Seismologically Observed Spatiotemporal Drainage Activity at Moulins

Abstract

Hydrology is important for glacier dynamics, but it is difficult to monitor the subsurface drainage systems of glaciers by direct observations. Since meltwater drainage generates seismic signals, passive seismic analysis has the potential to be used to monitor these processes. To study continuous seismic radiation from the drainage, we analyze geophone data from six stations deployed at the Kaskawulsh Glacier in Yukon, Canada, during the summer of 2014 using ambient noise cross-correlation techniques. We locate the noise sources by backprojecting the amplitude of the cross correlation to the glacier surface. Most of the ambient noise sequences are found in two clusters, with each cluster located in the vicinity of a moulin identified at the surface. Stronger seismic radiation is observed during the day, consistent with expected variability in melt rates. We demonstrate that the sparse seismic network array with 2 km station separation has the ability to detect moulins within the array with a precision of 50 m. We confirm that seismic activity is correlated with air temperature, and thus, melt, on a diurnal timescale, and precipitation correlates with the activity at longer timescales. Our results highlight the potential of passive seismic observations for monitoring water flow into subglacial channels through moulins with an affordable number of seismic stations, but quantification of water flow rates still remains a challenge. The cross-correlation backprojection technique described here can also potentially be applied to any localized source of ambient noise such as ocean noise, tectonic tremor, and volcanic tremor.