Information Technology Journal

Year: 2014 | Volume: 13 | Issue: 9 | Page No.: 1577-1590
DOI: 10.3923/itj.2014.1577.1590

Analysis of Heavy Metal Pollution in Urban Topsoil

Yalin Miao, Xin Ji, Wei Wei and Wen Wen

Abstract: With the rapid development of economic and the increasingly urban populations, the impacts of humans’ activities on the environment quality of the city is prominently increased. In this study, for the spatial distribution and concentration of the sampling points of heavy metal elements in the city, evaluate soil heavy metal pollution by using Single-factor index method, Nemero index method and cumulative index method, determine the main causes of pollution in each region by the cluster analysis method and the principal component analysis method. Finally, we get the position coordinates of pollution sources in the city by building and solving the three dimensional partial differential model of migration of heavy metal pollutants in soil environment system. In order to provide better research for the evolution model of the urban geology environment, we collected more information and evaluate the advantages and disadvantages of the models.

Keywords: partial differential model, single-factor index method, Soil heavy metal, nemero index method, cluster analysis method, pollution and principal component analysis method

INTRODUCTION

In the past, soil pollution is not as highly valued as air pollution and water pollution. The soil pollution is often wide and volume and is more difficult to control and governance than air pollution and water pollution. However, after effectively governance of air and water pollution, the seriousness and urgency of soil pollution in the developed country is fully revealed.

At present, method to improve and manage the heavy metal pollution contaminated with precipitation, elution method, antagonistic method, electrochemical method and magnetization method. These methods have common disadvantage include investment cost is too high and can only be used for the place of very seriously pollution. And the analysis before only use cluster analysis and principal component analysis is rarely consider the human factors in the cause of pollution, in this study, we solve the influence of artificial factors based on the use of cumulative index method. The propagation characteristics of pollutants are analyzed with the degree of the pollution deviation from and the center, the obtained characteristics is that by the center spread to the surrounding, along with the increase of distance, the more distance far from the heavy pollution area and the more degree of heavy metals pollution decreased. Generally, we rarely focus on the soil water migration and movement of soil heavy metal, in this study, we consider the soil heavy metal pollutants in the process of migration also will be affected by the chemical and physical properties of soil water and solute itself and other chemicals in the soil and treat the factors as the main basis of pollutant transmission characteristics, establish three dimensional partial differential model, calculate each direction changes of the concentration accurately, provide a more reliable method to determine the location of pollution sources.

In this study, Along functional lines, urban area can generally be divided into living areas, industrial areas, mountain areas, trunk roads, parks district and so on, respectively, are recorded as class 1 area, class 2 area,Y and class 5 area, different areas affected by human activities to varying degrees. Now, we carry out the investigation of geologic environment of the soil in the city. Therefore, the study area is divided into the grid area intervals of 1 km, getting sample and number according to 1 sampling points per square kilometer of topsoil sampling (0-10 cm depth) and using GPS to record the position of the sampling point. Application of special equipment testing analysis, we obtain every sample contained many kinds of chemical elements concentration data. Using special test equipment to analysis, the concentration data of various chemical elements contained in each sample was obtained. On the other hand, sampling according to the distance 2 km away from people and industrial activities in the natural area and use the result as the background value of elements of urban topsoil. The obtained data include, (1) Location, altitude and its subordinate function area information of 319 sampling point; (2) A list of concentration of 8 kinds of heavy metals (As, Cd, Cr, Pb, Hg, Cu, Ni, Zn) in the sampling point. (3) Lists the background values of 8 main elements of heavy metal (specific data in the appendix 1). In this study, we build the mathematical model and analysis the soil heavy metal pollution according to the obtained information.

MATERIALS AND METHODS ANALYSIS ON SPATIAL DISTRIBUTION AND POLLUTION DEGREE OF HEAVY METALS

This study presents five different environmental region of a city, in consist of living areas, industrial areas, mountain areas, trunk roads and parks district and concentration values of large number of sampling points and the 3D coordinates (specific data in appendix), due to the discrete degree of sampling point distribution, it is necessary to use excel to select and integrate the large amounts of data, result in the scatter distribution of the five regional sampling points in Fig. 1.

