On the use of water stable isotopes to estimate snow and ice melt contribution in a glaciated catchment

<p>Glaciated catchments in the Swiss Alps are subject to intense changes due to rapid glacier retreat. Combined with earlier snowmelt, water availability in summer and especially during drought periods will become limited, which will have a strong influence on the local ecology but also for downstream water usages.</p><p>Water stable isotopes have become a commonly used tracer in hydrological studies to separate the contribution of snow and rain, due to their distinct signature and conservative behaviour, but also due to their simple sampling procedure and low analysis costs. In high alpine catchments, their application is however challenging because the temporal and spatial isotopic composition of ice and snow varies strongly (which requires intense sampling) and because their isotopic signal range is often strongly overlapping</p><p>Based on a case study of a glaciated catchment in the Swiss Alps, we propose in this study to compare the estimation of snow and ice melt contribution using two separate approaches: a glacier surface mass balance modelling approach, and a separation using a large database of isotopic observations. We show the large spatial and temporal variation of the snow isotopic signal at the catchment scale and propose a simple modelling approach to reconstruct its temporal evolution. On this basis, we show that snow and ice melt isotopes show very similar ranges of values during the summer due to the effect of rain on snow and snow fractionation. As a result, it appears that the estimation of their respective contribution is very sensitive to the accuracy of the estimated snow melt end-member. In comparison, the mass balance modelling approach may provide more robust results but requires a more field intensive work in order to measure snow accumulation and ice ablation. Finally, we insist that water stable isotopes should only be used with a proper statistical assessment of their spatial variability and a proper characterization of their temporal evolution in order to provide any realistic estimations in glaciated catchments.</p>