Groundwater vulnerability assessment in Savar upazila of Dhaka district, Bangladesh — A GIS-based DRASTIC modeling

Highlights

• DRASTIC model utilized to identify vulnerable aquifers to industrial pollution in a rapidly growing urban city in Bangladesh.

• Depth to aquifer, Topography and Soil media are the most influential attributes to groundwater vulnerability for this study.

• Aquifers currently have low vulnerability, but further contamination risk cannot be ruled out, and is indeed rather likely.

• Information on aquifer vulnerability may facilitate precautionary measures for sustainable groundwater management.

Abstract

Revealing the status of groundwater contamination before it reaches a critical level is crucial to effective groundwater management. With this in mind, the research primarily aims to identify vulnerable aquifers of Savar upazila (sub-district), a highly industrialized zone of Dhaka district, by analyzing existing hydrological attributes through a GIS-based DRASTIC model. Seven DRASTIC attributes viz., depth to aquifer, net recharge, aquifer media, soil media, topography, impact of vadose zone, and hydraulic conductivity of aquifer were created and integrated using a weighted-overlay method in a GIS interface. The resultant vulnerability map reveals that about 34% of the study area is in low-vulnerability zones, 45% is moderately vulnerable and 21% highly vulnerable to groundwater contamination. The results also reveal that aquifers contiguous to floodplain areas are highly vulnerable whereas those adjacent to terrace areas have low vulnerability. The model outcomes affirm that the depth to aquifer, topography and soil media had the greatest link to vulnerability. A positive correlation was also noticed, while validating the final DRASTIC map, between the vulnerability classes and the three groundwater quality parameters-electrical conductivity, nitrate and chromium concentrations. Though the current levels of contamination are within permissible limits, the risk of further contamination cannot be ruled out, and is indeed rather likely. Information on the current status of contamination of groundwater might act as an early warning for the responsible authorities to take prudent measures to prevent further stress on this invaluable resource.

Introduction

Groundwater resources are becoming vulnerable to contamination due to the increasing stress of anthropogenic activities such as agriculture, urbanization and industrialization (Dimitriou and Moussoulis, 2011; Khan et al., 2011a; Shirazi et al., 2010). Human activities commonly threaten groundwater quality, leading to temporary or permanent loss of the resource, and therefore incurring a significant cost for the remediation and/or removal of harmful contaminants from the water prior to use.

The concept of groundwater vulnerability to contamination was introduced by Margat (1968), in which groundwater vulnerability was defined as the opposite to natural barriers against contamination. The National Research Council of United States of America defined groundwater vulnerability in a simplified way as “the tendency or likelihood for contaminants to reach a specified position in the groundwater system after getting introduced at some location above the uppermost aquifer”(Council, 1993). In a similar vein, Ball et al. (2004) argued that groundwater vulnerability is the tendency and likelihood of contaminants to reach the aquifer after being introduced at the ground surface.

Groundwater vulnerability depends on the hydrogeological characteristics of an area which largely control the residence time of water that fell as rain, infiltrated the soil, reached the water table and flowed into the aquifer (Prior et al., 2003). This is categorized as either intrinsic or specific vulnerability (Frind et al., 2006; Gogu and Dassargues, 2000; Stigter et al., 2006). Intrinsic vulnerability is the vulnerability controlled by the geological and hydrogeological attributes of the aquifer. Whereas, specific vulnerability is defined as the vulnerability associated with a specific source of contamination, its characteristics and connection to the various components of intrinsic vulnerability (Gogu and Dassargues, 2000; Vrba and Zaporozec, 1994).

Considering the aquifer characteristics, researchers have adopted various methods and techniques to estimate groundwater vulnerability (Javadi et al., 2011). All the current methods can be grouped under three major categories, namely: (1) process-based models; (2) statistical models; and, (3) GIS overlay and index (Council, 1993; Nobre et al., 2007; Tesoriero et al., 1998). The choice of the most appropriate method to determine vulnerability depends profoundly on the purpose and scope of a particular study, scale of the task, data availability and, most importantly, the user's time and cost (Liggett and Talwar, 2009).

Compared to the other two methods, process-based simulation models are relatively complex and involve a substantial amount of data input as well as considerable computing power (Almasri, 2008; Iqbal et al., 2012). Process-based models are restricted to predicting physical, chemical and biological processes distributing contaminants (spatially and temporally) and in assessing groundwater contamination. In contrast, statistical techniques are a simpler alternative approach for predicting groundwater contamination. This method involves correlating various measured parameters with the concentration of contaminants (Mclay et al., 2001). Monitoring changes in chemical concentration over time is the major disadvantage of this method. However, this method is applicable in areas where the groundwater contamination is governed by similar physical factors (e.g., hydrogeological attributes). The simplest and most widely used method, the GIS overlay and index method, in combination considers the physical factors relevant to potential contamination, such as soil media, geological material, aquifer depth, recharge rate, and environmental factors (Aller, 1985; Almasri, 2008; Gogu and Dassargues, 2000).