In order to get spatial distribution map of heavy metals more specific, we use concentration distribution of eight kinds of heavy metal in consist of As, Cd, Cr, Cu, Hg, Ni, Pb and Zn and the three-dimensional map of terrain distribution of five regions in consist of living areas, industrial areas, mountain areas, trunk roads and parks district to instruct. First the concentration distribution map of eight kinds of heavy metals in the city are obtained by interpolation function (Yue and Wang, 2012) reference in MATLAB (Li, 2012) in Fig. 2-9.

Then get the 3D terrain distribution map of five area in the city by interpolation function in MATLAB in Fig. 10.

Since each region has eight kinds of heavy metals, the evaluation model is divided into Single-factor index method and the Nemero (Zhu et al., 2012) index method. First of all, establish single exponential for pollution:

Type in: i take 1-8, respectively corresponding to eight kinds of elements in consist of As, Cd, Cr, Cu, Hg, Ni, Pb and Zn, I_i is the index of pollution elements i, C_i is the average concentration of elements i in the region, S_i is the national evaluation standard of element i, B is background values of element i in the soil. Then i kind of pollution elements in partial of five regions obtained by solving the single pollution index in Table 1-5. (Note: The unit for heavy metal elements is μg g^-1).

Nemero index method: The Single-factor index method (Xu et al., 2007) result in the pollution degree of heavy metals in each region, in order to obtain the comprehensive pollution level of each region, we use the Nemero index method Nemero index expression:

Type in: P is the comprehensive pollution index, I_Avg is the average value for each element of pollution index, I_Max is the maximum value of each element of pollution index.

The comprehensive pollution index of each region is obtained by solving the Nemero index formulation in Table 6 and 7.

Cumulative index method: Although the Single-factor index method (Yang et al., 2007a) and the Nemero index method can be more comprehensive evaluation for heavy metal pollution level of soil, since this two kind of method cannot separate the artificial exception from the natural exception, the cumulative index method make up for its insufficient, the cumulative index method expression is:

Type in: Igeo is the cumulative index, C_n is the average of element n in each region, B_n is the geochemistry background value of element n, k is the coefficient of the change of background that lead by different rock in each areas (value = 1.5). Result in the degree of heavy metals pollution in each area by solving the cumulative index expression in Table 8.

Evaluation results: Analyze the degree of heavy metals pollution in different regions of the city through the Nemero index method and the cumulative index method, result in that, in living areas, elements As, Cd, Cr, Hg and Pb belong to light pollution, element N_i belongs to none pollution, elements Cu and Zn belong to medium pollution, comprehensive pollution level is alert level. In industrial areas, elements As, Ni and Cr belong to light pollution, elements Cu, Pb, Zn and Cd belong to medium pollution, Hg belongs to serious pollution, comprehensive pollution level is serious level. In mountain areas, elements As, Cd, Cr, Cu, Hg, Ni, Pb and Zn belongs to none pollution, comprehensive pollution level is clean level. In trunk roads, elements As, Cd, Cr, Cu, Ni, Bp and Zn belong to light pollution, only Hg belongs to serious pollution, comprehensive pollution level is light level. In parks district, most elements belong to light pollution, few places belong to medium pollution, comprehensive pollution level is clean level.

REASON ANALYSIS OF HEAVY METALS POLLUTION

This study presents five different environmental area of a city, in consist of living areas, industrial areas, mountain areas, trunk roads and parks district, through the analysis of question 1, we get contaminated degree of five areas is different and the content of eight kind of heavy metals are also different in each region. In order to analyze the main reason for the heavy metal pollution of the city, we analyze through cluster analysis method (Ma et al., 2008) and principal component analysis method (Liu and Guo, 2003), result that, for one part of the heavy metal pollution is serious and for the other part the heavy metal pollution is light, the serious part becomes the main cause of pollution.

Cluster analysis method: Clustering analysis (Wang and Sun, 2007) is an ideal multivariate statistical technique, mainly include the hierarchical clustering method and iterative clustering method. Clustering analysis is also called group analysis and point group analysis, is a multivariate statistical method to study the classification. For this problem, first of all, to standardize the raw data (part of data in appendix 2) to obtain the content analysis results of heavy metal elements in each area in Table 13.

