A review on the application of the DRASTIC method in the assessment of groundwater vulnerability

Water is very important to living organisms. So definitely water is called the life of human beings and all living creatures. Water is also called the origin of life. Groundwater is the most precious source of fresh water on the earth among all water sources. Nearly half of all drinking water on the globe comes from groundwater, and 43 percent of all water used for irrigation in agriculture comes from groundwater.

Among fresh water sources about 30.10% of water is available as groundwater. So, groundwater is a globally important and valuable renewable resource for human life and the economic development of humans as well as whole countries. In developing countries like India, groundwater contamination has become a serious concern in the post-industrialization era (Vijayakumar et al. 2022). Basically, groundwater is water that exists underground in saturated zones below the land surface. It is divided into two parts: the unsaturated zone and saturated zone. The unsaturated zone lies above the water table and the saturated zone lies below the saturated zone. The body of rock and/or sediment which holds the groundwater is called an aquifer.

Groundwater systems are extremely dynamic, with water constantly moving downward due to the down-gradient from recharge to discharge sites. Nowadays for many different reasons groundwater is polluted and this creates a serious health hazard for the life of living creatures. Groundwater can be contaminated due to different types of human activities which include the disposal of sewage and leaching of pesticides and fertilizers, since due to increasing pesticides and fertilizers in agricultural fields, nitrate concentration may also increase. Pollution control and removal of pollutants from groundwater are a challenging and expensive task (Khodabakhshi et al. 2015), since after 1970 the exploitation of groundwater has also increased. According to the World Water Development Report 2020 published by the United Nations Educational, Scientific, and Cultural Organization (UNESCO), India is the world's greatest extractor of groundwater. Groundwater levels in important aquifers have dropped dramatically, lowering stream flows and degrading riparian and wetland ecosystems. Other critical regions for global food security, such as the North China Plains, Rajasthan, Gujarat, Punjab, Haryana, and Andhra Pradesh in India, and Eastern Australia, are facing catastrophic groundwater losses. Groundwater has received insufficient attention when it comes to pollution protection, despite its importance as the most critical component of sustainable development. Groundwater vulnerability assessment is the best way to safeguard groundwater resources in places that are prone to pollution or are already contaminated (National Research Council 1993). Groundwater levels are depleting, requiring long-term development planning to protect these crucial resources (Narendra & Rao 2006).

Groundwater is mostly affected by human activities. The velocity of groundwater below earth strata is much less and so the flow of contamination also reduces in the groundwater. Contamination flow is measured by the advection cohesion equation, which gives an idea about the velocities and flow of the contaminant with the groundwater. When the aquifer system is the fracture type of aquifer system then the velocities of the groundwater flow become higher and vice versa. Velocities of the groundwater flow also depend on the aquifer type, soil strata, topography, recharge rate etc. Velocities can be measured in kilometres per day in limestone with a well-developed solution or in some volcanic aquifers with extensive lava tubes or cooling fissures. Soil stratum is a critical component for contaminants that degrade over time, both for providing groundwater velocity and for controlling disease-causing microbes such as bacteria, viruses, and protozoa (Morris et al. 2003).

The primary risks to groundwater sustainability are the continual rise in water demand owing to population growth and increased agricultural productivity. To meet the water requirement, roughly 40% of India's land area is irrigated utilizing massive canal water delivery networks. Between 1950 and 2001, India's irrigated land increased from 23 Mha to 80 Mha (Chowdary et al. 2005), due to increase in the population food demand also increasing and due to the usage of fertilizers and pesticides also increasing day by day. From 1950 to 2001 the usage of chemical fertilizers and pesticides in India increased from 50 million tons to 210 million tons. So, seepage and deep percolation also reduce the quality of the groundwater.

India is a developing country. So, the urbanization and the industrialization of the country also affect groundwater quality. In many Indian towns and industrial clusters, groundwater pollution caused by industrial effluent and municipal waste in water bodies is a major concern. According to the Central Pollution Control Board's 1995 survey, groundwater at 22 sites across 16 Indian states has been contaminated by industrial effluent.

