Isotopic characterization (δ18O, δD and d-excess) and modelling of stream flow partitioning along an altitudinal gradient in the Satluj River basin (SRB) Himalayan-alluvial plain using end member mixing analysis (EMMA) and associated uncertainty

Abstract

Stable isotopic (δ¹⁸O, δD and d-excess) signatures were employed to gain valuable insights regarding key hydrological processes and modelling of stream flow runoff partitioning of Satluj River Basin (SRB) in the western Himalayan region and plains of Punjab, where surface water is extensively used for hydropower, agricultural and domestic supply. The SRB has not yet been unexplored. Spatially extensive and comprehensive stable isotope data are meagre, which constricts the thoughtful comprehension of hydrological processes, sources of moisture, the isotope-elevation relationship and provenance quantification in this transboundary river across the mountain-plain basin. The results showed a significant difference in δ¹⁸O, δD and d-excess among the main stream, tributaries and groundwater during the observed period, mainly regulated by snow-glacier melt input, evaporation and recycled moisture contribution. The deviation of the water line in each hydrological compartment from the Global Meteoric Water Line, as well as its similarity to the local, regional and local water lines, suggests that these compartments have undergone significant evaporation during the recharge process. This similarity in moisture dynamics implies a potential link between the local and regional hydrological systems. The isotopic signature integrated with back trajectories reflects the combined role of moisture sources via both westerlies and southwest monsoons followed by extensive local recycling. The isotopic altitude lapse rate (IALR) of the various hydrological compartments mimics the IALR results from the Himalayas and their surrounding catchments. Moreover, the isotopic composition (δ¹⁸O) of the main stream showed a closer fit to a second-order polynomial relationship between δ¹⁸O and elevation compared to other models. Furthermore, we employ end member mixing analysis (EMMA) principles to identify the possible end members and calculate contributing fractions for respective end members. Modelling of stream flow partitioning using a dual isotopic tracer (δ¹⁸O-EC and δ¹⁸O-d mixing model) based EMMA revealed the three possible end members, namely glacier melt, snow melt and groundwater, contributing 29.5%, 31.7% and 38.7%, respectively, via a δ¹⁸O-EC mixing model, while 29.6%, 31.3% and 39.0%, respectively, via a δ¹⁸O-d mixing model. The mathematically propagated uncertainty in the respective computed mixing fractions of EMMA is approximately 28%, 41% and 25% with the δ¹⁸O-EC mixing model and 19%, 54% and 35% with the δ¹⁸O-d mixing model, respectively, reflecting realistic uncertainties. An attempt is made to acknowledge, explore and hypothesize the realistic reason for the larger documented uncertainty. The study highlights the significant contribution of groundwater originating from the mountain-plain nexus to demarcate the river isotopic variability and underlines the need to concentrate on this important hydrological reservoir in the near future to learn about groundwater hysteresis and flow dynamics in this nexus mountain-plain watershed.