The Hydrological System of Storglaciären, Sweden: Integrating Modelling with Observations

Change log
Williamson, Andrew G. 

There is a need for more studies in glacier hydrology that integrate numerical models with detailed empirical data collected over one/multiple summer(s). This is especially true for physically based models whose main advantage lies in ability to predict basal water pressure because of its significance for ice dynamics, thus helping inform knowledge of runoff and ultimately sea-level rise both presently and, when forced with climate projections, into the future. This study helped fulfill this requirement by applying a physically based glacier-hydrology model to the predominantly temperate ~3.22km2 Storglaciären, Sweden, forced with 2012 empirical data. The model has three subelements: a high temporal (hourly) and spatial (20-m) resolution surface-energy-balance model to generate meltwater across Storglaciären’s surface for the entire summer, a surface-routing model to route this meltwater (and precipitation) across the surface either until it runs off Storglaciären’s edges or is intercepted by moulins extending into the internal system, and a subglacial-hydrology model where inflow from moulin hydrographs is routed in R-channels from moulins to the terminus to produce proglacial discharge. The surface-energy-balance model was validated using Storglaciären’s summer mass-balance data and measurements of surface lowering at an automatic weather station on its surface. It performed well, though slightly (~4%) underestimated melt. The surface-routing model was qualitatively evaluated using estimates of direct supraglacial runoff derived from previous studies and reproduced these to within 2%. The subglacial-hydrology model’s outputs of proglacial discharge and subglacial flow-routing times were quantitatively compared with empirical discharge measurements in the three proglacial streams Nordjåkk, Centerjåkk and Sydjåkk, and with flowrouting times from ~25 tracer injections across Storglaciären in 2012 and 2013. To replicate these data, the subglacial model required high conduit roughnesses to be specified (Manning’s n = 0.125 for conduits leading to Centerjåkk/Sydjåkk and n = 0.075 for the conduit producing outflow at Nordjåkk). However, this is encouraging and follows postulations of braided, broad and low Hchannels beneath Storglaciären, which cannot be explicitly accounted for by the model, so rough Rchannels are instead required. Overall, the subglacial model performed well, though slightly underestimated discharge and overestimated flow-routing times. The internal drainage system was reevaluated, with suggestion that the englacial network in the upper ablation and firn areas may feed Centerjåkk/Sydjåkk, not Nordjåkk, contrasting with suggestions in previous studies. The physically based model performed less favourably than simpler linear-reservoir models applied to Storglaciären; however, its key worth was in generating continuous basal water pressure for the entire summer, the first data of this kind to be generated for this glacier. These data compared well with suggestions from empirical borehole measurements in previous years. Specifically, the ‘Spring Event’ (marked by a high-pressure period at the end of May), continuously high pressures (at/above ice-overburden) within the overdeepened (up to ~230m) area and the response of water pressures to major meteorological forcings were captured. The fact that continuously high pressures within the overdeepening were reproduced despite the model’s inherent specification of R-channels suggests this flow morphology is possible here, informing our understanding of processes within overdeepenings.

Arnold, Neil
Master of Philosophy (MPhil)
Awarding Institution
University of Cambridge