As shown in Figure 1, solid waste disposal on land surfaces could be another source of groundwater pollution. The toxic minerals of the solid waste can seep into the groundwater via soil and affect the groundwater quality. Also, in landfill the metal content also degrades the quality.

Despite threats from potentially harmful activities, groundwater is often surprisingly resilient, and water quality is generally high throughout broad areas of the globe. Sometimes groundwater naturally minimizes the impacts of contamination in specific aquifer types. Groundwater is difficult to contaminate, but once it is, it is tough to rehabilitate. The intrinsic qualities that govern the sensitivity of the water to be adversely affected by the applied contaminated load are referred to as groundwater vulnerability.

In addition, the cost of a failing local aquifer is very significant, and its loss may put a strain on other water supplies that are being considered as replacements. As a result, in the poor world, it is both costly and impractical. As a result, assessing which aquifer systems and ecosystems are the most vulnerable to degradation is crucial. Vulnerability assessment has proven to be an efficient and cost-effective tool for protecting groundwater resources from contamination (Machiwal et al. 2018). Studies of vulnerability provide insight into the effects of pollution loading, whether severe or light. Lobo-Ferreira & Cabral (1991) termed groundwater pollution vulnerability to be defined in accordance with the conclusions and recommendations of the International Conference on the Vulnerability of Soil and Groundwater Pollutants.

The notion is of assessing groundwater vulnerability in places that are more susceptible to pollution than others (Piscopo 2001). There are various sorts of vulnerability assessment methodologies that are utilized depending on the situation of the place. In many regions of the world, process-based methods, statistical methods, and overlay and index methods are commonly employed to assess groundwater risk.

Due to the low requirement of field data, the overlay and index approaches are the most important and accurate methods for determining vulnerability mapping. This strategy takes into account all of the variables that influence the groundwater vulnerability in a specific area. Different physical and hydrological parameters are used in this strategy to restrict the passage of contaminants that impair groundwater quality through the unsaturated zone until the water table is reached (Aller et al. 1987). The process of vulnerability mapping is very easy and less time-consuming as well as requiring less data, so these methods are most preferable. These methods are easy but in spite of this there are certain limitations of the overlay and index methods. Among them the first limitation is that these methods assume linear relationships between the vulnerability and the parameters while they show a non-linear superposition relation (Neukum & Azzam 2009). Secondly, all the weights of the parameters require the proper subjective judgement, which requires experience, and this directly affects the result of the vulnerability (Frind et al. 2006). Finally, the ratings' values are discredited, introducing an inaccuracy in the vulnerability mapping. There are many overlay and index methods, and among them DRASTIC (Aller et al. 1987), SINTACS (Civita & De Maio 1997), GOD (Foster 1998), AVI (Stempvoort et al. 1993), and PI (Goldscheider et al. 2000) are the commonly used and known methods.

In these methods of vulnerability mapping, predictions of the contaminant flow are done by a different technique. Basically, the data of the process-based methods are found out by different indirect techniques (Barbash & Resek 1996). These methods work based on the advection–dispersion methods for contaminant transport in groundwater. Process-based simulation models require analytical or numerical solutions to mathematical equations that demonstrate linked processes affecting contaminant transport. Basically, there are different approaches of the simulation methods, including saturated–unsaturated, single phase–multiple phases as well as one-dimensional–two-dimensional–three-dimensional models. Based on a one-dimensional advection–dispersion transport model that replicates vertical movement through a soil to the water table, Meeks & Dean (1990) created a leaching potential index. Soutter & Pannatier (1996) defined groundwater susceptibility as the ratio of cumulative pesticide flow reaching mean water table depth to total pesticide applied. Visual ModFlow is a physical-based model that forecasts water flow and pollution transfer under various hydrologic circumstances. This method produces superior results, but it has the drawback of requiring a large amount of input data, as well as more complicated computational processes and calibration issues (Iqbal et al. 2012).