In the table above, compare average value of the content of soil heavy metals element in living area with its corresponding background value of soil in China, we can find that As, Cd, Cr, Cu, Hg, Ni, Pb and Zn are all over the city soil background values. In contrast, elements Cr, Hg, Pb and Zn have larger coefficient of variation, all above 30%, Centered, elements Cd and Cu have centered coefficient of variation, As and Ni have smaller coefficient of variation Then we put the data of Table 9 into the SPSS software, using the hierarchical clustering method to obtain the hierarchical diagram in living area clustering in Fig. 11.

According to Fig. 1, divided Cd and Cu into the first category, Cr, Hg, Pb and Zn into the second category, As and Ni into the third category.

The first two elements content of the soil is relatively smaller than background element content, so it is a natural source that has less influence of personal factors; The second four elements content of the soil is relatively larger than background element content, so it is a natural source that has more influence of personal factors; The third two elements content of the soil is relatively larger than background element content, so, it has more influence of personal factors. We only list the living areas here, other regional analysis method is same as the above (specific data in the appendix 2), do not do too much description, other regional cluster pedigree chart in Fig. 12-15.

Principal component analysis: Principal component analysis (Li, 2010) is also called PCA, aimed at using the idea of dimension reduction, change the multiple indicators into a few more composite indicator.

In the empirical research, in order to analyze the problem comprehensively and systematically, we must consider influence of many factors. These factors that involved in are generally referred as index, also known as a variable in the multivariate statistical analysis. For every variable in different extent reflects the research problem by certain information and the indexes have a certain correlation between each other, so the obtained statistical data reflect information to a certain extent of overlap. When we use the statistical method to study multivariate problems, too many variables can increase the amount of calculation and complexity of the problem analysis, it is hoped that in the process of quantitative analysis, less variable can be involved in, the we can get more information.

First we lead the raw data of heavy metal elements in living areas into SPSS software to analyze the principal components, obtain the main component load matrix in Table 14 of various kinds of heavy metal elements in living areas in Table 15.

The principal of number extraction of principle component is the first m principle component that corresponding eigenvalues of component larger than 1, through Table 15, three principal components are extracted.

Get the convenient of each index in three ingredients with the data in Table 14 divided by extract square of the number of principal components corresponding to eigenvalue, namely eigenvector A1, A2, A3 (Tan et al., 2006), result in the principal component expressions by multiplication of the feature vector by the standardized data:

Take the proportion of eigenvalue of each main component and the corresponding the total the sum of the eigenvalues that extract from as the weighting to calculate the principal component comprehensive model, use the corresponding coefficient of each index in the first principal component F1 multiplied by the first principal component contribution rate divided by the extraction of the F1 of the sum of the three principal component contribution rate and then add the second principal component corresponding coefficient multiplied by the percentage contribution of F2 and add the third principal component F3 corresponding coefficient multiplied by the percentage contribution, obtained the weight coefficient A of the comprehensive principal component F in each index (Ma et al., 2011). Descend order according to the weight coefficient A values, the main reason of pollution is the heavy metal elements that corresponding larger coefficient. Here only order the living areas, other regional analysis method are the same, we do not do too much description, principal components load matrix of various heavy metal elements of the other four in the region Table 16-20.

Evaluation results: For different regions in the city, we analyze the collected elements As, Cd, Cr, Cu, Hg, Ni, Pb and Zn of soil by clustering analysis and principal component analysis, clustering results of soil elements showed in Fig. 11-15, principal component results of soil elements showed in Table 20. Through the comparison of cluster analysis and principal component analysis, Principal component analysis is the better one.

Obtained by clustering and principal component analysis:

DETERMINE SOURCE OF HEAVY METAL POLLUTION

Heavy metal pollutants in the soil move with the migration of soil water, heavy metal pollutants will be affected by the chemical and physical properties of the soil substrate and soil solute itself and other chemicals in the soil in the process of migration (Wu et al., 2009). The physical mechanism of soil solute transport mainly include convection, diffusion and mechanical dispersion. Due to the limited conditions in this study, select diffusion and gravity potential as its propagation characteristics, through the analysis of transmission characteristics, set up the partial differential model.

