MODELING OF RELATIONSHIPS BETWEEN TRAFFIC PARAMETERS AND VEHICULAR LEAD AND CADMIUM DISTRIBUTION IN URBAN ROADSIDE SOILS

Saeed Samani1., Gholamreza Shiran2., Amir Taebi3., Majid Afyuni4., Mohammad Mehdi Abtahi3 and Hamidreza Pourzamani5 1Zenderood Environmental Research Center, Isfahan, 8174673461, Iran 2Civil Engineering Department, Faculty of Engineering, University of Isfahan, Isfahan 8174673441, Iran 3Department of Civil Engineering, Isfahan University of Technology, Isfahan 8415683111, Iran 4Department of Agriculture, Isfahan University of Technology, Isfahan 8415683111, Iran 5Environment Research Center, Isfahan University of Medical Sciences, Isfahan 8174673461, Iran


INTRODUCTION
The disease burden by a population and how that burden is distributed across sub population is important information for defining strategies to improve population health. Global found of the diseases burden from Lead have been introduced for 14 different reasons and different age groups worldwide in developed countries. When Leaded gasoline has usually been phased out, the highest environmental exposure to Lead generally affects children of lower income families. In countries where Lead is still used, Leaded gasoline will likely be a major contributor to exposure directly through air, or indirectly through food and dust of Leaded ceramics (WHO). The evidence for carcinogenicity is stronger in animals, its carcinogenicity in human has recently been questioned [1].
By increasing social needs for transportation and communication, day to day vehicular usage has been increased. There will be more than 950 million vehicles by the year 2020 in the world [2]. Automobiles are known to be the main sources of producing heavy metal pollutions in cities. These pollutions enter the environment from automobile bodies and exhaust as particles. Vehicular pollutions are distributed in roadsides by traffic flow and cause air, water and soil pollution [3,4].
Lead and Cadmium are parts of vehicular pollutants. These pollutants, because of their poisoning characteristics and because of being dangerous for nature and human body such as blood and nerve system, have been researched by previous researchers [5,6].
Lead usually enters the urban environment by using Leaded petrol. Although in recent years, Lead has been omitted from fuel and its entering the environment has been decreased, its pervious effects still persist. Cadmium exists in automobile tires. It partially departs from vehicles and disperses with traffic flow and enters the urban environment [5,7,8]. Moreover different parameters affect vehicular pollution in environment, which in large scale contains conditions of road, traffic and environment.
Road parameters related to road physics and design are: length, slope, road canyon and average height, average aspect ratio (i.e. the ratio of height of the building to width of the road), number of lanes, and the type of green cover and its height [9][10][11]. Traffic flow as one of the most important factors in dispersing vehicular pollution contains speed, traffic volume, traffic congestion and driver behavior in start time and movement, fleet composition ,traffic management method and traffic sign arrangement in roads [12][13][14]. Environmental parameters contain climatic conditions like temperature, wind direction, speed and also roadside soil characteristics [10,12,13].
Past research works have measured heavy metal concentrations in soil, as well as their impacts on distribution trend. Rahmani [15] through soil sampling from different distances in four highway sides in Iran illustrated that sharp decreases exit in lead concentrations as distance increases from the main road. Ward et al. [16] in Auckland New Zealand measured Lead and Cadmium concentrations in surface roadside soils in 17 sites. A gradual decrease of heavy metal concentrations has shown in the study as distance increases from roads. Lead content in soil has shown to have a definite relationship with traffic. However, this relationship has been weak for cadmium concentrations. Carlosena et al. [17] in Lacrona, Spain studied vehicular traffic effect on heavy metal concentrations and concluded that high load of Lead concentrations and low load of Cd, Cu and Zn concentration were results of traffic distributions. Garsia and Milan [18] measured Cd, Cu, Zn and Pb concentrations at two distances in eight sites experiencing different traffic flow in urban highways of Gipuzequa, Spain. Their results showed Pb and Cd concentration levels changed as distance from highway varied.
This study of vehicular pollutions and their impacts on roadside environment and also impacts of different traffic parameters on distribution amount and trend of pollution is a requirement when urban architecture, design and planning and urban traffic management are of concerns. This study investigates the roadside soil and gutter soil pollution because of Lead and Cadmium. During the course of this research, the impacts of different traffic parameters on the distribution of pollution will be studied and finally the trend of Lead and Cadmium distributions as function of distance and traffic parameters will be analyzed and then modeled.

