AbstractSamples of nearsurface atmosphere dustfall and different pollution endmembers were collected in the urban area of Jining City. The element characteristics of the nearsurface atmosphere dustfall and pollution endmembers were analyzed systematically and the spatial distribution of the dustfall and its pollution sources were studied. The results showed that the contents of elements varied obviously in different pollution endmembers. The contents of As, Cd, Cu, F, Pb, S and Se within coal dustfall were the highest, higher than those in atmosphere dustfall and had great influences on the environment. The contents of Cd, Pb, Se, Zn, Hg and CaO within nearsurface atmosphere dustfall were affected by human activities to different degrees. Results of correlation analysis and factor analysis showed that Cd, Pb, Se and Zn mainly came from enterprise coal consumption, with a contribution ratio of 26.32%. The contents of the above four elements near chemical factories, steel factories and high populated regions were generally high, which was consistent with the spatial distribution of the coal pollution sources. CaO was related to traffic pollution, with a contribution ratio of 10.06%. Hg was mainly due to automobile emissions, with a contribution ratio of 8.12%. The contents of As, Cr, F and Ni within dustfall were seldom influenced by human activities and mainly came from soil sedimentation (natural sources), with a contribution ratio of 30%. The above four pollution sources (enterprise coal consumption, traffic pollution, automobile emissions and soil sedimentation) were the main sources of atmosphere dustfall in Jining City. The enrichment coefficients of As, Cr, F and Ni were smaller and the spatial correlations of the four elements were better, while the enrichment coefficients and variable coefficients of elements influenced by human activities, such as Cd, Pb, Se and Hg were larger, and the spatial distributions of these elements were consistent with those of the pollution sources.
　　Key wordsPollution endmember; Atmosphere dustfall; Enrichment degree; Factor analysis; Jining City
　　
　　Received: September 10, 2018Accepted: November 27, 2018
　　Supported by Program of Ecological and Geochemical Survey in the Lower Reach of the Yellow River in Shandong Province (1212010310306).
　　Jierui DAI (1977-), male, P. R. China, senior engineer, devoted to research about agricultural geology and environmental geology. 　　*Corresponding author. Email: daijierui@126.com.
　　At present, urban environmental monitoring has been highly studied on total suspended particulates, aerosol (PM10, PM2.5), atmosphere dustfall and floating dust. There are also many studies on related aspects, while few people have done systematic research on nearsurface atmospheric dust. Nearsurface atmosphere dustfall is atmospheric dust at the average breathing height (1.5 m) of human body, which is a mixture of atmospheric particulate matter and surface dust［1］. With the rapid development of urbanization and industrialization, manmade sources have become the main sources of toxic and harmful elements in nearsurface atmosphere dustfall［2-5］, which carries more complicated pollutants, including nitrogen oxides and hydrocarbons in addition to various heavy metals and organic pollutants. Nearsurface atmosphere dustfall not only affects peoples living environment, but also carries pollutants which would cause secondary pollution to soil, plants and water bodies after settlement, thereby endangering human health. Therefore, understanding the distribution characteristics and changing laws of urban nearsurface dustfall and studying the main pollution factors affecting dustfall are of great significance to the prevention and control of dust pollution, improvement of urban atmospheric environmental quality and guarantee of human health and urban construction planning.
　　General Situation of the Study Region
　　Jining City is located in the junctional zone between Taiyi low mountains and hills in south Shandong and the Huanghuaihai Plain in southwest Shandong. The geological structure belongs to the faultblock sunken area in southwest Shandong, North China region. The terrain is dominated by plain depressions, and high in the east and low in the west. There are mountains and hills in the east, and a flat Yellow River flood plain in the west. The central region of Nansi Lake (Nansi Lake is the general name of Weishan Lake, Nanyang lake, Zhaoyang Lake and Dushan Lake) which runs through from north to South. Jining City belongs to the zone of the warm temperate continental monsoon climate, with an average temperature in the range of 13.3-14.1 ℃ and an average annual precipitation in the range of 597-820 mm. The total area of the city is 10 685 km2, and the urban area is 1 262 km2.
