ASSESSMENT OF GROUNDWATER VULNERABILITY IN AN ALLUVIAL INTERFLUVE USING GIS

Abstract

The growth of human population often corresponds with change in land use, including expansion of urban areas, which necessitates increasing the available amount of drinking water. As the surface water sources are more amenable to pollution, it has become necessary to use groundwater at an increasing rate. Groundwater is normally abundant in the alluvial region where the urban areas are often located. Such areas face a greater risk of pollution of groundwater due to several factors. Keeping these aspects in view, groundwater vulnerability studies have been carried out in a selected alluvial area of northern India. The study area is situated in the upper part of the Ganga-Yamuna interfluve, and is considered to be the major recharge zone for the deep aquifers of the region. The aim of this study is to identify the groundwater vulnerability in the area so that the groundwater can be protected from pollution. In present work, it was envisaged to review the methods currently available for assessment of the groundwater vulnerability and to develop an appropriate method suitable for the alluvial aquifers of the Ganga-Yamuna interfluve area. Attempts were also to develop a multipurpose database in GIS environment, and to validate the developed method by comparing its findings against the observed water quality characteristics ofthe region. The study area is the northern part of the vast Indo - Gangetic Plain in India and lies , . , o I II o I II o I II 0 I II between latitudes 29 33 51 to 30 19 10 Nand longitudes 77 06 20 to 78 20 15 Ewith a total geographical area ofapproximately 4900 km2. Administratively, the study area covers the districts of Hardwar in Uttaranchal and Saharanpur in Uttar Pradesh, and has apopulation of about 4.3 million as per Census of 2001. The Ganga River and its tributary, the Yamuna, are the two major rivers in the region. These two rivers are perennial in nature and form the eastern and western boundaries of the study area. Other small intermittent streams like Ratmau River, Solani, and Banganga - the tributaries of the Ganga, and Hindon River, a tributary of the Yamuna, drain the area. Anetwork of canals exists in the study area for meeting the irrigation needs; the notable being the Upper Ganga canal and Eastern Yamuna canal along with their distributaries and branches. The climate of the area is humid and subtropical. The rains occur mainly during July to middle of September with annual average rainfall of about 100 cm. ~* LISS-III data of India Remote Sensing (IRS) satellite was used to prepare the land use map of the study area. On the basis of these data, the whole area was classified into three land use classes, i.e. urban, rural (including agriculture), and forest. The field monitoring was carried out during January 2002 to December 2003, through extensive field surveys covering entire study area. Groundwater samples (136 no.) were collected from various sites well distributed among various land use types in the study area and. Also, soil samples (48 no.) were collected from different sites covering all the land use categories. Geological and hydrogeological settings were examined for the study area. The depth to groundwater was monitored from the observation wells (119 no.). Besides, historic data available for past 10 years was also taken from Uttar Pradesh Groundwater Department, Roorkee. Groundwater recharge was estimated by Tritium tagging method at five locations during the period from Jun 2002 to Nov 2002. These, alongwith earlier data available at National Institute of Hydrology, Roorkee were used. A Digital Elevation Model (DEM) was prepared by digitization of bench mark and topographic points from relevant maps. The DEM was in turn, used for the construction of the slope map. The depth to groundwater map of the study area was generated from the observed data. The soil samples were analyzed to ascertain the soil texture and to prepare the soil map. This map showed that the most of the northern part, paleochannels and active floodplains of rivers have soils of sandy loam texture whereas the remaining part of the study area is covered by soils having silty loam. Considering the soil texture as an appropriate source factor of variation, the average recharge percentage was calculated as 6.3 % in silty loam soils and 15.5 % in sandy loam soils. The software RockWorks99 was used to prepare the geological fence diagram. Hydraulic conductivity map was prepared using Hydraulic conductivity data in GIS environment. The flow direction map, showed that the ground water flows from the northern and northeast part (the hilly area) to southern and southwestern part and follows the general topography of the study area. Hindon River and Solani River are fed by groundwater in the southern part. The hydraulic conductivity in the study area varies between 10 m/day and 48 m/day. The south-western part shows higher values in general, whereas the north-western part shows lower values. The quality of the soils in the study area was analyzed. The results exhibit that in general the urban soil has the highest values of all the physicochemical parameters followed by the rural and forest soils indicating an important role of urban activities The groundwater samples were analyzed for various physicochemical parameters like pH, EC, TDS and major ions (Ca2+, Mg2+, Na+, K+, HC03", CI", S042", C032", F), nutrients (N03\ VI P04 "), total organic carbon (TOC) and heavy metals (Cd, Fe, Mn, Ni, Zn, Pb, Cr, Cu). In general, the groundwater quality in the study area does not indicate much variation between postmonsoon and premonsoon periods. Further, all major ions, except N03"and K+. show an increasing trend from north to south and southwest. The groundwater is generally alkaline in nature with pH ranging from 7.01 to 8.90. The TDS values range between 117 to 1002 mg/1. HC03" is the dominant major anion followed by C1>S042> N03">F">P04 3" whereas Ca2+ is the major cation followed by Na+>Mg2+>K+. The heavy metal Zn is dominant followed by Mn>Fe>Pb>Cd. The calcium-bicarbonate facies are dominant in the groundwater of the study area. indicating that a substantial part ofground water is derived as recharge from the Bhabar zone in the north. Generally, the concentration of chemical parameters, except N03" and S042\ follow a decreasing trend for urban > rural > forest land use categories, whereas N03" and S04 "follow the decreasing trend for forest >urban >rural land use. Besides, a few samples show some parameters like total alkalinity (TA), nitrate, TDS, and Calcium having higher ranges than acceptable limits ofIndian standard (BIS: 10500) for drinking water. Among the heavy metals, tolerance limits of cadmium, manganese, lead and iron are violated in several samples. On the basis ofdemonstrated violation ofthe acceptable limits, the quality parameters like TDS, Ca2+, Total Alkalinity (TA), N03\ Cd, Mn, Pb, and Fe have been selected for computation of an Index of Aquifer Water Quality (MIAWQ), utilizing the framework as proposed by Melloul and Collin (1998). The index was modified for the present case in the sense that the weights to these eight parameters were, however, assigned as per their analytical hierarchy in the human health (effecting) significance and not in a subjective manner (as attempted in the original work of Melloul and Collin). The values ofMIAWQ (modified index) show an increase from north, north-east to south and south-western parts of the study area. The groundwater vulnerability mapping was carried out using two approaches viz. standard DRASTIC method and a modified DRASTIC-MOD method the study area. The DRASTIC parameters were evaluated in GIS environment as seven restart-map layers. The rating percentages were subsequently added to obtain the total cell rating. The DRASTIC index in the study area ranges from 122 to 183. The east and south-west corners ofthe study area and the paleochannels in the southern part show higher vulnerability index values. While applying the DRASTIC method on the present study area for assessment of vulnerability, following limitations were noticed: vn <&- > The proposed rating scale for the parameter "Impact of vadose zone" did not adequately address the implicit variability among the geological constituents of the vadose zone viz. sand gravel, silt and clay, andthe resulting complexity. > The "Hydraulic conductivity" values observed in study area mostly surpassed the highest limit/range of the rating scale rendering observed spatial variability meaningless with respect to the aquifer vulnerability. > Observing definitive signals about the influence of land use (urban > rural and agricultural > forest) on the soil and groundwater quality, the parameter "land use " appeared to also have on important bearing the status of aquifer vulnerability alongwith other parameters proposed earlier. Due modifications were incorporated in the original DRASTIC model in view of the above and the modified (DRASTIC-MOD) index map was sub-divided into four classes, (i) 100 - 119 with low risk in the north part of the study area (forest area), (ii) 120 - 159 with moderate risk in the Bhabar zone with deep depth to groundwater and forest area, (iii) 160 - 199 with high risk in most parts of the study area, these values resulting mainly from cumulative effects of rural and agricultural land use, low to moderate depth to groundwater and high recharge coefficient, (iiiv) Indices of 200 and above with very high vulnerability in some parts of the study area, reflecting the shallow depth to groundwater, high recharge and high urbanization related activities. DRASTIC-MOD indicates high vulnerability in the southern parts of the study area indicating higher risk of groundwater pollution. In order to validate the projected risk of vulnerability with actual groundwater quality statues in the region, DRASTIC and DRASTIC-MOD maps were correlated with modified Index for Aquifer Water Quality (MIAWQ). The MIAWQ showed high significant correlation with DRASTIC-MOD map. The differences observed in the spatial distribution of vulnerability estimates obtained from both the methods (DRASTIC and DRASTIC-MOD) indicate that in the areas with existing well defined land use practices, vulnerability estimation should necessarily include "land use" as a parameter. Further, in view of a good correlation between the DRASTICMOD and MIAWQ maps, it may be inferred that the "risk of vulnerability" corresponds quite well with the existing water quality scenario in the study area, a finding not commonly reported by researchers earlier. This also highlights the need of initiating corrective measures in many parts of the study area as well as to establish a suitable monitoring protocoal to detect adverse quality trends in the future.