The augmentation of human population regularly corresponds with change in the land cover, including expansion of urban areas, which imposes increasing the available amount of domestic and drinking water. The study area, Halabja-Saidsadiq Basin, is situated in the Northeast of Iraq and is one of the major groundwater sources of the region. As the surface water sources are not enough in the studied area, it has become necessary to use groundwater at an increasing rate. Usually, a huge amount of groundwater is plentiful in the alluvial deposits or rock outcrops where the urban areas are frequently situated. Such areas face a huge risk of pollution of groundwater due to producing different sources of a contaminant from human's activity. Keeping these aspects in view, groundwater vulnerability studies have been carried out in the current studied basin. The main objective of this work is to investigate the environmental impacts on groundwater quality and recognize the groundwater vulnerability in the area so that the groundwater can be protected from probable contaminations.
In the current study, DRASTIC model has been applied since it is one of the most proper useful methods available for the assessment of the groundwater vulnerability. This model has been modified in different ways to achieve the obvious vulnerability condition in the area; likewise, different further methods have been applied for comparison purposes such as: weight modified VLDA and standard COP models. In addition, the applied models were validated by comparing its findings against the estimated groundwater ages and the observed water characteristic qualities within the region in two successive seasons.
According to the spatial distribution of irrigation water quality index, groundwater at the studied basin classified into three group for both dry and wet seasons, namely, Sever Restriction (SR), High Restriction (HR) and Moderate Restriction (MR). The coverage area of all three classes are (1.4%, 52.4% and46.2%) for dry season and (0.7%, 83.3% and16%) for wet seasons, respectively. While, refer to the water quality index for drinking purpose, groundwater in this basin reveals a permissible to excellent groundwater quality of the dry season and a good to excellent groundwater quality of the wet season. The high level of a good groundwater quality in the wet season compared to the dry season might be expected to the groundwater recharge during the winter and spring periods, which lead to dilution of chemical component, in contrast high irrigation and agricultural activities and groundwater discharge in the dry season leads to increase the concentration of chemical component.
Field and official data were collected to review several environmental impacts and were used to map standard DRASTIC vulnerability model for the study basin. Based on this model, the study area was classified into four zones of vulnerability indexes, comprises a very low, low, moderate and high vulnerability index with a coverage area of (34%, 13%, 48% and 5%) respectively.
In the first modification step, the rate and weight value of each parameter in DRASTIC model is modified. Nitrate concentration from 39 groundwater samples was used for modifying the recommended standard rating value based on the Wilcoxon rank-sum nonparametric statistical test and then sensitivity analysis was used to modifying recommended standard weighting value of each parameter. To calibrate the modified rate, the Pearson's correlation coefficient was applied to estimate the relation between DRASTIC values and nitrate concentrations in groundwater samples. For the first modified model, the correlation coefficient was 72% that was significantly higher than 43% achieved for the standard model. The modified model classified the area into five vulnerability classes, including (very low, low, moderate, high and very high) with covered area of (7%, 35%, 19%, 35% and 4%), respectively.
The second modification of DRASTIC model was based on land use and land cover for the studied area. The land use and land cover (LULC) map prepared using ERDAS IMAGINE software from two different scenes of Landsat Thematic Mapper (TM). The LULC map indicates that only five classes of LULC can be identified: these are: barren land, agricultural land, vegetation land, urban area and wet land or water body. The modified DRASTIC based on LULC map classified the area into five classes with different coverage area of each class: very low (1.17%), low (36.82%), moderate (17.57%), high (43.42%) and very high (1.02%).
The third modified method of the current study is the modification of DRASTIC model based on Lineament feature of the study basin. A lineament map is extracted from Enhanced Thematic Mapper plus (ETM+) satellite imagery using different techniques in remote sensing and GIS. The lineament density map demonstrates that only six classes of lineament density can be identified ranged from (0-2.4). The third modified DRASTIC model classified the area into four vulnerability categories: very low (28.75%), low (14.31%), moderate (46.91%) and high (10.03%).
The fourth effort to modify standard DRASTIC model is the application of Analytical Hierarchical Process (AHP) to assess the weight value of each parameter. The modified DRASTIC vulnerability index values based on AHP method ranged between (65.82–224.1) with five vulnerability classes comprises (very low to very high).
Weight modified VLDA and standard COP models were also applied to map vulnerability system in the study basin. The vulnerability outcome based on weight modified VLDA model revealed that a total of four ranges of vulnerability indexes had been distinguished ranging from low to very high with vulnerability indexes (2.133-9.16). Subsequently, based on the standard COP model, the area is also divided into four vulnerability classes ranging from very low to high with index value ranged from (0.79) to (6.2).
All applied models in the study basin were compared to each other and validated to clarify the validity of the theoretical sympathetic of current hydrogeological conditions and to show the accuracy of the modeled vulnerability system. Two methods were applied for the validation of the result, in the first approach; nitrate concentration analysis has been selected; the nitrate differences between two following seasons (dry and wet) were analyzed from (39) water wells. In the second approach, groundwater vulnerability was assessed based on estimated groundwater age from range of tritium (3H) value in the groundwater samples from different groundwater aquifers in the studied area. The results of both validation methods verify the sensibility of the gradation and distribution of vulnerability levels acquired using the modified DRASTIC model based on (rate and weight modification, weight modification based on AHP process and effect of LULC on DRASTIC model) and also applying weight modified of VLDA model.