Analysis of transmission characteristics: Diffusion is caused by the thermal movement of solute molecules, the trend is migrate from high concentration to low concentration (Yang et al., 2007b), in order to achieve equal concentration, in a free aqueous solution, the solute diffusion follows the Fick’s first law, expressed with mathematical expression:

In the type, J_d is the solute in the free molecular diffusion of water flux, D₀ is the diffusion coefficient of solute molecules in free water, ∂c/∂z represents the solute concentration gradient, minus represents the direction of the concentration of solute molecules along the lower diffusion. In the soil solution, the soil solute diffusion also follow the Fick’s first law, the mathematical expression:

In the type, J_ds is the diffusion of solute molecules in the soil in flux, D_s represents the solute molecular diffusion coefficient in the soil, either in saturated soil or in unsaturated soils, D_s are less than D₀, generally will expressed solute diffusion coefficient in the soil of only as the function of water content and has nothing to do with the concentration of the solute, its mathematical expression is commonly used empirical expression:

In the type, τ called tortuosity factor, is a dimensionless physical quantity, according to the research of Wagenet, for most of the soil, the range of 0.3-0.7. Soil heavy metal pollution of gravitational potential is due to the effect of gravity, its value is equal to move the unit weight of heavy metal pollutants from the soil heavy metal contaminant reference point to the research height of the point, work value of the heavy metal pollutants in the soil (Zhu and Wei, 2011). Gravitational potential selected location coordinates with the size of the height and direction of the coordinate system, generally in the process of research, often selected the surface of the earth’s surface or groundwater as the origin of coordinates on and the direction of the axis can choose up or down according to the needs of calculation, if the research point height is z from the coordinates, the unit weight of soil heavy metal pollutants in the gravitational potential in unit length can be represented as:

Type in: ψ_g is the gravitational potential for heavy metal pollutants.

Establishment of the partial differential model: Assuming that the environment medium (such as soil) at any point P(x, y, z), this point as a unit around, the side length o branch is 2dx, 2dy, 2dz, (Fig. 3-5) P point velocity v(x, y, z) in three directions respectively is u_x, u_y, u_z (Lin, 2011), then pollution material flux in the three directions:

Now the molecular diffusion coefficient, turbulent diffusion coefficient and dispersion effects spread into a dispersion term, namely:

E_x, E_y, E_y respect X, Y, Z direction of the dispersion coefficient. Mass flux of dispersion effect in the X direction are as follows:

Quality flux of direction Y and Z:

Total mass flux of X, Y and Z:

Mass flux of the element get into from the X direction and get out:

Flux difference between the end faces (Qian, 1990):

Similarly, flux difference in Y and Z direction:

In the element, variation of pollutant in infinitesimal time:

According to the law of conservation of mass :

Put relationship between type (3-l) to (3-4) type into type (3-5), result in:

That is:

Considering the additional increase or decrease of the pollutant in the environment medium caused by physical (Zhu et al., 2012), chemical and biological, such as dispersing inflow and out flow along the river, the degradation, settlement, raise and adsorption of pollutants. Other pollution sources of pollutants incorporated from the air, absorb by vegetation and buildings of soil, the above mentioned can be combined as an additional source term S(x, y, z, c, t), is to:

Simplified of the basic equations: Type (22) is the basic equation of environment pollutants migration and transformation, vortex with environmental fluid equation, so we can simulate pollutants migration and transformation process in various environmental media. Although type (19) is complete in theory, But if vortex with the fluid motion equations, it is necessary to master the parameters and model coefficient data of each space in different time and the data contains the extremely complex relationship between the space and time, therefore, it is very difficult to solve this equation. In the actual work, often do all kinds of simplify according to the predicted target, hydrological characteristics, the condition of medium or pollutants.

River water quality simulation and prediction can use one-dimensional model.

Large rivers, estuaries, bays, shallow lake water quality simulation and prediction, can use two-dimensional model.

Solving the model: According to the data given in the question, this is a 3D model, solve the location of pollution sources on the 3D model, due to limited by conditions and data, we assume the biggest concentration place as pollution source, regardless of the element enrichment, we get the points of largest concentration according to the space distribution of heavy metal elements in the first question 1 of, the initial concentration of the biggest points out, According to the:

Under steady-state conditions of the point source ∂C/∂T = 0, we can ignore the longitudinal diffusion and flow in the Y direction and Z direction, that is E_x = 0, u_y = 0, u_z = 0, The type can be simplified as:

This type of analytical solutions for:

Evaluation results: Then programming the 319 sampling points (part data in appendix ) with MATLAB and solve the greatest concentration, obtained as follows: As(4400,7400,8), Cd(4500,1110,6), Cr(3400,5910,4), Cu(2400,3600), Hg(2600,2350,22), Ni (3200,5900,4), Pb (2000,3300,7), Zn (13600,9600,18), unit is m.

