Assessment and Estimating Groundwater Vulnerability to Pollution Using a Modified DRASTIC and GODS Models (Case Study: Malayer Plain of Iran)

This study deals with the intrinsic vulnerability of groundwater reservoirs to pollution, by the use of two models DRASTIC and GODS, this study is done by taking samples from 17 water resources of Malayer plain Aquifer area of southern Hamedan Province, Iran. 30 physicochemical parameters and heavy metals have been studied and vulnerability of this aquifer to the nitrate concentration, was determined. The study showed that results from DRASTIC were better than GODS in Assessment and Estimating groundwater vulnerability to pollution, also DRASTIC model has been corrected, and compared the ability of these two models in vulnerability zoning has been evaluated. According to high correlation between DRASTIC index and nitrate concentration, ranking and weighting of nitrate pollutant is inserted in the DRASTIC equation, and zoning map of DRASTIC method has been calibrated by nitrate concentration. By this method, vulnerability zoning is determined between very low to very high, which shows the increase of DRASTIC index by nitrate concentration. DRASTIC parameters uncertainty has affected the zoning results in this method, but its calibration with nitrate concentration, gives more accurate vulnerability results.


Introduction
Water plays a vital role in every biological society in the globe.The socioeconomic development of a region predominantly depends on the availability of good quality water.Groundwater vulnerability assessment could be defined as the degree of assimilation capacity of the area to the contaminant from surrounding surface above the aquifer [1].Groundwater pollution is a universal problem.Sometimes the nature is the cause of water quality decrease, but in the most cases, human is the main cause of water pollution [2].The concept of aquifer vulnerability was first introduced by Marget.Vulnerability assessment has been conducted as an essential part of protection strategies for land use planning and groundwater protection zoning [3].In fact, the term ''vulnerability of groundwater to contamination was first used by [4].This concept refers to the sensitivity of an aquifer to deterioration due to an external action and is based on the assumption that physical environment may provide some degrees of protection to groundwater against contaminants entering the subsurface zone.Consequently, some land areas are more vulnerable to groundwater contamination than others [5,6].Groundwater pollution vulnerability mapping is an important tool to identify areas that are more sensitive to contamination.GIS is an effective technique for the zone mapping and risk assessment on environmental health problems.GIS can be useful for taking quick decisions as graphical representation would be easy to take a policy decision by the makers [7].GIS techniques have been becoming the most commonly used platform for assessment of groundwater vulnerability [1].Several methods have been developed, throughout the world, to assess groundwater vulnerability.The NRC (1993) has classified these methods into three major classes: (1) the process-based simulation models; (2) the statistical methods; and (3) the overlay and index methods [8].In hydrogeology, vulnerability assessment typically describes the susceptibility of a particular aquifer to contamination that can reduce the groundwater quality [9].The National Research Council (1993) defined it as the tendency or likelihood for contaminants to reach a specified position in the groundwater system after introduction at some location above the uppermost aquifer.Two terms are used to describe groundwater vulnerability: intrinsic and specific.Intrinsic vulnerability is the natural susceptibility to contamination based on the physical characteristics of the environment while specific vulnerability is defined as an accounting for the transport properties of a particular contaminant or group of contaminants through the subsurface [10].The first term refers to the intrinsic property of groundwater system to human or natural impacts.The most leading model of the intrinsic vulnerability is DRASTIC index [11].
The evaluation of vulnerability is a mean to gather complex hydrogeological data in a way that can be used by nonspecialist people such as decision-makers.Vulnerability assessment of groundwater aquifers provides a basis for initially protective measures for important groundwater resources and will normally be the first step in a groundwater pollution hazard assessment and quality, when it interest [3].In recent years, vulnerability assessment of groundwater aquifers considered as an essential part for putting suitable plans to protect groundwater aquifers around the world [9].