　　Jining City is one of the three major industrial center cities of Shandong Province, which is the focus of the planning and construction of Shandong Province. It has been initially formed a relatively complete industrial system. With the rapid development of city modernization and industrialization in the past 30 years, Jining City and the surrounding area have set up a number of enterprises, such as chemical plants, chemical fiber factories, cement factories, brick factories, chemical fertilizer factories and brewery, etc. Building industry has developed rapidly over the past 10 years, the amount of vehicle increases sharply, and coal, industrial dust, road dust, vehicle exhaust dust and construction dust have a serious impact on the citys air quality. In recent years, many researchers have investigated soil environmental quality and air quality in Jining City［6-7］, but no studies have been conducted on the distribution and pollution characteristics of heavy metals in the near surface atmospheric dust. Based on the study of the pollution source and pollution characteristics of near surface air dust in Jining City, correlation analysis and factor analysis were used to calculate the contribution of each source to the atmospheric dust, and the possible sources were discussed. 　　Sample Collection and Analysis
　　Pollution end sample collection
　　The distribution of industry, transportation and construction in Jining City were investigated in detailed. There were traffic dust, construction dust, coal dust and vehicle exhaust dust place points which are exhaust emissions and serious pollution. Each kind of dust was sampled at 6 points. Traffic dust was sampled in urban areas with large traffic flow. Building dust samples were collected in large scale urban construction sites. Coal dust should be taken into account when sampling the dominant wind direction, and samples were collected in the downwind area of coal enterprises. Floating dust was collected from 1.5 to 2.0 m around the upper part of the artificial platform and the building. Coal dust samples can also be collected on the inner wall of chimney and surrounding. Automobile exhaust dust sampling points were distributed in the large parking lot, bus station and gas station: exhaust dust was directly in scraped in the automobile exhaust pipe, vehicle exhaust dust of the same type was merged into one sample, and car types and the type of gasoline were recorded in detailed. The above samples were between 20 to 50 g, which can meet the requirements of the quantity of analysis and preservation.
　　Nearsurface atmospheric dust collection
　　The nearsurface atmospheric dust samples were collected by the grid pattern acquisition method (1∶5 million topographic map), and the sampling density was one sample per square kilometer. A portable GPS was used to determine its geographical coordinates, and the fixed point error was not more than 1 mm on the graph, which meant the actual distance was not more than 50 m. The samples were collected in the middle and lower layers, mainly flat roof room, windows of residential buildings, poles, trees, bus stop, etc., by sweeping with a brush and holding the dust with a clean plastic bag. Sampling should be away direct pollution sources (such as industrial pollution, civil coal, paint, etc.) as far as possible. The method of multipoint and equal volume mixing was used to collect 306 samples per point, which was about 20 g. In order to compare with native soil, the surface soil samples were collected at the same point of dust in the green belt or park and so on. The 0-20 cm soil was collected vertically after removing surface debris when ensuring uniform collection, and animal and plant residue, gravel and fertilizer mass were discarded. The sampled soil was put into a clean bag, and the original sample weight was greater than 200 g. Soil samples were air dried and crushed, and sieved with a 20 mesh nylon sieve, and 80 g was taken to laboratory for analysis. 　　
　　Fig. 1Sampling positions of the pollution endmembers and nearsurface atmosphere dustfall in Jining
　　Sample analysis
　　The pollution endmember samples were analyzed for such 19 indicators as As, Cd, Co, Cu, Cr, F, Hg, Mn, Ni, Pb, S, Ti, Zn, Se, Na2O, Al2O3, CaO, K2O and MgO. Nearsurface atmospheric dust and soil samples were analyzed for such 10 indicators as As, Cd, Cr, Hg, Ni, Pb, Zn, F, Se and CaO.