CONCLUSION

For the first question, the national evaluation standard is the second level, to analyze the error, now we use the first and third level to calculate (Zhou et al., 2012) and analyze the single pollution index in question 1 whether it’s reasonable, then, we analyze the error in the first area as an example, other areas are the same to the first. This process is presented in Table 9-12.

Analyze relative error the two single pollution index and the solution on the calculated values, we find the results are less than 0.1%, it can be concluded that the error of question 1 in an acceptable range.

For the second question, There are a lot of road in living area, on both sides of the road were the burning of leaded gasoline and dust of tire wear will increase elements Zn and Cr in soil and because of human activities are frequent in living areas, supplies discarded as waste in daily life can also lead to the increase of these three elements. The “three wastes” emissions and metal smelting lead to the high content of those elements. Due to the mountains far away from urban and industrial pollution sources, only part of the metal mine mining can lead to leakage of some metal elements. In trunk roads, the main reason of pollution is caused by motor vehicle exhaust emission, it also significantly cause the soil heavy metal pollution of road on both sides, car petrol, engine, tires, lubricating oil and gold plated parts, that all can burn or wear and tear and release Hg and Zn elements. In parks district, the use of fertilizers will be different degree of influence of heavy metal pollution in parks district and groundwater irrigation and plastic film can increase the soil content of elements Cu and Zn. It is not difficult to find different degree of pollution of element Cd in each region, this is because the Cd is mainly stored in the form of zinc cadmium sulfide ore, lead-zinc and copper in zinc, cadmium mainly comes from zinc ore, lead smelting, alloy, electroplating, chemical plant emissions of waste water, industrial solid waste dumps, garbage infiltration containing waste batteries, sewage sludge as fertilizer and excessive or inappropriate use of chemical fertilizer and pesticide and so on.

We analyze the stability according to the problems, when all the raw data elements density unchanged in addition to Ni, the Ni density change in (-4, -2, 0, 2, 4 ), then observe the changes of overall weight coefficient. Here we take living areas as an example.

When small changes in concentration of heavy metal elements, F1, F2 change is small, Comprehensive to get the stability of the model is very high, its composition coefficient does not do too big change through the change with heavy metal elements, so the scheme is suitable for the finishing of concentration.

Overall: Areas are more serious pollution when industrialization degree higher, urban area is higher than that of suburban and rural, surface is higher than that of underground, heavy metals accumulate more when the pollution time longer, the atmospheric media of soil heavy metal pollution in soil can be added strong, heavy metal content is higher when curing degree higher.

For the third question, There are advantages and disadvantages of models in question 3:

Advantages:

Disadvantages: Metal contaminants in soils is related to many factors, but we only selected two, at this point will produce deviation with the actual deviation.

In order to research evolution model of city geology environment better, through information query related to the city geological environment, geological environment is a kind of natural environment, the environment system composed of lithosphere, hydrosphere and atmosphere. Therefore, we should also collect heavy metal pollution elements in the lithosphere, atmosphere and hydrosphere with the distribution map of time, considering the effect of a number of other factors, we establish the N dimensional partial differential equations, in order to get evolution model. In order to determine the position of pollution source, First of all, we find the point according to the features of pollution source, then according to the 3D graphics of concentration distribution of each sampling points to solve maximum concentration, in solving the most value problems on a curved surface of 3D graphics, we will get the most value by establish and solve the partial differential equations.

ACKNOWLEDGMENTS

We would like to thank the anonymous reviewers for their valuable comments. This program is supported by Scientific Research Program Funded by Shaanxi Provincial Education Department (Program No. 2013JK1139) and Supported by China Postdoctoral Science Foundation (No. 2013M542370). The paper is partially supported by Natural Science Basic Research Plan in Shaanxi Province of China (Program No. 2012JM8047).

And this project is also supported by NSFC Grant (Program No. 11226173 and 61102163). This study is also supported by Scientific Research Program Funded by Xi’an University of Science and Technology (Program No. 201139). This job is also supported by Science and Technology Project of Xi’an (CX1262 ). Shaanxi Provincial Natural Science Basic Research Program (2009JQ1010) Shaanxi Provincial Department of Education Science Research Program (2010JK702) Xi’an University of Technology PhD project started (104-210901).

REFERENCES