Prediction of Groundwater Arsenic Contamination using Geographic Information System and Artifcial Neural Network Md. Moqbul Hossain a , Krishna Neaupane b , Nitin Kumar Tripathi c and Mongkut Piantanakulchai a a School of Civil Engineering and Technology, Sirindhorn International Institute of Technology, Thammasat University, Rangsit Campus, Thailand. b Principal Engineer, Geoservices, URS Infrastructure & Environment UK Limited, Birmingham, United Kingdom c School of Engineering and Technology, Asian Institute of Technology, Thailand. Abstract Ground water arsenic contamination is a well known health and environmental problem in in several countries including Bangladesh. Sources of this heavy metal are known to be geogenic, however, the processes of its release into groundwater are poorly understood phenomena. In quest for the mitigation of the problem, it is necessary to predict probable contamination before it causes any damage to human health. This research has been carried out to investiaget the factors affecting the mobility of the contaminant and develop the prediction model. Researchers have generally agreed that the elevated concentration of arsenic is affected by several factors such as soil reaction (pH), organic matter content, geology, iron content, etc. However, the variability of concentration within short lateral and vertical intervals, and the inter-relationships of variables among themselves, make the statistical analyses highly non-linear and diffcult to converge with a meaningful relationship. Artifcial Neural Networks (ANN) comes in handy for such a black box type problem. This research uses Back propagation Neural Networks (BPNN) to train and validate the data derived from Geographic Information System (GIS) spatial distribution grids. The neural network architecture with (6-20-1) pattern was able to predict the arsenic concentration with reasonable accuracy. Keywords: arsenic; contamination; groundwater; ANN; GIS; pH 1. Introduction Elevated concentrations of arsenic in groundwater are reported from several countries including United States, Mexico, Argentina, Vietnam, China, India, Nepal and Bangladesh among others (Bhattacharya et al., 2004; Smedley and Kinniburgh, 2005; Bhattacharya et al., 2010). In Bangladesh, groundwater from shallow aquifers of Pleistocene to recent fuvial origin have been extensively used as the main sources of drinking and irrigation. The aquifers in unconsolidated and estuarine sediments are often contaminated with arsenic and pose a serious health and environmental concern in this country. According to a British Geological Survey report, a large part of Bangladesh groundwater is found to contain arsenic concentration beyond World Health Organization limit (≤10 μg/L) and national recom- mended standard (≤50 μg/L) (NERC/DPHE/DFID, 2001). It is estimated that around 57 million people drink water with arsenic levels exceeding the limits set by the WHO (Polizzotto et al., 2005). It is generally accepted that the origin of arsenic is natural, and it is being released into the ground water through different processes, which are poorly understood. Two main processes on which the arsenic mobility in the groundwater depend are adsorption and desorption which are infuenced by physicochemical conditions such as pH, occurrence of redox (reduction/ oxidation) reactions, presence of competing anions, and solid-phase structural changes at the atomic level. Arsenate (As III) and arsenite (As V), the two forms of arsenic commonly found in ground water, are adsorbed to the surfaces of a variety of aquifer rock, including iron oxides, aluminum oxides, and clay minerals. Groundwater arsenic concentration and distribution in the Bengal basin are not well understood. Variability of concentration within short lateral and vertical distances makes it further diffcult to predict the level of arsenic of a given well, even when the concentration of the adjacent wells are known (Van et al., 2003). It is, however, essential to establish relationships between arsenic level with simple yet measurable and identifable indices such as channel proximity, pH level, organic matter content etc. for long term engineering solution of the problem. When the relationship between the input and output is complicated or application of other available The international journal published by the Thai Society of Higher Education Institutes on Environment Environment Asia Available online at www.tshe.org/EA EnvironmentAsia 6(1) (2013) 38-44