HYDROLOGICAL PROCESSES Hydrol. Process. (2008) Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/hyp.7149 Understanding the extent of interactions between groundwater and surface water through major ion chemistry and multivariate statistical techniques Manish Kumar, 1† Al. Ramanathan 1 * and A. K. Keshari 2 1 School of Environmental Sciences, Jawaharlal Nehru University, New Delhi-110067, India 2 Dept of Civil Engineering’ Indian Institute of Technology, New Delhi-110016, India Abstract: The present work examines the possible use of major ion chemistry and multivariate statistical techniques as a rapid and relatively cost-effective method of identifying the extent of groundwater and surface water (GW–SW) interaction in an urban setting. The original hydrogeochemical dataset consists of groundwater (n D 114), stream water (n D 42) and drain water (n D 24) samples, collected twice in a year for the pre- and post-monsoon seasons, for three successive years along an 8 km reach of the Delhi segment of River Yamuna, India. The dynamic and similar seasonal changes of hydro-geochemical facies and major ion trends of river, drain and groundwater samples indicate the existence of an empirical relationship between GW and SW. Results of both R- and Q-mode factor and cluster analyses highlight multi-scale control of the fluid exchange distributions, with distinct seasonal alteration in mode and extent of GW–SW interaction, namely, the influence of the mixing zones between urban river and groundwater and the pattern of groundwater flow through the river bed. Hierarchical cluster analysis (HCA) of sampling locations efficiently illustrates different groups that comprise samples severely influenced by contaminated surface water downstream and the upstream fresh water samples. These results substantiate the strong exchange processes between GW and SW all along the stretch. The study shows that the combination of an empirical and statistical relationship between different ionic species and sampling locations can provide greater confidence in identifying the extent of GW–SW interaction/exchange processes. Copyright 2008 John Wiley & Sons, Ltd. KEY WORDS groundwater/surface water interaction; factor analysis; cluster analysis; hydrogeochemistry; Yamuna River; Delhi Received 21 September 2007; Accepted 20 August 2008 INTRODUCTION The quality of groundwater evolves from its pri- mary composition (i.e. precipitation) due to constant interaction with the ambient environment, before and after recharge; thus interactions with surface water, atmospheric and anthropogenic inputs, soil matrix and mixing/non-mixing of different types of groundwater are important (Matthess, 1982; Kumar et al., 2006; 2007). Although the physical characteristics of groundwater and surface water have been considered to be distinctive, sur- face water can be regarded as a ‘perched groundwater aquifer’, in recognition of the isotopic history of both components (Freeze and Cherry, 1979; Sear et al., 1999; Lambs 2004). This is because in applied hydrology, the role of groundwater in sustaining low flows of surface water has been well recognised on the basis of simi- larity in the stable isotope ( 2 H and 18 O) signatures of surface water and groundwater. Moreover, groundwater * Correspondence to: Al. Ramanathan, School of Environmental Sciences, Jawaharlal Nehru University, New Delhi-110067, India. E-mail: alrjnu@gmail.com † Present Address: Dept of Urban Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, 113-8656, Japan. E-mail: manish.env@gmail.com. and surface water (GW and SW) are integral components of any hydrological system. Therefore, the contamina- tion or development of one is bound to have an effect on other. Understanding GW–SW interaction is impor- tant for determining contaminant migration pathways as well as for water resources management, especially in an urban setting (Chapman et al., 2006; Ellis et al., 2007). The degree of interaction may depend on a number of factors including topography, underlying geology, sub- surface hydraulic properties, temporal variation in precip- itation, and local groundwater flow patterns (Cey et al., 1998, Oxtobee and Novakowski., 2002; Westbrook et al., 2005). Henceforth, understanding GW–SW interactions presents a unique challenge (Sophocleous, 2002). To study the interaction between GW and SW, five general approaches/methodologies can be followed: (i) experimental determination and estimation through mathematical modelling with NETPATH, MODFLOW or models based on a representative elementary water- shed (REW) (Laaksoharju et al., 1999; Fencia et al., 2005; Ellis et al., 2007); (ii) isotopic characterization of surface water and groundwater using 18 O, 3 H and 2 H (Ojiambo et al., 2003; Lambs, 2004; Tsujimura et al., 2007); (iii) comparison between surface water chemistry and that of groundwater (Sami, 1992; Banaszuk et al., Copyright 2008 John Wiley & Sons, Ltd.