ASSESSMENT OF GROUNDWATER QUALITY AND RECHARGE POTENTIAL IN KABUL, AFGHANISTAN

Abstract

Groundwater is an integral part of water resources. It has a vital role in water use in Kabul, Afghanistan. It is the only available source of water supply in the city. Since groundwater is the only accessible source for water supply purposes, studying the quantity and quality of underground water is of particular importance. The present study aimed to evaluate groundwater quality and its recharge potential in Kabul, Afghanistan. This study comprehensively analysed 35 groundwater samples and determined their hydrogeochemical characteristics, quality, water types, and suitability for drinking purposes. Various parameters were measured using different instruments and methods, including total dissolved solids (TDS), pH, electrical conductivity (EC), hardness, chloride, bicarbonate, sodium, calcium, magnesium, potassium, fluoride, sulphate, and nitrate. The distribution pattern of these parameters and the Water Quality Index (WQI) was spatially modelled using the ArcGIS tool. The results indicate that the main anions and cations follow an ascending order of Iron < Nitrate < Sulphate < Chloride < Bicarbonate and Potassium < Calcium < Sodium < Magnesium, respectively. Bicarbonate, chloride, nitrate, magnesium, sodium, calcium, and potassium exceeded the World Health Organization (WHO) permissible limits in drinking water samples. The Piper diagram analysis shows that the major water type is Mg-HCO3 (about 83%). The rest is Na (11.4%), Ca-Na-HCO3 (5.7%), and Ca-Mg Cl (5.7%). According to Gibbs' plot results, all water samples are of rock dominance and precipitation dominanceAccording to the WQI, approximately 88.57% of the study area has excellent to good water quality, while 11.43% has poor to very poor water quality. The Artificial Neural Network (ANN) model was developed in MATLAB to predict groundwater quality. The results of the ANN model for simulating sodium concentration in groundwater based on input data (EC, TDS, Salinity) have an average variance of 11.53%. The average variance for chloride and sulfate is 3.83% and - 3.41%, respectively. However, the average variance for potassium and total hardness is 259.6% and 45.25%, respectively. These different mean percentages of variances show the models' accuracy and suitability. Based on these percentages, one can conclude that the model is very suitable for simulating the concentrations of sodium, chloride, and sulfate in groundwater with the suggested inputs (EC, TDS, and Salinity). Therefore, the model is unsuitable for predicting potassium and total hardness in groundwater with the same inputs. Data on groundwater quality from 54 monitoring wells were collected by the National Water Affairs Regulation Authority of Afghanistan, including data from both dry and wet seasons. The analysis focused on specific water quality measures such as EC, TDS, hardness, nitrate, chloride, fluoride, sulfate, sodium, and some heavy metals such as iron and manganese. Spatial distribution maps and temporal variations were created to examine trends in groundwater quality and seasonal fluctuations. Statistical analysis revealed significant seasonal changes in magnesium, sodium, chloride, fluoride, iron, and manganese concentrations. Out of the 20 water quality assessments conducted, 14 during the dry season and 15 during the wet season showed concentrations exceeding the WHO recommendations. The variations in water quality metrics were influenced by factors such as recharge volume, hydraulic conductivity, vi and the geological formation of the region. Notably, the levels of qualitative parameters were higher during the wet season, particularly in wells located near river routes or in agricultural areas. The study also analyzed groundwater level trends and assessed drought dynamics in Kabul city. Cluster analysis was used to classify observation wells based on long-term trends in groundwater level data. The Mann-Kendall statistical test was employed to determine seasonal and annual variations in groundwater depth. The Standardized Groundwater Level Index (SGI) was used to measure groundwater drought. The trend analysis revealed that water levels in 82% of the observation wells were significantly decreasing. From 2014 to 2020, the study area experienced increasingly severe and persistent drought, according to the SGI results. The analysis of land use and land cover (LULC) indicated that the built-up area in the study area increased from approximately 15% in 2005 to 32% in 2020, while bare land decreased from about 67% in 2005 to 52% in 2020. The significant decline in groundwater level can be attributed to changes in LULC, excessive groundwater exploitation, and declining annual precipitation. The study employed an integrated application of the analytical hierarchy process (AHP) and ArcGIS to identify potential groundwater zones in the study area. Ten different thematic variables were analyzed in the Arc GIS environment with various numerical weightage values in the basin. These variables included geology, geomorphology, land-use land cover, lineament density, drainage density, soil, slope, rainfall, elevation, and water depth. Static groundwater level records have been utilized to acquire precision and reliability for discovering groundwater potential zones. According to the final output of the results, most parts of the study area are covered by a reasonable and very good capacity of groundwater potential zones. Based on the results, four categories of the GWPZs were eventually recognized. According to the statistics, the area is divided into zones with very poor potential (16%), poor (18%), good (35%), and very good (31%). Cities in arid and semi-arid regions face challenges in managing urban floods and water shortages. Kabul City in Afghanistan is particularly vulnerable to groundwater decline and urban floods. This study explores using rainwater harvesting (RWH) to manage floods and recharge groundwater in Kabul City. The research analyses rainfall patterns, including variability, rainy days, seasonality, probability, and maximum daily precipitation. The findings show rainfall greater than 30mm occurs approximately every 3-4 years. Rainfall in Kabul is seasonal, with a coefficient of variation of 127% in October and 46% in February. The study also investigates the potential of RWH in Kabul City for stormwater management and groundwater recharge. Based on LULC, implementing an RWH and recharge system could increase mean annual infiltration from 4.86 million cubic meters (MCM) to 11.33 MCM. A weighted curve number (CN) of 90.5% indicates the dominance of impervious surfaces. The study identifies a rainfall threshold of 5.3 mm for runoff generation. Two approaches to rainwater collection for groundwater recharge were explored: RWH for a residential house (yielding about 88m3 /year) and RWH for a street sidewalk (collecting water from streets and sidewalks). These findings highlight the potential of RWH for effective management of urban floods and artificial groundwater recharge.