　　The sample test was completed by Shandong Institute of Geological Survey. The indicators such as Co, Mn, Ni, Zn, MgO and CaO were determined by plasma emission spectrometry after digestion with HF+HNO3+HClO4+aqua regia. During the determination of indicators such as Pb, Cr, Ti, Al2O3 and K2O, the sample powder was pressed under 35 t pressure to pieces and then analyzed by Xray fluorescent spectrometry. As to indicators such as As, Hg and Se, the sample was digested with 1+1 aqua regia into a boiling water bath, and subjected to redox treatment by adding potassium permanganate and oxalic acid, and then analyzed with an atomic fluorescence spectrometer. The Cd element was determined by digesting with HF+HNO3+HClO4+ aqua regia, adding matrix modifier, and analyzing by graphite furnace atomic absorption spectrometry. For the determination of F element, the sample was melted with NaOH at 700 ℃, extracted with water, and determined with an F ion selection electrode on an ion activity meter after the addition of an ion intensity regulator. S element was determined by burning the sample with a tube furnace burn and analyzing by iodometric method. Various monitoring methods such as standard sample, code sample and monitoring sample were applied, to guard the analysis quality, and the testing quality passed the test by the expert group from China Geological Survey Bureau.
　　Result and Discussion
　　Characteristics of pollution endmember element contents
　　As shown in Table 1, the contents of most indicators from building dust and coal combustion were mostly stable, while the variable ranges of indicator contents from automobile exhaust were larger, for example, the maximum of Zn content was 34.67 times of the minimum in the automobile exhaust, and the maximum of Ti was 30.18 times of the minimum. These conditions might be connected with automobile performance and fuel quality, but the mean data still could reflect the overall level of element pollution from automobile exhaust.
　　Compared with nearsurface dust, the contents of all elements from coal combustion were slightly higher (the specific values were greater than 1.2). Among them, the content of Se was 502.54 times of dustfall; the content of As was 9.61 times of dustfall; and the contents of other elements were 1.33 (Cr)-8.33 (Cd) times of dustfall. In traffic dust, except Cr and Ni, other elements were 1.42 (F)-7.11 (Zn) times of dustfall. In building dust, the contents of Pb, CaO, Se, As and Zn were slightly higher. Among them, the content of Pb was 11.25 times of dustfall, and other elements were 1.63 (Zn)-5.11 (Cao) times of dustfall. In automobile exhaust, the content of Hg was 3.11 times of dustfall, and Zn, Se, Ni and Cr were 1.34 (Cr)-1.78 (Zn) times of dustfall. All the data indicated the elements from the pollution endmembers can enter atmosphere and speed up accumulate of heavy metals in dustfall. 　　Jierui DAI. Geochemical Characteristics and Pollution Source Identification of the Nearsurface Atmosphere Dustfall in Jining City, Eastern China
　　Table 1Statistics of the contents of pollution endmember elements in Jining
　　Indicators
　　Traffic dust
　　Content range Mean
　　Building dust
　　Content range Mean
　　Coal dust
　　Content range Mean
　　Automobile exhaust dust
　　Content range MeanAtmosphere dustfall
　　As13.53-20.0916.0214.48-22.8819.1313.01-222.0582.402.48-9.046.048.57
　　Cd0.71 -1.741.260.12-0.340.200.36-2.611.500.06-0.230.150.18
　　Co11.33-13.5411.988.11-14.089.9714.31-27.9922.152.46-16.706.29--
　　Cr64.01-91.1876.8938.99-80.8251.9175.01-123.5092.3736.06-156.3292.8969.20
　　Cu49.88-90.4872.3027.27-190.6273.3576.39-288.74164.7218.99-143.3877.83--
　　F533-940669272-451384336-95113 34480-178134471.00
　　Hg335-722460.847-11869.4274-830498240-840570183.00
　　Mn446-518487259-1472681220-130558745-135104--
　　Ni22.11-29.4926.529.64-26.3916.1533.24-99.4163.237.70-201.0061.8727.14
　　Pb116.3-268.0172.341.0-1 231.2272.847.1-613.4362.16.0-17.511.332.20
　　S0.40-0.610.500.23-0.530.390.46-12.643.581.36-2.251.87--