There is a growing concern in the scientific community regarding the assessment of groundwater contamination. With the increasing availability of spatial data and the use of GIS, groundwater vulnerability mapping has become a popular tool for groundwater resource protection and management (Jha et al., 2007). Of all the available GIS-based mapping techniques, DRASTIC — acronym depth to aquifer (D), net recharge (R), aquifer media (A), soil media (S), topography (T), impact of vadose zone (I) and hydraulic conductivity (C) — is the most popular, and was initially proposed by Aller et al. (1987a). Inspired by the DRASTIC method, but with a different nomenclature, Civita (1994) proposed the SINTACS method, the acronym coming from depth to the water table (S), net infiltration (I), unsaturated zone (N), soil media (T), aquifer media (A), hydraulic conductivity (C) and slope (S). SEEPAGE, standing for system for early evaluation of pollution potential of agricultural groundwater environments, is another mapping approach, proposed by Navulur and Engel (1998). The SEEPAGE method uses depth to the water table, soil topography, soil depth, impact of vadose zone, aquifer material characteristics and attenuation potential. The EPIK method proposed by Doerfliger and Zwahlen (1997) was developed for karst aquifers, and includes Epikarst development (E), protective cover potency (P), infiltration stipulation (I) and karst mesh growth (K). With fewer parameters than the previous approaches, the GOD method is considered to be the quick and an easy technique for pollution mapping. GOD, initially proposed by Foster (1987), stands for groundwater occurrence, overall lithology of the aquifer and depth to the groundwater table.

All these techniques aim to assess the intrinsic vulnerability of aquifers, with the main factors the depth to the aquifer (water table), the soil properties and the characteristics of the saturated and unsaturated areas (Gogu and Dassargues, 2000). Numerous projects based on DRASTIC method, the most popular method, have been undertaken in many parts of the world. Those in the United States (Rupert, 2001); Sweden (Rosen, 1994); South Korea (Kim and Hamm, 1999); South Africa (Lynch et al., 1997); Portugal (Lobo Ferreira and Oliveira, 1993); North America (Ducci, 1999; Fagnan et al., 1998; Fritch et al., 2000; Navulur and Engel, 1998; Stark et al., 1999) and China, Italy, and Algeria (Dai et al., 2001; Lobo Ferreira and Oliveira, 1997; Lynch et al., 1994; Menani, 2001; Napolitano and Fabbri, 1996; Shahid, 2000) are prominent.

In Bangladesh, groundwater research has mostly been confined to hydrogeological and hydrogeochemical aspects, such as groundwater potential, groundwater quality, hydrogeological modeling and heavy-metal pollutants (Zahid, 2003; Khan et al., 2011a, 2011b; Salam and Alam, 2014; Zakir et al., 2006). However, very little has been attempted in the context of vulnerability due to urbanization and industrialization. Seddique and Matin (2013) and Jamil (2010) both targeted groundwater vulnerability assessment in two rapidly urbanized and densely populated cities Narayanganj and Tongi, respectively. However, the authors did not use conventional DRASTIC parameters in their groundwater vulnerability studies rather they considered only thickness of the upper clay, depth to the water level and land use pattern. Haque et al. (2018) on the other hand, conducted a research on vulnerability of groundwater pollution from mining activities in an open-pit coal mine in Bangladesh.

To begin to bridge this research gap, this study was undertaken in Savar upazila in the Dhaka district of Bangladesh where around 1500 industries, together with a major export processing zone, the Dhaka Export Processing Zone (DEPZ), are currently in operation. Among the industries, chemicals, ceramics, leather processing, and pharmaceutical, textile, dyeing and washing activities are the major sources of contamination. About 195 brick factories are also operating in the vicinity. According to the pollution categories of the Department of Environment of Bangladesh, all types of industries (Green, Orange-A, Orange-B and Red) are present in the study area (DOE, 1997). This leads to the assumption that groundwater resources in the study area could be under threat from industrial effluent. Thus, the primary aim of this study was to identify vulnerable aquifers using the GIS-based DRASTIC model with the intention of providing contemporary information on groundwater contamination which could be beneficial for sustainable management of groundwater resource.