Due to the complexities of groundwater vulnerability, there are different types of factors and parameters and so the statistical methods are suitable only in specific regions. They establish a link between explanatory variables and pollution concentrations (McLay et al. 2001). Basically, statistical methods use different parameters which work on the basis of probability and their results are also expressed as probability. Different probabilistic models are used in these types of methods. Multiple independent variables are used in these models, with the dependent variable being a pollutant concentration or a likelihood of contamination (Iqbal et al. 2012). A logistic regression model including soil-texture-related independent variables was proposed by Teso et al. (1996). The vulnerability assessment is done using dependent and independent variables in this method. In this method, the dependent variable is the variable, which is defined as the contamination status of soil sections (uncontaminated vs contaminated), and groundwater vulnerability is assessed by predicting the probability of a section containing a contaminated well (Teso et al. 1996). Worrall & Kolpin (2004) developed a groundwater contamination logistic regression model that combines chemical property fluctuation with land use, soil, and aquifer parameters. Compared with the overlay and index methods, the statistical method and the simulation models give better results and the only concern is about the input data or the availability of the data (Focazio et al. 2008).

As shown in Table 1, names of the procedures are usually made up of letters that represent the layer being processed; a list of acronyms is given as used in these approaches.

DRASTIC is a parametric vulnerability mapping technique that is well known and widely used. It was created in the United States as part of an EPA (Environmental Protection Agency) effort with the goal of assisting managers, planners, and administrators. Because of the minimal cost of deployment and the ease with which data may be collected, DRASTIC can be employed in a wide range of locations (Aller et al. 1987). Panagopoulos et al. (2006) claim that ‘the selection of many parameters and their interrelationship decreases the probability of ignoring some important parameters, restricts the effect of an incidental error in the calculation of a parameter and so enhances the statistical accuracy of the model’. This method is a well-known overlay and index method that evaluates intrinsic vulnerability based on the physical and hydrogeological properties of the aquifer (Aller et al. 1987). The DRASTIC approach is simple to use, and numerous researchers have used it to assess groundwater vulnerability all across the world. The DRASTIC model is used in many parts of the worlds such as Africa, Indonesia, India, Iran, Iraq, Jordan, Europe, and the USA. Basically, in this model seven parameters are used for the assessment of vulnerability. These model parameters are considered for the ratings and weights of the parameters and give the DRASTIC index value (Rosen 1994). D stands for depth of water, R for net recharge, A for aquifer, S for soil, T for topography, I for vadose zone impact, and C for conductivity in DRASTIC. Figure 2 denotes the different parameters which are used in the DRASTIC modelling.

The DRASTIC model is a valuable tool for assessing groundwater vulnerability and, more typically, for research purposes. Over time there has been work on the basis of modification and the including of different parameters which gave better results and accuracy in the assessment of vulnerability. Also, at the time of first introduction of the DRASTIC method and model, Aller et al. (1987) used different hypotheses or we can say assumptions, and some hypotheses among them are listed below.

• At the earth's surface, the pollutant is released (use of fertilizers, burning of coal and leaching of metals from coal-ash tailings etc.).

• The pollutant is transferred into the groundwater by rainfall.

• The pollutant moves simultaneously with water.

• The affected area should be at least 100 acres (0.4 square kilometres).

Based on the hypothesis of the DRASTIC model some limitations are also included in this modelling of the vulnerability assessment, some of which among them are included here.

• The vulnerability index is calculated using the DRASTIC model, which includes seven criteria, each with its own weight and rating value. Because the model does not account for the influence of specific area factors, some of the weights and rates are considered universally, and we may conclude that it is not accurate for a detailed examination of groundwater vulnerability assessment.

• There is no standard algorithm for testing and validating the model in the environment for aquifers.

