DRASTICAI, a New Index for Groundwater Vulnerability Assessment—A Portuguese Case Study

Abstract

Groundwater vulnerability assessment has become a useful tool for groundwater pollution prevention. Groundwater vulnerability maps provide useful data for protecting groundwater resources. Identification of agricultural patterns is an important issue for optimized land management. The watershed of the Tagus River is the backbone of this study. Naturtejo UNESCO Global Geopark, in the central interior of Portugal, corresponds to a rural area. Intensive agricultural practices showed an increasing trend in the last decades. The method that is most used internationally to assess vulnerability is the DRASTIC index. In this study, the DRASTICAI index is introduced. A new attribute—anthropogenic influence—is added here. Five levels of increasing vulnerability, from low to high, can be recognized here. The municipality of Idanha-a-Nova is most affected by intensive agricultural activities, showing spatial patterns of DRASTICAI with a clear influence of anthropogenic activities, with high needs for groundwater abstraction. A robust assessment of groundwater quality has a key role. Climate change scenarios and water scarcity are important issues in the coming years, and particularly in the studied area. Therefore, optimized groundwater management is essential to consider in policy-making strategies.

1. Introduction

Nowadays, groundwater vulnerability assessment has become a useful tool for groundwater pollution prevention. Groundwater vulnerability maps provide useful data for the protection of groundwater resources and serve as a tool for water management, as they allow changes in agricultural practices to be considered and, thus, optimization of land use [1,2,3,4,5,6,7].

Groundwater vulnerability assessment is a recognized complex process, since natural conditions and land-use practices vary substantially around the world. Thus, making the associated risk of pollution dependent on specific factors and the available data, two different types of groundwater vulnerability are widely recognized in literature: intrinsic and extrinsic, or specific, vulnerabilities [8]. The first is a function of hydrogeological factors, and the second is defined by adding attributes that define the potential anthropogenic influence. The most widely used method for assessing intrinsic vulnerability is the DRASTIC index [9], as it is easy to calculate and requires a minimum amount of data. The DRASTIC model was developed by the U.S. Environmental Protection Agency (EPA) to assess groundwater contamination potential for the entire United States. The DRASTIC Pesticide uses the same parameters but a different weighting scheme to stress the importance of agricultural activities [9,10,11].

In the last decades, many authors have proposed different extensions of the DRASTIC index, e.g., by testing different approaches of weighting techniques [1]; introducing a DRASTIC calibration technique for the specific contamination by nitrates [2]; or introducing other approaches, such as the modified SINTACS method based on a NO₃^- and SO₄^2- dataset [7]. Another approach is the susceptibility index (SI). This index provides the assessment of the specific vertical vulnerability to pollution mainly due to agricultural activities and, therefore, with high influence of the presence of nitrate. Its computation is based on the DRASTIC, considering four of the original parameters: depth to water (D), annual efficient recharge (R), aquifer media (A), and topography (T), and adding a fifth and new parameter, land use [12].

In the present study, the main objective is to introduce an empirical approach for a modified DRASTIC to assess the vulnerability of a groundwater system in a region where anthropogenic activities are very diverse, from mining activities to intensive agriculture, and where available information is scarce. In other words, how can one approach groundwater management in an inherently complex region that is subject to diverse anthropogenic influences and where scarce information is available? How can one create a possible decision support tool?

The study area is in the center of the Portuguese interior, close to the Spanish border, and belongs to the district, Castelo Branco (Figure 1a). This area largely overlaps with the area of Naturtejo UNESCO Global Geopark, created in 2006, an area of outstanding geological heritage that includes 176 geosites, 9 of which have hydrogeological significance, such as springs, hot springs, fountains, or medicinal water sources [13]. Geographically located within the Central Iberian Zone of the Iberian Massif, it lies at the junction between the Central Cordillera (to the north) and the Alentejo peneplain (to the south).

The area under study lies within the hydrographic watershed of the Tagus River, one of the largest hydrologic systems of Western Europe, which traverses almost the entire Iberian Peninsula. This watershed includes rivers such as Zêzere, Erges, Ocreza, Ponsul, or Tagus (which forms the southern border of the district), and streams such as Meimoa, Aravil, or Sertã. The altitude ranges from 31 m, in the lower Tagus river, to 1220 m, in the Gardunha Mountain (Figure 1b). The clime is the Mediterranean, with a hot and dry summer and with mean precipitation ranging from 500 mm to 800 mm, and a mean temperature between 22.5 °C and 25 °C [14].