Study Area
About 90% of water demand across Malayer Plain is supplied by groundwater, deep and half-deep wells, aqueducts and fountains [12].Annual groundwater resources discharge from deep and semi-deep wells, ghanats, and springs, is 495 million cubic meter.As can be seen in the Figure 1; the studied area is located between 48 o , 25' to 49 o , 05' longitudes east and 34 o , 05' to 34 o , 25' latitude north.And is a part of Malayer plain in the Hamedan province in Iran, approximately 922 km 2 .Malayer height is 1760 m above sea level.The main river in this plain named Harm-Abad (Ab Malayer), which originates from Sarband Arak heights, after immerging with the branch streams from Borujerd heights (Kalan branch), flows to Malayer city.Average yearly rainfall is 313 mm and average temperature degree is 13o C. Meteoric rainfalls are caused by Mediterranean flows, mainly.The humidity and raining are caused by west weather fronts too (about 8 mounts (from December to July) is influenced by the system) [12].Malayer Plain's alluvium comprises the erosion of schist, slate, phyllite, granite, granodiorite, limestone and sandstone and deposition of the destructive materials carried in valleys, where the size of particles tapers off from the brinks to the center.Thin gravelled sandy and silty layers are spread in the great part of the plain.Malayer plain is almost flat and its gradient is 0-6%.Surface water, penetrates the aquifer and the rest of it flows in the Harm-Abad seasonal river, and at the end enters the Nahavand limit.Groundwater direction is as the same as the surface water, which flows to centre from all main entrances.Metamorphic, igneous and sedimentary rocks across the region play an unimportant role in feeding the aquifer.However, they are significant in terms of forming the aquifer's rocky bed.The Plain's aquifer is composed of rubble, sand and clay with a thickness of 25 m in the brinks of the plain, expanding to 150 meters in the southwestern areas.According to the observed number of wells, the depth-to-water table in Malayer Plain ranged from 6 to maximum 30 meters during April 2007 (Figure 1).The overall direction of groundwater flow within Malayer Plain begins from East, North, West and South ends in the Center.The hydraulic gradient in the southeast and northwest brinks is higher than the center and outflow of the plain.The average transmissivity in Malayer Plain ranges from 1051 m 2 /day to 8035 m 2 /day.Over the 12-year span (1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006), the aquifer experienced a 15.48-meter drop in ground water.The summary of results indicated that the aquifer's balance in Malayer Plain is negative due to factors such as excess exploitation [13].
Malayer plain is located in the semi-pressurized zone of Sanandaj-Sirjan.This zone is south west margins of central Iran flat, actually.The oldest petrologic unit outcropped in Malayer plain is related to Triassic and the newest sediments are quaternary structures.Malayer plain alluvium is formed by schist, slate, phyllite, granite, granodiorite, limestone and sandstone of plain margins erosions and sedimentation of these sediments in the depression of region.Particles sizes decrease from margin to centre.
With a total of 217 zones registered, Malayer is one of the important plains in terms of agricultural and industrial growth.The quality downgrade in groundwater of Malayer Plain is mainly the outcome of human activities involving nitrate wastewater due to urban development and the sprawl of a few pollutant industries such as refining and packaging of raisins, dairy production and livestock slaughterhouses, which are released into the surrounding environment and then leaked into the aquifer.In addition, the contamination of groundwater has intensified by the excessive use of nitrate and animal fertilizers, pesticides, return water flow in agriculture, water passage through solid waste left from cowshed and poultry farms in rural areas and husbandry complexes especially in other areas in plain as well as inappropriate sanitation systems and Malayer's leachate from the urban landfill in plain.The contamination of drinking water refers to altered physical, chemical and biological properties of water, making it harmful for direct human consumption [14].The quality downgrade of groundwater takes place slowly and is difficult to identify or even control.It might render the water unusable and dangerous for decades [15].Therefore, it is crucial to preserve groundwater sources so as to guarantee health and reduce the risk of diseases caused by consumption of contaminated water.
Groundwater quality decrease in this plain is mainly due to nitrate sewage produced by some industries, such as raisin purification and packing, dairy products, and slaughters complexes, which widespread in the plain with no attention to environmental problems and diffuse to groundwater.
In addition, animal and chemical manure irregular use; accumulation of animal waste in villages' margins, animals' husbandry complexes and Malayer landfills, increase the probability of groundwater vulnerability in the plain.So, there are enough reasons to assess the vulnerability of this plain to pollutants.Vulnerability is not an absolute characteristic, but is a relative one, which can be propounded in the polluted area.The prognosis of groundwater vulnerability is similar to the weather forecasting [16].This not only gives more data in order to decision-making, but also helps to detect and purify high risk and polluted areas.It will prevent more pollutant diffusion to groundwater and saves aquifer internal characteristic.
If possible, pollutants must be far from water resources, and if pollution occurred, it must be removed or compensated.And if complete removal is not feasible, at least the preferences must be distinct [17].
Hot spots, rainfall decrease and over-using causing the quality of groundwater getting worse, getting more information in order to increase groundwater protection strategies, information documentation in order to decision making in the area of groundwater pollutants, selecting special regions for rural and urban sewage management, groundwater quality protection [16] and saving public health [17] attract attention to Malayer plain aquifer vulnerability zoning.