　　Se2.21-5.493.361.68-5.892.8528.85-852.07296.501.83-4.963.620.59
　　Ti0.35-0.390.370.19-0.260.210.31-0.720.440.003-0.0850.022--
　　Zn307-1 81342559-292140158-46530637-129767286.10
　　Al2O310.26-11.3110.799.07-11.089.759.59-23.7814.830.18-0.660.38--
　　CaO9.62-19.1317.0713.88-27.0118.154.05-14.479.000.43-1.580.983.55
　　K2O1.53-1.761.651.50-2.091.820.61-1.170.840.04-0.110.06--
　　MgO1.43-2.061.751.06-2.711.570.05-0.530.200.07-0.220.13--
　　Na2O1.60-1.821.682.10-2.332.240.78-1.531.240.08-0.170.13--
　　The content of Hg should be multiplied by 10-9; the contents of S, Ti, Al2O3, CaO, K2O and MgO should be multiplied by 10-2; the contents of remaining elements should be multiplied by 10-6; and -- means not counted.
　　The comparison of different pollution endmember elements are shown in Fig. 2. For convenience of comparison, Fig. 2 adopts standardized mapping, that is, the element contents were multiplied by 10n (n is an integer), and the data results were controlled within 25.
　　The contents of elements were very different between different kinds of pollution endmembers. The contents of As, Cd, Co, Cu, F, Ni, Pb, S, Se, Ti and Al2O3 were higher than others in the coal combustion dust. The content of MgO was the highest in traffic dust, and in addition, the contents of Cd, Co, Hg, Ti, Zn, CaO, K2O and Na2O had a high level. The contents of Hg, Cr and Zn were the highest in automobile exhaust. The content of Hg was 1.14 (coal combustion dust)-8.21 times(building dust) of others, while the contents of As, Cd, Co, F, Mn, Pb, Ti and oxide were minimum in all kinds of endmember dust. The elements showed a polarizing trend in building dust. Compared with other endmember, the contents of Cr, Hg, Ni, S, Se and Zn were the lowest. The contents of Mn, CaO, K2O and Na2O were the highest. 　　On the whole, coal combustion dust (As, Cd, Co, Cu, F, Ni, Pb, S, Se, Ti and Al2O3) had the highest impact to the environment. Automobile exhaust (Hg, Cr, Zn, S and so on) and traffic dust (Cd, Co, Hg, Ti, Zn and oxide) were in the middle, while building dust (Mn, CaO, K2O and Na2O) had the lowest impact to environment. Meanwhile, the element combination can deed symbolic element of different kinds of pollution endmembers.
　　
　　Fig. 2Comparison of the contents of different pollution endmember elements in Jining
　　Characteristics of atmospheric dust element contents
　　As shown in Table 2, the elements in the nearsurface atmospheric dust and in the soil were all in significant positive correlation at the 0.01 significant level, indicating that the dust and the soil had a close relationship.
　　The content ranges of As, Cr, Ni and F in dust changed slightly, with the coefficients of variation less than 0.4, and the correlation coefficients of the corresponding elements in the soil were greater than 0.4. The scatter distribution was approximately linear and the contents in soil were almost equal or slightly higher (Fig. 3). It can be believed that the above elements were not enriched at all relative to the soil and mainly from soil dust. Although Pb and Zn in dust were similar to those mentioned above elements, That is, they all had the characteristics of large correlation coefficients, almost equal contents and small coefficients of variation, seeing from the comparison of geochemical map of Pb element between dust and surface soil (Fig. 4), Pb abnormalities also occured in urban areas. The content of Pb in dust was more than 40×10-6, which was consistent with the spatial distribution of urban enterprises, and this range was slightly larger than the soil. This fully reflected the human activities leading to lead pollution and zinc pollution in dust, and affected the distribution of related elements in soil. Although the average content of Hg in dust was close to the soil background value (the enrichment coefficient was 0.92), the coefficient of variation was as high as 1.35. It showed that the distribution of Hg element in the dust was uneven, and the mercury pollution was especially serious in densely populated urban areas.
