Reply to the comments of referee N°2

Abstract

The subglacial drainage system is one of the main controls on basal sliding, but remains only partially understood. Here we expand the analysis of the eight-year dataset of borehole observations presented in Rada and Schoof (2018). These observations were made on a small, alpine polythermal valley glacier in the Yukon Territory, and were previously used to describe the seasonal evolution of the drainage system and to underline the importance of hydraulic isolation at the glacier bed. Now, to explore the spatial structure of the drainage system and its seasonal progression, we automatically cluster boreholes based on similarities in their pressure records and follow their evolution through the melt season. Some of these borehole clusters show pressure variations that suggest they are part of a drainage system connected to the surface meltwater supply, while others show features consistent with hydraulic isolation. The distribution of connected and isolated boreholes suggests that the distributed drainage system we observe comprises a network of small conduits with spacings smaller than the borehole bottom diameter (approximately 25–50 cm). Within these hydraulically connected areas, pressure phase lags, and amplitude attenuation rarely shows the behaviour expected in a diffusive system. This observation suggests that the diffusivity distribution in such areas presents a fine structure at scales smaller than our minimum borehole spacing of 15 m. However, at a glacier-wide scale, we observe that hydraulic connections are ubiquitous in some regions of the bed and permanently absent in others, suggesting large contrasts in difuisivity. Within disconnected areas, boreholes often show small amplitude pressure variations associated with horizontal normal stress transfers. Such stress transfers seem to play a more important role than previously considered for controlling the effective pressure distribution at the bed. Through the melt season, the evolution of borehole clusters suggests that the diurnal meltwater supply promotes the growth of the low-efficiency drainage systems found early in the season while stimulating the shrinkage and fragmentation of the more efficient drainage systems that appear later in the season. Therefore, an increase in drainage efficiency is associated with the growth of disconnected areas. Our observations support the traditional view of a distributed drainage system early in the melt season that gradually evolves into a progressively more channelized system. However, the most notable difference is the highly heterogeneous distribution of diffusivity that our results suggest and the robust support for disconnected areas. The extent of disconnected areas could be an essential control of basal speed variations. It is possible that even relatively small disconnected areas could have a disproportionate effect on basal speed.