. Soot aggregates were attributed to vehicle exhaust emissions from gasoline and diesel combustion and the incomplete combustion of biomass and fossil fuels53,54. Fly ash originated from the full combustion of biomass and fossil fuels, as well as the spontaneous combustion of coal gangue during coal mining operations55,56 in the open-pit mining area. Biological impurities were present in all functional areas, and based on previous researches, they might stem from biological excreta, plant debris, pollen and spores, as well as human hair and dander7.

Except for the desert area, the other four functional areas are located in the northeast of Wuhai City on the east bank of the Yellow River. The study period was in the windy period, which has the most days of high wind activity over the year, and the prevailing wind direction was northwest. Due to the prevailing wind direction, the other four functional areas were less influenced by the desert area as the source of atmospheric dustfall. Sources of silicate particles, carbonate particles, Fe-rich particles, and other particles in atmospheric dustfall in the industrial area and the open-pit mining area were influenced by the commercial and residential area and the off-mine coal road. The sources of atmospheric dustfall from these areas were mainly affected by human factors. Therefore, it is necessary to focus on road surfaces, industrial emissions, and coal mining operations to take reasonable and effective dust suppression measures to reduce the impact of atmospheric dustfall in Wuhai and other cities located downwind of the air pollution.

Conclusion

  1. (1)

    The atmospheric dustfall was mainly composed of coarse silt, very coarse silt, very fine sand, fine sand, and medium sand in the industrial area and the open-pit mining area. In the commercial and residential area, the desert area, and on the off-mine coal road, the atmospheric dustfall mainly contained very coarse silt, very fine sand, fine sand, and medium sand. The atmospheric dustfall in all functional areas was dominated by coarse-grained particles.

  2. (2)

    The shape of the atmospheric dustfall deviated from spherical shape in all functional areas. There was no significant difference in roughness of the atmospheric dustfall in the five functional areas, indicating that it was less affected by the different functional areas. Except for the desert area, the differences in Roundness, Aspect ratio and Convexity of the atmospheric dustfall in each functional area were small and were found to be mainly influenced by the dominant wind direction and the proximity of each functional area.

  3. (3)

    The variation of shape factors with size distribution was similar in each functional area, and with decreasing particle size, the difference in shape factors increased in each area. The study on the microscopic characteristics of atmospheric dustfall can be further deepened from the correlation in the future.

  4. (4)

    The EDS and XRD analysis indicated the presence of quartz, feldspar, kaolinite, muscovite, calcite, dolomite, Fe-rich mineral and biogenic impurities in each functional area, while gypsum was present in the commercial and residential area, the industrial area and the open-pit mining area. Fe-rich particles, soot aggregates and fly ash was present in each functional area except the desert area.

  5. (5)

    The atmospheric dustfall in each functional area was not transported over long distances and was mainly influenced by local sources. Source analysis was conducted for the different types of particles of atmospheric dustfall, and the specific dust sources in each functional area were analyzed, which indicated that the main sources of atmospheric dustfall were soil dust in all functional areas. And other 15 sources were identified but were not existed in every functional area, because of their different specific functions and development priorities. But there is limitation to the source research in this paper. While the method provided qualitative analysis of source types, it did not allow for quantitative analysis of them.