　　Compared with the soil background value, the contents of Se, Cd and CaO in dust were significantly high and the enrichment degrees of elements from high to low in turn were Se (2.07), CaO (1.67) and Cd (1.32). Meanwhile, the correlation coefficients of the corresponding elements in soil were less than 0.3. As shown in the scatter plot of Se, Cd (Fig. 5), the distribution pattern shows that there were a few extremely high content points in dust. These high content points might be caused by the superposition of artificial point source pollution. Through the above analysis, it could be seen that Hg, Se, Cd, Pb, Zn and CaO in dust may be affected by human activities. However, the elements of As, Cr, Ni and F may mainly come from soil dust (natural sources). 　　Table 2Statistics of the contents of the nearsurface atmosphere dustfall elements in Jining
　　ElementContent rangeMeanMidvalueStandarddeviationCoefficientof variationSoil backgroundvalue EnrichmentcoefficientCorrelationcoefficient
　　As3.07-17.798.578.312.440.298.561.000.343
　　Cd0.08-0.760.180.160.090.490.141.320.209
　　Cr33.03-186.9869.2067.5816.970.2574.250.930.342
　　F146-781471459104.200.225480.860.388
　　Hg16-2 188183110248.221.351990.920.519
　　Ni12.92-46.9627.1425.836.520.2432.990.820.514
　　Pb14.01-99.6232.2129.7010.520.3331.291.030.620
　　Se0.14-6.460.590.390.621.050.292.070.208
　　Zn42.4-301.886.177.334.280.4084.01.020.345
　　CaO1.19-29.073.552.562.800.792.161.640.295
　　The content of Hg should be multiplied by 10-9; the content of S should be multiplied by 10-2; and the contents of the remaining elements should be multiplied by 10-6; the number of statistical samples is 306, and at the confidence value α=0.01, the relevant critical value is about 0.207.
　　
　　Fig. 3Scatter diagrams of As and Ni in nearsurface atmosphere dustfall and in surface soil in Jining
　　
　　A: Geochemical distribution of Pb in near-surface atmosphere dust-fall; B: geochemical distribution of Pb in surface soil.
　　Fig. 4Geochemical distribution of Pb in nearsurface atmosphere dustfall and in surface soil in Jining
　　
　　Fig. 5Scatter diagrams of Se and Cd in nearsurface atmosphere dustfall and in surface soil in Jining
　　Correlation analysis between elements in dust
　　The correlation coefficient table (Table 3) of the contents of elements in dust in Jining City shows that the correlation between nonenriched (nonpolluting) elements which may originate from soil particles such as As, Cr, Ni and F and had enrichment coefficients less than 1 was better. The correlation coefficients between Ni and As, Cr and F and that between Cr and F were all above 0.55. Specifically, the correlation coefficients were 0.556, 0.776, 0.744 and 0.579, respectively. The correlation coefficients between As and F and Cr were 0.448 and 0.450, respectively.
