ARSENIC CONTAMINATION OF GROUNDWATER IN KATHMANDU VALLEY, NEPAL, AS A CONSEQUENCE OF RAPID EROSION

Elevated levels of As in groundwater in the flood plain of the Ganges River in Bangladesh, West Bengal (India) and the Terai (southern Nepal) have been well-documented over the past decade. The objective of this study was to measure As and the transition elements normally associated with As in Kathmandu Valley in Nepal, a heavily populated tectonic valley in the upper reaches of the Ganges River system. Water samples were collected from six shallow tubewells (depth < 50 m), eight deep tubewells and 13 dug wells and stone spouts. Electrical conductivity, temperature and pH were measured on-site and concentrations of As, Fe, Cu, Ni, Co, Mn, Zn and Cr were measured with a spectrophotometer. Five tubewells and four dug wells had As levels exceeding the Nepal Interim Standard (As = 0.05 mg/L). The maximum As level was As = 2.07 mg/L in the deep tubewell of Paropkhar Maternity Hospital. There was no statistically significant clustering of As levels either with depth or horizontal location. Arsenic was uncorrelated with either Fe (R² = 0.096), Mn (R² = 0.0004) or any combination of transition elements (R² < 0.083), which is inconsistent with both the reduction-dissolution and the sulfide oxidation models for As release. The geometric mean As level of groundwater (As = 0.015 mg/L) was indistinguishable from the geometric mean As level of surface water (As = 0.013 mg/L) obtained from 48 river samples from Kathmandu Valley in a previous study. Although the rivers of Kathmandu Valley are heavily polluted, the geometric mean of surface water As is indistinguishable from the geometric mean of surface water As (As = 0.010 mg/L) found in 102 river samples throughout central Nepal outside of Kathmandu Valley. We are suggesting that elevated groundwater As results not from subsurface redox conditions or from direct human activity, but from losing streams with elevated As, which is a consequence of rapid erosion caused by a combination of monsoon climate, rapid tectonic uplift and deforestation. Other recent studies have reached different conclusions by studying a different set of wells and not comparing groundwater with surface water. It is crucial that a common database be created for all studies of groundwater quality in Kathmandu Valley and that this study be repeated in other tectonic valleys in the upper reaches of the Ganges River system.