The oldest bedrock comprises mainly Neoproterozoic and Lower Cambrian, with impermeable metasediments (shales and metagraywackes) from the Beiras Group [15,16] forming wide flattened areas. The catchment is truncated by the Armorican quartzite formation of the Lower Ordovician age [17,18]. Quartzites characterized by fractures and fissures, with minor intercalations of pelite, are the most common lithology. These rocks are found in thick quartzite beds in the ridges of Penha Garcia, Monforte da Beira, Muradal, Unhais-o-Velho ou Talhadas, with the main direction NW–SE directly related to Variscan deformation. Variscan and Ordovician magmatic bodies intruded the bedrock of the Beiras Group in Castelo Branco [19], Penamacor-Monsanto [20], Fundão [21], Covilhã [22], and Oledo [23] (Figure 2a). Continental sediments of the Cenozoic age with variable permeability (arkoses, conglomerates, and breccias) are found in the basins of the Sarzedas and Lower Tagus [24].

The main objective of this study was to evaluate the vulnerability of groundwater to pollution, more specifically, the watershed of the Tagus River in Naturtejo Geopark (Figure 1), using a modified DRASTIC method in a GIS environment. The modified DRASTIC index (DRASTICAI) was created by adding a new attribute called “anthropogenic influence”.

2. Materials and Methods

The DRASTIC model is based on the concept of the hydrogeologic environment, which is defined as a composite description of all the major geologic and hydrologic factors that influence and control groundwater movement into, throughout, and out of the area. The acronym represents seven hydrogeologic parameters considered in the assessment process: depth to groundwater (D), aquifer recharge (R), aquifer media (A), soil media (S), topography (T), impact of vadose zone (I), and hydraulic conductivity (C) of the aquifer. Each DRASTIC parameter is evaluated concerning the others to determine the relative importance of each parameter, and is then given a relative weight ranging from 1 to 5. The most important parameters are given a weight of 5, while the least important are given a weight of 1 [9] (

Table 1. Assigned weights for DRASTIC parameters (Source: [9]) and DRASTICAI.

Parameters	Weight
Depth	5
Recharge	4
Aquifer media	3
Soil media	2
Topography	1
Impact of vadose zone	5
Hydraulic conductivity	3
Anthropogenic influence	5

and

Table 2. Data type, source, format, and respective references of the DRASTIC/DRASTICAI parameters.

Parameters	Data Type	Source	Format	Reference
Depth	Interpolation	SNIRH and IGEM	Table	[25,26]
Recharge	Location Map	ARH Tejo	Raster	[27,28]
Aquifer media	Geology Map	Global Lithological Map database v1.1	Polygon	[29]
Soil media	Soil Map	FAO	Polygon	[30]
Topography	Elevation Map	Global Digital Elevation Model	Raster	[31]
Impact of vadose zone	Geology Map	Global Lithological Map database v1.1	Polygon	[29]
Hydraulic conductivity	Driven Soil Map	FAO + 3D Hydraulic DB	Raster	[30,32,33]
Anthropogenic influence	Land Use 2018	COS 2018	Polygon	[34]
Study Area	Administration	CAOP 2020	Polygon	[35]

). The purpose of the DRASTIC index implies the multiplication of each factor weight by its category rating (

Table 3. DRASTIC (Source: [9]) and DRASTICAI parameters.

DRASTIC/DRASTICAI	Range	Rating
Depth to groundwater (m)	30.5	1
	22.9–30.5	2
	15.2–22.9	3
	9.1–15.2	5
	4.6–9.1	7
Recharge rate (net) (mm)	<50	1
	50–100	3
	100–150	6
	150–300	8
	>300	9
Aquifer media	Metamorphic rocks Sedimentary Rocks	3
	Acid Volcanic Rocks Intermediate Volcanic Rocks Basic Volcanic Rocks	3
	Acid Plutonic Rocks	4
	Limestone	6
Soil media	Coarse	6
	Medium	9
	Null/No information	10
Topography (slope) (%)	>18	1
	12–18	3
	6–12	5
	2–6	9
	<2	10
Impact of the vadose zone	Metamorphic Rocks Sedimentary Rocks	4
	Acid Volcanic Rocks Intermediate Volcanic Rocks Basic Volcanic Rocks	4
	Acid Plutonic Rocks	4
	Limestone	6
Hydraulic conductivity of the aquifer (m/d)	<4.1	1
	12.2–28.5	4
	>28.5	6
Anthropogenic influence	Water body and waste land	1
	Forest and shrub land	2
	Built up with very low density Agriculture	5
	Built up with low density	7
	Built up with medium density	8
	Built up with high density	9