　　Table 3Correlation coefficients of the element contents in nearsurface atmosphere dustfall in Jining
　　ElementAsCdCrFHgNiPbSeZnCaO
　　As1.000
　　Cd0.1301.000
　　Cr0.450-0.1031.000
　　F0.448-0.1610.5791.000
　　Hg0.0860.192-0.013-0.0441.000
　　Ni0.556-0.1530.770.744-0.1211.000 　　Pb0.2590.4400.004-0.0410.184-0.0091.000
　　Se0.1080.216-0.128-0.2010.125-0.0680.3771.000
　　Zn0.3200.4540.0610.0160.1660.10.6060.3811.000
　　CaO0.0080.181-0.393-0.223-0.001-0.3980.1290.2330.3861.000
　　The correlation between enriched (polluting) elements which may originate from manmade pollution such as As, Cr, Ni and F with enrichment coefficients greater than 1 was general. Among them, the correlation between elements of Cd, Pb, Se and Zn were relatively high. The correlation coefficient between Pb and Zn was 0.606. However, the correlation coefficients between other elements were between 0.216 (Se and Cd) and 0.454 (Cd and Zn). CaO was negatively correlated with Cr, F, Ni and other elements and the correlation coefficients were below -0.4. This showed that the sources of these pollution elements had a certain correlation, but the correlation was less obvious than the nonpollution elements, which might be because that the sources were more complex. However, the correlation between Hg and other elements was poor (the correlation coefficient was less than 0.2), and the main source was obviously different from other elements.
　　Analysis of the contribution of polluted end to dust
　　In order to analyze the relationship between elements in dust in Jining City and further determine their sources and control factors, principal component analysis of data was carried out by SPSS software and an initial factor load matrix was obtained. In order to eliminate the influence of the contents of different elements on the order of magnitude, the maximum variance orthogonal rotation reduction was used to obtain the factor load matrix. It could be seen from Table 4 that the cumulative contribution rate of the 4 factors was 74.50%, that is, 4 types of pollution sources were selected.
　　Table 4Factor load matrix of the elements in nearsurface atmosphere dustfall in Jining
　　Element variable
　　Factor load
　　Factor 1Factor 2Factor 3Factor 4
　　As0.7270.2010.2040.046
　　Cd-0.1440.889-0.007-0.029
　　Cr0.806-0.019-0.2990.042
　　F0.823-0.08-0.108-0.101
　　Hg-0.0800.194-0.1510.817
　　Ni0.933-0.055-0.143-0.023
　　Pb0.0730.6700.1460.176
　　Se-0.0170.6060.2060.206
　　Zn0.1900.7030.2790.103
　　CaO-0.2760.1690.794-0.167
　　Characteristic value3.002.631.010.81
　　Contribution rate∥%30.0026.3210.068.12
　　Cumulative contribution rate∥%30.0056.3266.3874.50 　　In factor 1, As, Cr, F and Ni had larger factor coefficients. The critical value of correlation coefficient was 0.207 at the 0.01 confidence level. Therefore, the 4 elements are significantly correlated with the first factor. The previous studies showed that for F, Ni and Cr as the main elements of coal combustion emissions, their average contents in burning coal dust were up to 3 344×10-6, 63.23×10-6 and 92.39×10-6, respectively. Se was the indicator element of coal combustion［8-9］, and the average content was 269.50×10-6. If Se and these 4 elements As, F, Ni and Cr appear simultaneously, the 4 elements may be derived from burning coal. There was Se without As, F, Ni and Cr, so coal was not the main source of the 4 elements. The correlation coefficients of Se and F, Ni and Cr in study area were -0.201, -0.068 and -0.128, respectively. The correlation between Se and As was also poor (the correlation coefficient was 0.108). In addition, the contents of the 4 elements in dust were close to or slightly smaller than the soil values. There was significant positive correlation between them and they had the same source. It could be seen from the above analysis that the first factor represented the natural source, that is the settlement of soil dust from local or adjacent areas and the contribution rate was 30%.