• It is expressively designed for use in GIS software. The first time the DRASTIC model was used with GIS was by Merchant et al. (1987).

• Each parameter is weighted based on the standard value and the standard weights, so it is not so accurate for all areas related to the groundwater contamination.

• It is not so applicable for a study area of less than 100 acres.

The DRASTIC index is a dimensionless metric that assesses the vulnerability of groundwater to pollution. If the value of the DRASTIC index higher then we can say that the area is more affected by pollution and if the value of the DRASTIC index is lower then the area is less affected by pollution. Basically, the range of the DRASTIC index is between 70 and 200.

Here the methods of measurements of the DRASTIC parameters and how they affect the groundwater vulnerability assessment are explained below.

Depth of water table is an important parameter of the DRASTIC model which is defined as the distance from the ground surface to the water table (Aller et al. 1987). In the aquifer, it is the vertical distance from the ground surface to the water table, or top of the saturated zone. It is calculated using the country's topography and contour map as well as the study area's topography and contour map. In India the Central Ground Water Board (CGWB) and state ground water bodies give data of the depth of water table. The general concept is that if the depth of the water table is greater then the chances of pollution are much less and vice versa because if the depth of water table is decreased then the pollutants easily reach the groundwater and pollute it.

The quantity of rainwater that infiltrates into the ground and reaches the groundwater table is referred to as net recharge. It serves as a channel for solid and liquid pollutants to flow through. The groundwater recharge for the study is estimated using two methods: one provided by the Ground Water Resources Estimation Committee, which was established under the National Water Policy by the Ministry of Water Resources, Government of India, and the other provided by various researchers, in which the recharge is determined using runoff, evapotranspiration, and precipitation values.

Basically this method gives the broad concept of the net recharge. In this method net recharge is find from another four to five parameters, which are the following:

• a.

  Recharge due to rainfall

• b.

  Recharge due to canal seepage

• c.

  Recharge due to return flow from land under irrigation

• d.

  Recharge due to river

• e.

  Recharge due to tanks and ponds.

Water from irrigated fields seeps into the ground, causing fertilizers and pesticides to drain into the soil and eventually into the water table. The command areas under the branch canal, branch minor, distributary, sub-distributary, and minor are taken into account when calculating the recharge from irrigated land. From a general approach, the recharge from return flow is computed as 35 percent of the water given at the outlet for use in the field.

Recharge is also done by the river so it is an important parameter in finding out the net recharge value. Basically, recharge from the river is found from government reports on the river and river infiltration data or reports from bridge construction on the river. In general, according to the Ministry of the Water Resources, the recharge from the river is taken as 10 to 15 m/year for the major river and for the tributaries or the minor river it is taken as 5 to 10 m/year.

Using Equation (5) net recharge data are effectively found.

This method of finding the net recharge data has been applied by researchers like Arora et al. (1996), Shirazi et al. (2013), and Khakhar et al. (2017) in their respective study areas.

In this method, precipitation and evapotranspiration data is collected from the different websites like WRIS, NASA Earth data etc. Total runoff is also calculated according to the different methods which are given by different researchers and different authors. Among all methods of runoff calculation, the SCS-CN method is the most suitable method, as has been applied by Ali (2007).

Net recharge is the main factor in contaminants reaching the aquifer. Greater recharge indicates the possibility of contaminants reaching the groundwater and affecting the groundwater quality (Aller et al. 1987). River leakage or rainfall are the two main sources of recharging.

This is another crucial parameter for determining groundwater vulnerability. The material property or kind of aquifer present in the study area is referred to as aquifer media. The geological map of the basin and lithologs of the study area or bore logs in the study area are used to create the aquifer media. Aquifer is defined by Aller et al. (1987) as a geological formation that produces enough water for human consumption. Consolidated and unconsolidated rocks (such as sand, gravel, or limestone) that act as an aquifer are referred to as aquifer media. Controlling the route, path length, and movement of pollutants requires this parameter. Pollution sensitivity can be raised by a variety of factors, including sediment size, permeability, and attenuation capacity. Geological data can be used to create an aquifer media map.