), where the ratings (r) reflect the importance of the classes, and the weights (w) reflect the importance of the parameter as follows:

$$\mathrm{DRASTIC}=\left({\mathrm{D}}_{\mathrm{r}}\times {\mathrm{D}}_{\mathrm{w}}\right)+\left({\mathrm{R}}_{\mathrm{r}}\times {\mathrm{R}}_{\mathrm{w}}\right)+\left({\mathrm{A}}_{\mathrm{r}}\times {\mathrm{A}}_{\mathrm{w}}\right)+\left({\mathrm{S}}_{\mathrm{r}}\times {\mathrm{S}}_{\mathrm{w}}\right)+({\mathrm{T}}_{\mathrm{r}}\times {\mathrm{T}}_{\mathrm{w}})+\left({\mathrm{I}}_{\mathrm{r}}\times {\mathrm{I}}_{\mathrm{w}}\right)+\left({\mathrm{C}}_{\mathrm{r}}\times {\mathrm{C}}_{\mathrm{w}}\right)$$

(1)

In this study, an additional parameter was added to the DRASTIC model to represent groundwater vulnerability more accurately by including anthropogenic influence (AI). This new parameter is rated depending on land use and was assigned a weighting value of 5, and the modified DRASTIC index, DRASTICAI, was calculated using the following equation:

$$\mathrm{DRASTICAI}=({\mathrm{D}}_{\mathrm{r}}\times {\mathrm{D}}_{\mathrm{w}})+\left({\mathrm{R}}_{\mathrm{r}}\times {\mathrm{R}}_{\mathrm{w}}\right)+({\mathrm{A}}_{\mathrm{r}}\times {\mathrm{A}}_{\mathrm{w}})+({\mathrm{S}}_{\mathrm{r}}\times {\mathrm{S}}_{\mathrm{w}})+({\mathrm{T}}_{\mathrm{r}}\times {\mathrm{T}}_{\mathrm{w}})+({\mathrm{I}}_{\mathrm{r}}\times {\mathrm{I}}_{\mathrm{w}})+({\mathrm{C}}_{\mathrm{r}}\times {\mathrm{C}}_{\mathrm{w}})+({\mathrm{AI}}_{\mathrm{r}}\times {\mathrm{AI}}_{\mathrm{w}})$$

(2)

where D is rated depending on the depth of groundwater, R is rated depending on the recharge rate (net), A is the aquifer media, S depends on the kind of soil, T is rated in relation to the terrain slope (slope), I is rated depending on the hydraulic conductivity of the aquifer on the lithology of the vadose zone, C is the conductivity (hydraulic) of the aquifer, and AI is anthropogenic influence (Table 1).

For the mention parameters, different data sources were used (Table 2), and the geographical information projected using the EPGS (European Petroleum Survey Group—Geodetic Parameter Dataset) code projection 4326 | WGS 84 reference system. Furthermore, the maps were clipped using the survey area polygon as geometry. ArcGIS 10.3 software was used for computation purposes.

It is worth stressing that the predominant lithology corresponds to metamorphic rocks and mixed sedimentary rocks (in the global lithological map). Concerning soils, the most common are the dystric regosol, haplic aluvisol, and dystric cambisol (in the FAO harmonized soil map). Finally, the map representing the slope classes was computed using the elevation, driven by the SRTM elevation map. The computation was performed in degrees, in the Spatial Analyst tool. The land-use map aggregates the characteristics of the land use, namely agricultural, agro–forest, and forest (Figure 2).

The piezometers’ network data for Spain and Portugal, between 1986 and 2020, was compiled [25,26], and a point dataset of 23 locations, 13 locations in Portugal and 10 in Spain, was gathered. The inverse distance weighted (IDW) interpolator was used to produce the raster data file containing the depth information, in meters. The recharge layer was gathered from [27], and contains the annual net recharge (mm/year) within a 250 m cell resolution. Concerning the hydraulic conductivity, a layer driven from SoilGrids included in EU—SoilHydroGrids ver1.0 [32] was downloaded for 2 m depth from saturated hydraulic conductivity cm day⁻¹ (KS), in a 250 m cell size.