MATERIALS AND METHODS
Thirteen sites along various roads of varying traffic volumes in the built up areas of Isfahan were selected. Table ( Figure  (2) presents typical cross sections of roads along sites number 10 to 13. Sampling was carried out using linear soil sampling method. In doing so, samples were taken at a distance of 25 cm from a line parallel to road axis and then mixed. Standard Methods /Techniques [19] were adapted as the basis of sample extraction and measurement of Pb and Cd concentration. These include Atomic Absorption Spectroscopy (AAS) in the sites and laboratories. Data and information were collected on the total traffic volume, daily traffic volume, total (accumulated) traffic volume, total traffic volume to directional road width, total traffic volume to total capacity, vehicular speed and street aspect ratio. Their definitions and calculation methods are as follows: Average annual daily traffic (AADT) is defined as the yearly traffic volume divided by 365. AADT has been used by many researchers in order to compare concentrations in various roads [15][16][17][18]. Average daily traffic in year and yearly traffic volume in studied roads were defined from existing data of transportation research center of Sharif University of Technology [20]. Road ages also were defined from their construction documents. On the basis of the vehicular possession coefficients, Isfahan population and age of each road, annual traffic volume in different years for roads were estimated. Information on the construction year of road extraction was extracted from the data base of Isfahan Highway Department and Municipality of Isfahan. Total traffic volume for each road was estimated up to the years 2013. The total traffic volume in all studied roads was about 52 to 520 million vehicles. Total traffic volume was assumed as an effective parameter on extinction of cumulative Lead concentration on the roads under study.

Fig 1 Map of Isfahan and sampling sites
Total capacity is the maximum number of vehicles that can pass through a point on a road per unit time in a year in both or all directions under the prevailing condition of road and the traffic. Total traffic volume is the demand or number of vehicles that pass through a given point on a road per unit time in both directions. The ratio of (Volume/Capacity) or (V/C) shows the extent to which volume of a road is approaches capacity. It also presents the degree of saturation of a road. Such a ratio has been sought to be another influencing parameter in the amount vehicular heavy metal pollution [10].
Total capacity values were calculated by EMME/2 software. This parameter shows relative value of using roads in their usage age. In each road, by measuring the speed of two types of dominant vehicles in different hours of day, mean speed was defined.
The vehicular mean speed on the vicinity of study sites varied between 30 and 76 kilometers per hour (km.h -1 ). Previous studies have confirmed the influence of street aspect ratio on traffic generated air pollutants distribution. Aspect ratio is defined as the mean height of building along streets to the mean width of streets. Linaritakis (1987) has modeled CO pollution concentrations using data on streets of central London. Models developed based on traffic load was setting and geometry of streets located in central basin districts where high building are set on both sides of roads are known as "Street Canyon models" [21]. Zero height denotes to area where there is no building along streets.

RESULTS
Statistical characteristics of total Lead and Cadmium in street soils up to 50 meters from kerb and background soil are presented in Table (2).
In Figure (3) relative frequent distribution (probability) of total Lead in roadside soil has been presented. The distribution curve of Lead in background soil is also shown in the Figure (  The statistical difference of Lead and Cadmium mean concentrations are investigated by t-student test in significant level of 5% (reliable level 95%) in Table (3).         (5) and (6). Table (5) shows traffic data, used for modeling which were total traffic volume, daily traffic volume, total traffic volume to directional road width, total traffic volume to total capacity, speed, and aspect ratio.  (7) and (8), respectively.

DISCUSSION
Background Lead and Cadmium concentrations are those measured in remote areas of Isfahan regions which were lesser exposed by anthropogenic activities. As shown in Table (2), mean concentrations of metals in the soils of streets and up to distance of 50 m are greater than the background values. Lead distribution in roadside surface soil is more dispersed when compared with lead distribution in background soil. As shown in this Figure 4, cadmium distribution in surface has an outward distribution from its background distribution. (3) and (4) shows that Lead is outer than Cadmium from their related background distribution. As a result of planting trees in gutter soil, it is required to use agricultural soil standard for its health measurement. The British and Australian standards were used for Pb is 100 and for Cd is land 5 mg.kg -1 , respectively. As seen in Table (2), the mean concentration of Pb in gutter soil up to 50 m is greater than maximum allowable concentration. In the case of Cd, mean concentration exceeded the British standard. Such comparison reveals that more attention should be paid to roadside plantation and soil pollution.