Soil media is defined as the soil which is present in the uppermost part of the ground. The top, weathered portion of the unsaturated zone with substantial biological activity is represented by the soil media (Aller et al. 1987). It is the vadose zone's top layer, and its features are important in determining potential contamination, whereas increasing soil depth decreases infiltration. In general, a soil map can reflect pollutant infiltration rates. The type of clay, the grain size, and the shrink potential of clay are all important elements in determining the potential contamination of soil. Indeed, aquifer vulnerability is reduced when there is less clay, less shrinkage potential, and smaller particle size.

The slope of the terrain in the research area is known as topography. Topography directly or indirectly affects the runoff of contaminants. If the slope of the area is steep then the recharge of the water is decreased and contaminants are also reduced, so it is an important factor in assessing the groundwater vulnerability (Aller et al. 1987). GIS tools can determine slope and use a digital elevation model (DEM) to create topography maps.

The unsaturated zone, also known as the vadose zone, is described as the area between the aquifer and the soil cover. Vertical movement of the water in the unsaturated zone of the soil is the main cause of contaminant transport in groundwater and so this factor is very important for the groundwater vulnerability assessment. The unsaturated zone's characteristics can influence the media's attenuation qualities above the water table. The subsurface geology and lithological properties of drilling logs are also used to create a vadose zone map.

The ability of an aquifer to carry water is known as hydraulic conductivity. A hydraulic conductivity distribution map is created using the results of the pumping test and lithology (Aller et al. 1987). Contaminant migration is controlled by permeability. In India, hydraulic conductivity is found then drafted in the Central Ground Water Board reports as well as state groundwater reports by different government organizations like the Central Ground Water Board as well as state groundwater boards. Greater pollutant transport and distribution can be seen in areas with higher hydraulic conductivity. Pore gaps and fractures inside the aquifer also play a role.

In order to construct the aquifer vulnerability index, the DRASTIC technique assigns a precise weight and rate to each parameter. The groundwater vulnerability map is prepared and vulnerable areas are classified according to Table 2 (Aller et al. 1987) and Table 3 (Engel et al. 1996), and then the vulnerability classes are defined as indicated in Table 4, after assigning the weights as shown in Table 5 (Aller et al. 1987). Ranges of the different parameters are indicated in Table 6 (Aller et al. 1987). Aller et al. (1987) suggest different colour-coding for the vulnerability map preparation and this is included in Table 7.

Vulnerability mapping using the DRASTIC technique has a number of shortcomings, despite its ease of use and popularity. The biggest disadvantage is its subjective nature, which raises questions about the selection of specific factors and the exclusion of others (Panagopoulos et al. 2006). DRASTIC has been criticized for the following reasons:

• By giving equal rates and weights to parameters, it ignores the effect of regional characteristics.

• A standard validation approach is not used in this procedure.

• The parameters were chosen based on qualitative assessment rather than quantitative research.

As a result, a number of researchers have worked to improve the DRASTIC model so that a more accurate vulnerability assessment may be obtained. Some investigations, for example, compared the DRASTIC index with chemical or pollutant data, but found no link in most situations. McLay et al. (2001) stated that models like DRASTIC, which include a land management index, could be effective for predicting locations where groundwater monitoring should be more intensive. It has also been suggested that further research into the association between nitrate leaching measures from various land use activities and groundwater nitrate concentrations below these activities is required (McLay et al. 2001). In a GIS setting, Panagopoulos et al. (2006) revised the factor weightings and ratings of all DRASTIC parameters using simple statistical and geostatistical methodologies. There are lots of different parameters which affect groundwater quality; hydraulic conductivity is one of them but it less affects the quality. In some research work, hydraulic conductivity was removed from the DRASTIC equation for its improvement, and land use was added as a new DRASTIC component. When compared with the conventional technique, the correlation coefficient between the groundwater vulnerability index and nitrate concentrations was dramatically boosted, rising by 33% (Panagopoulos et al. 2006). The standard method with the addition of land-use was improved to assess the specific vulnerability using the DRASTIC-LU model (Wei et al. 2021). The DRASTIC scores for groundwater nitrate concentration were published by Leone et al. (2009). The scores were divided into two groups: lower vulnerability (60–80) with no nitrate content and higher vulnerability (110–155) with a wide range of nitrate concentrations ranging from around 0 to over 160 mg/L. The obligation to undertake interpolation of restricted field data, which involves interpolation error, is one of the main sources of this imbalance. As a result, DRASTIC results should be modified to fit location and pollutant specifications. For this purpose, several methods for developing and modifying the DRASTIC algorithm have been suggested.