Methods
In this issue aquifer vulnerability of Malayer plain has been assessed by the use of DRASTIC and GODS methods.DRASTIC method was developed by EPA [18] in 1980, which estimates the groundwater vulnerability, empirically and the basis of hydro geologic data.
DRASTIC is a groundwater quality model for evaluating the pollution potential of large areas using the hydro geologic settings of the region [18].Data needed for this method, are depth to water table, net recharge, aquifer media, soil media, topography slope, impact of vadose zone and hydraulic conductivity of aquifer formed material.Four dependent parameters (D, A, S, I) are gained from drilling and T parameter from topographic maps.There's an estimated value for R and C is calculated according to the aquifer characteristic.
Vulnerability index is obtained from the equation 1.The higher the DRASTIC index is the greater relative vulnerability potential.r and w refers to rating and assigned weighting, respectively.For areas with a lack of information about the sub surfaces and the groundwater, GODS method can be used [3].GODS has four parameters.Groundwater existence (inexistent = 0 existent = 1), overlaying petrology (0.4-1), depth to groundwater (0-1) and soil type (0.5-1).Vulnerability index in this method is determined by multiplying these four parameters.
The risk of biological contamination of Malayer Aquifer was determined by collecting samples from a total of 17 water sources and analysing the samples under laboratory standards from December 1, 2007 to December 23, 2007 (Figure 2).There were 30 physical-chemical parameters such as EC, TDS, Alkalinity, Hardness, Turbidity involving heavy metal samples such as Ca, Mg, Na, K, Cl, SO4, Si, NH3, NO3, NO2 Cr, Co, Cu, Pb, Ni, Zn and Fe.specified by Tehran Water and Sewage Company.The samples were collected from zones in vicinity of industrial plants, poultry, livestock slaughter facilities and landfill sites across Malayer (Table 1, 2).

Results and Discussion
The results of chemical analysis on 17 samples collected from the study area based on the WHO standards (2000) and the Iranian standards indicated that groundwater across Malayer Plain mainly fall under facies Ca -Mg-HCO3 (71%) and Na -HCO3 (29%) and Bicarbonate calcic basic type as well as fresh and drinkable water according to the Piper and Stiff diagrams.In this regard, 94% of the fresh class, while 6% are in the brackish class.Moreover, 29% of the samples were averagely hard, 65% were hard and 6% were very hard.The location of the samples on the Schuler chart indicates that 94% of the samples were categorized as good quality, 4% were acceptable and 2% were average for drinking.In terms of agricultural potential, 76% of the samples fell under C3S1 while 24% fell under C2S1.In terms of industrial application, 76% of the samples inclined toward sedimentation and 88% were slightly corrosive.The majority of the samples collected from the Plain's groundwater did not contain any suspended particles at constant temperature of 25 °C.Hence, they are preferred over surface water or general water supply and distribution.However, the water turbidity is extremely high in Kartilabad Poultry Kartilabad (about 54%).In other areas, water turbidity was fine.As an indicator of biological contamination in Malayer Plain, the nitrate concentration varies from 9 to 61.3 mg/litre.Nitrate concentrations on the 17 sampling sites are shown in the Figure 3.The nitrate concentrations were over 20 mg/lit in about 65% of the samples.On the other hand, the increase in nitrate concentration at the wastewater outflow of raisin refining and packing plants and slaughterhouses is remarkable.Due to extreme blending with the wastewater outflow, the nitrate concentration is 61.3 mg/litre in the samples collected from the water well in Shirinabad Livestock Slaughterhouse.The maximum depth of the wells in the region is 90 meters.Hence, the impact of nitrate is limited to the depth of 90 meters.Since the bedrock of the alluvial plain is composed of schist, there is no chance of feeding.Therefore, the concentration of nitrate contaminants escalates over time by an increase in ammonium oxidation due to the leaching of inorganic fertilizers applied to farms.
The variations in nitrate concentration are normally proportionate to the hydrogeological regime.In fact, the nitrate flowing into the groundwater progresses like a plume in the direction of water towards greater depths as its concentration is curtailed.Based on the results of the tests, there is no contamination with heavy metals such as lead, copper, cadmium, chromium, nickel and zinc across Malayer Plain.
The iron concentration exceeded the standards at 0.376 mg/lit only in the water sample collected from Kartilabad Poultry.The intensity of variability is generally higher in zinc, manganese, iron, sulphate and chlorine than the average of ions, indicating erratic dispersion of data around the mean values.
In the case of nickel and lead ions, the intensity of variability is zero.Moreover, the intensity of variability for calcium, magnesium, sodium and potassium ions is higher than the average suggesting there are extreme alterations and dispersion of data in the constituent elements of different samples.Coexisting with humans, coliform bacteria decompose intestinal substances in the body.They tend to be highly resistant under temperature shift and variations in pH and eh.Moreover, they provide an indicator, measured based on E. coli.According to research conducted by the Health Department of Malayer, the E. coli test result in water well across the outskirts has been positive in recent years due to inadequate preservation measures taken to reinforce of walls and maintaining a distance from the sanitation network.However, this has not been resolved and the water contains little microbial load.

