Delineation of Recharge Areas of Some Karst Springs Using Hydrogeochemistry and Stable Isotopes of Hydrogen and Oxygen

Hydrogeochemical and isotopic study of water samples collected from Bringi watershed SE Kashmir, India was carried out to delineate the recharge areas of karst springs. About 90 water samples of precipitation, snowmelt, streams and springs were collected on bimonthly basis for one complete year from March-08 to January-09 and analysed for major ions, δ18O and 3H following standard procedures. Both stream and spring discharge showed an inverse relation with precipitation. Spring and stream discharges were less during January and high during July. However, precipitation was high during January and low during November. Major ion analysis of precipitation, streams and springs revealed that Ca and HCO3 were the dominant ions, making up more than 50% of total ions which indicated carbonate lithology as the dominant source of ionic species. However, increased Na in some samples particularly Kongamnag indicated the impact of silicate weathering on water chemistry. The dominant order of cations and anions in water samples was Ca > Mg > Na > K and HCO3 > SO4 >Cl. In both streams and springs, EC, TDS, Ca, HCO3 were high during winter when the discharge was low and was low during summer when the discharge was high. However, Kokernag and Achabalnag also showed higher concentrations during July, resulted due to the piston effect. The springs showed a significant variability of total dissolved solids, with highest value of 180mg/L observed at Achabalnag followed by Kokernag (130mg/L) and Kongamnag (90mg/L). The high variability of TDS indicated rapid and strong reaction of Achabalnag to hydrological events followed by Kokernag and Kongamnag. The δ18O data generated indicated a strong spatial and temporal variation in oxygen isotope composition. An altitude effect of - 1.48 ‰ to - 0.13 ‰/100m change in elevation (mean altitude effect: -0.5‰/100m) was discernible in precipitation isotopic composition. The precipitation was δ18O depleted in cooler seasons/months and at higher elevations. The δ18O of precipitation showed a narrow range (0.7 to 1‰) during the periods/months of more precipitation and wide range (4.6 to 6.8‰) during the other months. Like precipitation, the isotopic composition of streams is also controlled by the elevations of their catchments. The stream water was δ18O depleted in the mountain areas with their catchments at higher altitudes than at lower elevations. The streams/tributaries were most δ18O depleted in May and least depleted in September. The δ18O composition of springs is similar to that observed in the streams, being most δ18O depleted in May and least δ18O depleted in September. On the basis of local vertical isotopic gradient of δ18O, the mean altitude of the recharge was estimated. Widespread melting of snow, high discharges of streams and springs and δ18O depleted stream and spring waters indicate the dominant contribution of snow melt in May; negligible snow melting, precipitation in the form of snow only, less stream and spring discharges and less variance indicate the contribution from baseflow in January; while as exhausted snow, moderate stream and spring discharges, high precipitation and δ18O enriched stream and spring water during September indicate the contribution from rainfall. The best correlation was observed between the δ18O of springs and streams. The mean residence time of karst springs calculated from tritium analysis is very short (less than one year). The residence time is more for Kongamnag and short for Achabalnag. Chloride Mass Balance Equation used to estimate the precipitation contribution to groundwater recharge averaged at 18.5%, with higher contribution during summer and lower during winter. δ18O Mass Balance Equation used to estimate the surface water component to groundwater recharge averaged at 75% during summer and 18.6% during winter. The results were further verified in field through dye tests. Coloured dye was inserted at two sinks near Adigam and Gadol which discharged very fast at Achabalnag and Kokernag, after a lag time of 48 and 120 hours, confirming the hydrological connection between injection locations and the springs