　　Factor 2 was significantly related to Cd, Pb, Se and Zn. The correlation analysis showed that Cd, Pb, Se and Zn in dust in Jining city may have the same source. Compared with the soil background values, they all had a higher content or greater coefficient of variation. From the spatial distribution of element content, it could be seen that Cd, Pb, Se and so on were related to coal combustion activities. It was also found that the dust sample points which were near the urban central, iron and steel plants and chemical plants had the highest content of Cd (0.76×10-6) and higher contents of Se (4.16×10-6 ) and Zn (252×10-6). Therefore, it was proved that these elements were mainly derived from industrial coal. It was considered that coal combustion was the second pollution factor, and its contribution rate was 26.32%.
　　According to relevant research data［10-12］, Pb is a tracer element of automobile exhaust emissions. However, this research showed that in automobile exhaust dust the content of Pb was the lowest. In dustfall, the Pb was controlled by F2 factor representing coal source. This shows that with the compulsory use and popularization of unleaded gasoline, industrial coal has gradually become a major source of Pb in atmospheric dust. 　　Factor 3 was significantly related to CaO and had a relatively large factor coefficient with Zn, so we can infer that the third factor came mainly from the second traffic pollution. The traffic dust was collected in the dust of heavy traffic roads and was a mixture of soil particles and automobile exhaust dust. The previous analysis showed that Zn was the main element of coal combustion emissions. CaO had the highest content in construction dust and the soil also contained a certain amount of CaO. The correlation coefficient of two elements of different sources was 0.386. It was concluded that the two elements in dust came from the second polluted traffic dust. In addition, the ratio between CaO and Zn can also be confirmed. CaO/Zn in atmospheric dust was 412.3, which was close to traffic dust (CaO/Zn was 401.6), and far from the construction dust (1 296.4), burning coal dust (294.1) and exhaust dust (14.6). Therefore, the third factor represented the source of traffic dust and the contribution rate was 10.06%.
　　Factor 4 was only significantly related to Hg. Cars and coal will produce a certain concentration of Hg, but the correlation between Hg and coal characteristic elements such as As, Se and Pb and so on was very poor. In view of the above reasons, it was concluded that PC4 was automobile emission factor, and the contribution rate of atmospheric dust was 8.12%.
　　Spatial distribution characteristics of atmospheric dust elements
　　The spatial interpolation method of DTM module of Map GIS software was used to make the element distribution map. It was found that the contents of elements in dust were different in different regions of Jining. From the F2 factor score chart (Fig. 5), it could be seen that the second factor high score area associated with elements such as Cd, Pb, Se and Zn was in good agreement with the distribution scope of urban enterprises. This fully reflects the consistency of the geographical distribution of the pollution sources. Especially, the scores were the highest in the chemical plants, steel plants and densely populated urban areas. The highest score was far higher than that in other parts of Jining, which was related to the longterm burning of coal by residents and enterprises. Although influenced by regional soil dust, the F1 (As, Cr, F and Ni) scores had no significant differences in different regions, but were relatively higher in the western brick factory, the eastern chemical plant and the steel plant. This might be related to pollution emissions in the region. 　　The average content of CaO in dust near Jining bus station was 8.4%, and the maximum value was 20.4%. Meanwhile, the content decreased gradually around the center of the station, it was related to the large traffic flow in this area. The high content area of Hg element in dust was consistent with the population density area in urban area. In addition, the content was the highest in the bus station and along the main roads in the city, and in other parts of the city, the content was relatively stable, and low in suburban the content.
　　
　　Fig. 6Spatial distribution of F2 factor scores in nearsurface atmosphere dustfall in Jining
　　Conclusions
　　The contents of elements in different pollution ends were obviously different. The contents of As, Cd, Co, Cu, F, Ni, Pb, S, Se and other elements in coal dust were the highest. Among them, the contents of Se, F and As were 4 times more than other end dust. In coal dust, the content of Se was 81.91 times of that in tail gas and 104.04 times of that in building dust. Burning coal dust had a great influence on environment. However, most of the elements in the building dust were the lowest among all the pollution ends and only the content of oxides and Mn element were high, so building dust had little impact on the environment. The effects of other end dust on the environment were in the middle. In the past, Pb was a tracer element for automobile exhaust emissions, but with the popularity of the compulsory use of unleaded gasoline, industrial coalfired emissions have gradually replaced automobile exhaust as the main source of Pb in atmospheric dust.