Figure 3. Nitrate-to-water table contour map across Malayer Plain
By the use of DRASTIC method and seven parameters gained in the studied area between 80 to 143.Final vulnerability mapping (Figure 4) shows that, the vulnerability potential of the studied region -Malayer plain-with the area of 922 km 2 , is categorized in to very low, low and medium, which contains 51, 43, 6% of the area respectively.
Central parts have low to medium vulnerability, but northern, southern and margins have very low vulnerability.For DRASTIC method calibration, nitrate concentration of 17 water resources has been shown on the map.Table 3. shows final DRASTIC index calculations.The greatest and the smallest vulnerability potential are due to D parameter and soil type, respectively.There is more changes index in depth to water table and impact of vadose zone, which indicates the important role of these parameters in vulnerability evaluation in this area.Because of samples biasing, which were taken of pollutant resources, nitrate concentration doesn't express the exact values in the aquifer and doesn't correspond with the vulnerability maps.Therefore, it is necessary to change weighting of seven parameters of DRASTIC, to gain better correspondence.Although DRASTIC method gets good relative information, in order to have a more accurate evaluation, not only must the exploiter drillings be increased, but also nitrate concentration data must be entered in the DRASTIC linear equation, and vulnerability must be assessed by changing in weights of parameters and finally calibrated by nitrate concentration, to increase the validity of this method.There's a high correlation between DRASTIC index and average nitrate concentration (r = 0.86).After some steps of changing the weighting of parameters of DARSTIC and using the results of nitrate concentration in the samples and rating and weighting of nitrate pollutants, DI gained 85-190 (heuristically).Thus, from the total area of 922 km 2 in 28% equals to 258 km 2 , in 15% equals to 138 km 2 , 16% equals 148 km 2 , 38% equals to 350 km 2 and 3% equals to 28 km 2 , the vulnerability assessed very low, low, medium, high and very high, respectively.This estimation can be accepted because of expression the vulnerability zoning and increase of DI index related to nitrate concentration.Parameter deletion sensitivity analysis, expresses the sensitivity of vulnerability map to delete one or more parameters (Equations 2 and 3) [2]. (2) In this equations Si= sensitivity analysis, Vi= vulnerability index in the i-th cell, n= number of parameters used in cell vulnerability determining, Vxi= vulnerability index sign of i-th cell without any notice to xi, n= number of parameters used in the sensitivity analysis, VXi= changes index due to parameter deletion, Vi= vulnerability index gained from seven parameters deletion and Vxi= vulnerability index with no parameter deletion.Uni-parameters sensitivity analysis evaluates the effect of each parameter on the vulnerability index.This sensitivity analysis must be done to comparison the actual and theoretic weighting of parameters (equation 4) [2].Sensitivity analysis of DRASTIC model must be used in order to determine the effectiveness of each parameter on the aquifer vulnerability.
In this equation Wxi= effective weighting of parameter, Xwi= x parameter weighting in the I area and Xri= numeric value of DRASTIC to xi parameter.The most changes in vulnerability index in the parameter deletion sensitivity analysis of DRASTIC method, occurs after deletion the groundwater depth parameter.The most effective weighting average in the uni-parameter sensitivity analysis relates to groundwater depth (23.8%).Vulnerability index in the Malayer plain, by the use of GODS method and four defined parameters, ranges from 0 to 0.34.Ultimate vulnerability mapping is shown in Figure 5.In this way three groups of vulnerability very low, low and medium were obtained, which were 53%, 44% and 3% respectively.Central parts have medium vulnerability potential and northern and southern part and margins relate to very low vulnerability.In order to calibrate the GODS method.