Comparison of Figures
One-tailed t-paired student test in significant level of 5 percent was employed to determine the significant effect of distance on metal distribution. One tailed test was selected because it is expected that in nearer distances to streets, concentrations were high. In each site, in ten distance, metal concentrations were measured, but comparison of all is not necessary, and just comparison of concentrations in distances 1, 8, 15 and 50 m will be made.
As  16285 | P a g e concentration on that site. These values are always between 0 to 1. Using these curves and regarding their amounts, it is more possible to study and compare distribution trend of concentrations.
Figures (5) and (6) show that distribution trends are in wave manner. Wave summits are the same in all sites. This phenomenon relates to size of produced particles by vehicles and happens in the result of traffic flow speed and rotary flow of air near the roads.
Each wave moves to the extent that its movement energy dissipates and concentrations reach background concentrations or any obstacles near the road. Moreover heavy metal distribution is that concentrations in 5 m are less than 1 m from road ( Figures 5 and 6). This phenomenon refers (is related to) to road architecture and curb heights ( Figure 2) and rotary flow of air. Comparison of Figures (5) and (6) shows that wave manner distribution trend for Pb concentrations is more apparent than Cd concentrations which relates to differences of their particle sizes and weights. Plotting Pb and Cd concentrations versus distances for sites 10 to 13 show that distributions decrease with the increment in distance. In addition to this, the logarithmic models are the best. As Table  (6) presents, results of comparison at significant level of 5 percent shows that site 10 versus 11 and 12 with regard to its calculated t and critical t in two tailed test has a significant difference. Also site 13 versus sites 11 and 12 has t cal more than t cri . Results for Cd show that site 10 versus sites 11 and 13 have t cal more than t cri and site 11 versus site 13 has t cal more than t cri . So, it can be concluded that differences between sites are significant and it is clear that this difference relates to various traffic parameters in these sites.
Regression analysis suggests traffic related parameters have influenced heavy metal concentrations and their distribution in road side soil. From fitting different regression models to Pb and Cd concentration data in gutters of all sites versus each one of traffic parameters statute in Table (5), metal concentrations have a good correlation with the total traffic volume. Linear and exponential models explain the relationship between Lead and Cadmium and the explanatory variables well, respectively. Analysis shows a weak correlation between the Pb and Cd concentrations and the explanatory variable V/C (ratio of total traffic volume to capacity). Moreover, we found a weak relationship between daily traffic volume and concentrations of Pb and Cd, which is the opposite of what was found by other researchers [16][17][18]. Insignificant correlation between the metal concentration and daily traffic volume can be attributed to the short term nature of daily traffic volume when compared to the cumulative traffic volume. No significant correlation was found between the metal concentration and vehicles speed and street aspect ratio. One may conclude that parameters such as street aspect ratio, vehicles speed and daily traffic volume influence the distribution pattern of gaseous pollutants. Because of this, in this research, a significant relationship was seen between total traffic volume and metal concentration in soil.
The cumulative nature of Pb and Cd concentrations in soil may also explain both operational and prolonged vehicular volume.
To model the changes in metal pollution concentrations of Pb and Cd, various mathematical expressions were tested. Such models examined the relationship between the dependent variables Pb and Cd and explanation variables of total traffic volumes and the distance between the curbside and points where samples were taken.
Models (1) and (2) Where: C Pb is total Pb concentration as milligram to kilogram dried soil, C Cd is total Cd concentration as milligram to kilogram dried soil, V is total traffic volume as number of vehicles, and D is distances as meter.
As pointed, total traffic volume is not used as independent variable in calculating the metal concentrations in other researches. Therefore, to verify the results, a statistical method standardized residual was used. Standardized residuals are calculated from dividing differences of predicted values and observed values to mean squares of residuals; and they have a mean equal to zero and a standard deviation equal to one and when placed between 2 and -2 show that the selected model is suitable for regression. (7) and (8), almost all the residuals placed between 2 and -2 and scattered data has no specific pattern for Pb and Cd. Therefore, regression models presented in model (1) and (2) have a suitable validation.

As shown in Figures
These are taken from samples using mean concentrations of background samples. One way probability has been employed in this statistical comparison. This is because it is expected that measured Pb and Cd concentrations are greater than background values. In t-student test, the difference between samples will be significant if calculated t (t cal ) is more than critical t (t cri ). As presented in Table (3), Pb concentrations in all locations have significant difference at significant level 5% with background concentrations. Based on this, total Pb concentrations in sample soil are affected by pollution sources and because of their being in cities, they are anthropogenic. Cd concentrations in gutter soil have significant difference with background values and have affected by external factors. With regard to the fact that these samples are related to road side soils, it can be concluded that traffic flow has caused the roadside to be polluted. This finding is in line with the work of other researchers' [15][16][17][18].

CONCLUSIONS
From the review of literature on the related area of heavy metal concentrations and statistical analyses undertaken, the following conclusions can be drawn:  All of the sites specially gutter soils and soils within 50 m from curbs had Pb and Cd mean concentrations higher than background level and standards of agricultural soil in many countries.  Statistically, there was a significant difference between Pb and Cd mean concentrations in distances within 50 m of road curb and background level. This difference is related to the effects of traffic and transportation parameters on roadside soil pollution.  The significant discrepancy observed between Pb and Cd mean concentrations in different sites can be attributed to the difference shown in their traffic parameters.  Pb and Cd concentrations in road side soil exhibited more relationship with total traffic volume compared to other traffic parameters.  Traffic parameters with longer effect such as total traffic volume, total (accumulated) traffic volume, total traffic volume to directional road width, and total traffic volume to total capacity, showed a greater impact on relationship of soil heavy metal concentrations with traffic parameter.  By increasing distance, Pb and Cd concentrations in roadside soil decrease logarithmically.  Pb and Cd concentrations of gutter soils increase linearly and exponentially with increment of total traffic volume, respectively.  The models (models 1 and 2) developed in this research are capable of predicting Pb and Cd concentrations in soil within a distance of 50 m as a function of distance from road side and total traffic volume.