In recent years, sensitivity analysis has been used in a number of studies to change the weight values specified in the DRASTIC approach to increase assessment accuracy and dependability (Abdullah et al. 2016; Barzegar et al. 2016). For non-point-source pollution, Neshat et al. (2014a, 2014b) used SPSA to modify the weights of the DRASTIC model, concluding that the modified DRASTIC model performed better than the standard method. For the Kerman Plain in Iran, the correlation between the vulnerability index and nitrate concentration was 82 percent after correction, compared with 44 percent before modification, according to the regression coefficients. Pacheco et al. (2015) used the modified DRASTIC technique to evaluate sensitivity analysis on 26 aquifer systems in Portugal. As a result, vulnerability indices were on average 20% lower than the initial DRASTIC values (Pacheco et al. 2015). The removal of the impact of the vadose zone, the depth to water, the hydraulic conductivity, and the net recharge created a considerable difference in the mapping vulnerability, according to Ouedraogo et al. (2016), who utilized sensitivity analysis in aquifer systems in Africa. They also revealed that the data on nitrate concentrations are favourably associated with the intrinsic vulnerability index, with a correlation coefficient of 0.94. Furthermore, Abdullah et al. (2016) used sensitivity analysis in the Halabja Said Sadiq Basin in Iraq's northeast and found that the modified DRASTIC was far superior to the standard model. The modified DRASTIC index and nitrate concentration have a good relationship, with Pearson correlation factors of 0.69, 0.57, and 0.72 for modified rate (using nitrate concentration), weight (sensitivity analysis), and combined rate–weight techniques, respectively.

The overlay index models are less constrained by data shortages and computational challenges than the other approaches of groundwater vulnerability assessment statistical-inference-based models, but they have a number of conceptual shortcomings. Sometimes there are many different parameters that are used for the calculation of the indices, which gives an idea of vulnerability, but the availability of the data is very difficult and if some data are not available than they are assumed according to the field conditions and this creates errors in the actual results. Also, weighting and validation are very difficult for the parameters.

So according to the whole research on groundwater vulnerability assessment and within it especially research on the DRASTIC model, we can say that the DRASTIC model is a very accurate and easy as well as reliable method for groundwater vulnerability assessment. Modifications are essential in both the approach and the model. The finest results come from modification according to field investigations. In arid, semi-arid, basaltic, agricultural, and landfill zones, the modified DRASTIC approach outperforms the traditional DRASTIC method. The DRASTIC approach benefits greatly from sensitivity analysis. It identifies which parameter has the most impact on groundwater vulnerability. Sensitivity analysis demonstrates the discrepancies between theoretical and effective weights of DRASTIC parameters. Also the parameter calculation method is the main important part of the research and so based on the field conditions it is very important to select appropriate methods for calculation of the parameters. Also the model is validated using the basic parameters of the drinking water, and in agricultural areas nitrate concentration and carbonate concentration are also used for validation purposes.

This may improve the prospects of reducing the gap between policy enactment and implementation, which is frequently a barrier to attaining long-term water usage sustainability.

All relevant data are included in the paper or its Supplementary Information.