The methods are compared in what follows:
 Vulnerability zoning by the use of DRASTIC method is more accurate than GODS method.
 Although vulnerability results are relatively equal in both methods, the vulnerability zoning using DRASTIC method is in accordance with pollutants sites, more.Vulnerability increases by the increasing of pollutants, gradually.
 There may be more faults for both methods in the margins, because the data collected from the margins are less than the centre, in number.
 Without calibration the DRASTIC method by the nitrate concentration, the vulnerability was very low.After calibration, due to high correlation between DRASTIC index and nitrate, vulnerability index increased.Also, the zoning results correlated with nitrate concentration, relatively.

Conclusion
The results of chemical analysis on water samples according to the WHO and Iranian standards revealed that nitrate concentration as an indicator of biological contamination in Malayer Plain varies from 9 to 61.3 mg/lit.The nitrate concentrations were over 20 mg/lit in about 65% of the samples.Due to extreme blending with the wastewater outflow, the nitrate concentration is 61.3 mg/lit in the samples collected from the water well in Shirinabad Livestock Slaughterhouse.The increase in nitrate concentration at the wastewater outflow of raisin refining and packing plants and slaughterhouses is remarkable.In terms of heavy metals including lead, copper, cadmium, chromium, nickel and zinc, the aquifer in Malayer is not contaminated.The iron concentration exceeded the standards at 0.376 mg/lit only in the water sample collected from Kartilabad Poultry.
Vulnerability zoning by the use of DRASTIC method is more accurate than GODS method.Although vulnerability results are relatively equal in both methods, the vulnerability zoning using DRASTIC method is in accordance with pollutants sites, more.Vulnerability increases by the increasing of pollutants, gradually.
The most pollution potential relates to D parameter with the average of 20.88 m, and the list one relates to soil type.Changes indices in the parameter of depth to water table and vadose zone are more than other parameters and show the importance of these two parameters.In the parameter deletion sensitivity analysis of DRASTIC method, more changes occur with depth to water table deletion.In uni-parameters sensitivity analysis depth to water table has the most effective waiting in the parameters, nearly 23.8%.DRASTIC method is more accurate in Malayer plain, and in this method sensitivity analysis is feasible.Uncertainties of some data are neutralized with certainty of other data (especially for I, T and C parameters which do not exist in the GODS method).DRASTIC method uses more parameters and rating and weighting in DRASTIC method are more precise than GODS method.
In order to reduce the environmental impact of contaminants such as nitrate, it is recommended that further investigation be conducted on how and why the relatively shallow wells are contaminated [19].Since the urban waste has been amassed in Malayer's landfill, it is essential to insulate the site and prevent any leachate to water flows.Alternatively, the landfills can be relocated.Moreover, it is recommended that the outflow sewage at raisins refines and packing plants, slaughterhouses and poultry be continually sampled and examined.Furthermore, it is necessary to construct a refinery in highly pollutant factories.It is suggested that accumulation of biomass be prevented during rainy seasons, while restricting the application of pesticides with nitrate and livestock origins in the plain.

Figure 2 .
Figure 2. The scope of Malayer Plain and sample collection

Figure 4 .
Figure 4. Aquifer vulnerability zoning in the Malayer plain by the use of DRASTIC 100

Figure 5 .
Figure 5. Aquifer vulnerability zoning in the Malayer plain by the use of GODS