Furthermore, aquifer media, soil media, impact of the vadose zone, and anthropogenic influence raster layers were constructed in a 250 m cell resolution, using the ArcMap’s polygon to raster tool (Figure S1).

The final DRASTIC and DRASTICAI outputs were computed through map algebra (Figure 3).

For the further calculation of the indices DRASTIC and DRASTICAI, eight thematic maps were compiled as shown in Figure 3.

3. Results

The index DRASTIC was constructed using Equation (1). The obtained results were classified into different classes corresponding to low (54–119, 87.9%), intermediate (119–149, 10.4%), and high vulnerability (149, 1.7%; Figure 4a). The DRASTICAI index was compiled based on the DRASTIC index using Equation (2). The DRASTICAI index considers the reclassification of the obtained results into different classes corresponding to low vulnerability (53–119, 65.8%), intermediate vulnerability (119–149, 28.4%), and high vulnerability (149, 5.8%; Figure 4b).

The risk map DRASTIC highlights two different levels of vulnerability: low (87.9% of the total area) and moderate (10.4% of the total area) (Figure 4a). The northern areas, Idanha-a-Nova and Castelo Branco, show low vulnerability (105–119), while the rest of the area is moderately vulnerable (120–138). However, when analyzing the DRASTICAI map, it is possible to detect significant changes in the spatial patterns of vulnerability (Figure 4b). Three levels of increasing vulnerability, from low to high, can be identified. The municipality of Idanha-a-Nova is the most affected by anthropogenic influence due to intensive agricultural activities.

4. Discussion

Land use parameters can significantly affect groundwater quality. Anthropogenic factors affecting water quality in rural areas differ from those in urban areas. In rural areas, these include agricultural practices, such as the use of fertilizers, herbicides, and pesticides; siltation of streams through erosion; nutrient loading of water bodies; runoff from degraded forest land; and livestock production. Changes in land cover, and especially in land use patterns, play an important role in water flow and, consequently, in water quality [36,37]. The studied area is mainly intensively used for agriculture, which raises important issues related to groundwater monitoring and quality assessment (Figure 3). Indeed, groundwater resources in areas of high agricultural use generally have higher nitrate levels, leading to water quality degradation and high non-carcinogenic risk to human health [38,39,40,41,42,43]. Therefore, increasing attention is being paid worldwide to the identification of effective approaches aimed at prevention and management solutions for groundwater pollution assessment and water safety [40].

The index DRASTIC shows a predominantly low/moderate vulnerability (87.9%) and, thus, a reassuring geological and hydrogeological framework that provides natural protection against groundwater pollution. However, when the anthropogenic influence is considered using land-use parameters, the calculated index (DRASTICAI) shows a moderate–high vulnerability, increasing from 12.1% to 34.2%, indicating a high risk due to human influence. Here, the more vulnerable areas coincide with agricultural and urban areas.

The algebraic subtraction between the maps DRASTIC and DRASTICAI shows an important contribution of anthropogenic influence (Figure 5). It is possible to mention an increasing effect from low/moderate to high vulnerability in the whole studied area, but specifically in the municipality of Idanha-a-Nova.

In conclusion, it is important to emphasize that an understanding of field conditions associated with human activities provides a robust assessment of groundwater quality and plays a fundamental role in decision making and establishing effective management strategies for groundwater monitoring and abstraction.

Future developments include extending the methods studied to other neighboring areas to assess groundwater quality at a regional scale [37,42,43].

5. Conclusions

Groundwater is an essential and strategic resource for social and human development, especially in regions with water scarcity and cyclical droughts under climate change scenarios. This study aimed to assess the vulnerability of groundwater to pollution in the Naturtejo Geopark area using a modified DRASTIC index—DRASTICAI. This new index was constructed by adding a new attribute called anthropogenic influence. The spatial patterns of DRASTICAI show a clear influence of anthropogenic activities, mainly in the community of Idanha-a-Nova, which is an area of intensive olive tree production with high levels of water abstraction. Therefore, an urgent need to develop efficient management strategies for the sustainable use and protection of groundwater resources is needed. For this purpose, it is imperative to have appropriate tools for effective monitoring and assessment of groundwater resources and abstraction control at the watershed scale. Effective management is critical for the sustainable management of groundwater resources.

Water is one of the most strategic resources in the world. Portugal has significant groundwater resources that can be of strategic importance for the expected dry years.