Dense borehole network observations of a small surge-type glacier

Subglacial drainage is known to play an important role in glacier dynamics trough its influence on basal sliding. However, drainage is also one of the most poorly understood process in glacier flow due to the difficulties of observing, identifying and modeling the physics involved. In an effort to improve understanding of subglacial processes, we have monitored a small, approximately 100 m thick surge-type alpine glacier for seven years. Over 300 boreholes were instrumented with pressure transducers over a 0.5 km² in its upper ablation area, in addition to a permanent GPS array and weather station. To quantitatively characterize the morphology and evolution of the subglacial drainage, we use Self-Organizing Maps (SOMs) in combination with various filtering techniques. This allows us to identify clusters of boreholes exhibiting similar evolution in water pressure. Using moving time windows, we use this approach to identify connections between boreholes, and to track how these changes evolve over longer time intervals. Our results include: i) The different areas can be subdivided in three groups according to the water pressure variations: Hydraulically isolated, hydraulically connected with distributed drainage, and channelized areas ii) Transitions between connected and isolated areas are mostly discrete and rapid switch-like events, taking from a few minutes to a few hours iii) There is limited evidence to support the assumption that the hydraulic connection between instruments installed at the bed happens necessarily trough the bed interface itself, with engacial conduits also playing an important role iv) Transmissivity, conductivity and reflectivity measurements in a small subsets of boreholes suggest that most water pressure variation measured at the bed in areas of "distributed" drainage are not accompanied by substantial water transport, even where time lags and attenuation of pressure suggest behavior consistent with a "macroporous" water sheet at the bed vi) Preliminary evidence suggest that englacial conduits can persist over winter, potentially forming an hydraulic network that could play an important role in water transport We discuss how our results fit into current efforts to model subglacial drainage and alterations to drainage model physics that are suggested by our observations.