　　The elements of As, Cr, F and Ni in the dust which are not affected by human activities were more consistent in spatial distribution, and had better correlation coefficients and smaller enrichment coefficients. The contents of these elements in dust were mainly affected by soil dust deposition (natural source). However, the elements of polluting Cd, Pb, Se, Zn, Hg and CaO (indices) which are influenced by human activities to different degrees had greater enrichment coefficients or coefficients of variation, the correlation was poor, and the spatial distribution was consistent with the pollution source. Studies have shown that Cd, Pb, Se, and Zn mainly originate from coal combustion enterprises. CaO related to traffic pollution and Hg mainly came from vehicle exhaust emissions. The above 4 kinds of pollution sources were the main sources of dust in Jining. Specifically, the contribution rates of soil dust, coal dust, traffic dust and exhaust dust were 30%, 26.32%, 10.06% and 8.12%, respectively. 　　The results of this study show that that the content analysis, correlation analysis and factor analysis of the elements of different pollution ends and atmospheric dust are effective methods to indicate the sources of pollution elements in urban atmospheric dust. In view of the current situation in Jining, the prevention and control of coalburning pollution is an important task. With the rapid growth of car ownership, automobile exhaust pollution cant be ignored.
　　References
　　［1］ XI DL, SUN YS, LIU XY. Environmental monitoring［M］. Beijing: Higher Education Press, 2002. (in Chinese)
　　［2］ LIU HT, YU RL, HU GR, et al. Assessment of pollution by heavy metal street dust in the Quanzhou City［J］. Journal of Mineralogy and Petrology, 2010, 30(3): 116-120. (in Chinese).
　　［3］ YANG XZ, CHEN Y, XU DD, et al. Characteristics of heavy metal pollution and health risk assessment in subway dust in Beijing［J］. China Environmental Science, 2011, 31(6): 944-950. (in Chinese).
　　［4］ LIN X, LIU M, HOU LJ, et al. Soil and surface dust heavy metal pollution state and assessment in Shanghai City［J］. China Environmental Science, 2007, 27(5): 613-618. (in Chinese).
　　［5］ HU XM, WANG LP, BI JH. Research on the heavy metal pollution in city atmosphere［J］. Journal of Anhui Agricultural Sciences, 2008, 36(1): 302-303. (in Chinese).
　　［6］ SI MR, SU T, LI GZ. Investigation on soil actinomycetes in Jining City［J］. Shandong Agricultural Sciences, 2008, 21(4): 68-71. (in Chinese).
　　［7］ LI Y, WANG YQ, LIU Q. The impact of spring wind on the air quality in Jining［J］. Journal of Anhui Agricultural Sciences, 2009, 37(3): 1212-1213. (in Chinese).
　　［8］ YANG LP, CHEN FH. Study on the source apportionment of atmospheric dust pollutants in Lanzhou［J］. Acta Scientiae Circumstantiae, 2002, 22(4): 499-502. (in Chinese).
　　［9］ HE B, LIANG LN, JIANG GB. Distributions of arsenic and selenium in selected Chinese coal mines［J］. The Science of the Total Environment, 2002, 296: 19-26.
　　［10］ LEWIS J. Leand poisoning: A historical perspective［J］. EPA Journal, 1985, 11: 15-18.
　　［11］ WANG MX. The impact of space heating in winter on the air quality in Beijing［J］. Environmental Sciences in China, 1984, 4(3): 36-40. (in Chinese).
　　［12］ DAI SG, ZHU T, ZENG YS, et al. The source of fly ash analysis of heating period in Tianjin City［J］. Environmental Sciences in China, 1986, 6(4): 24-30. ( in Chinese).