Wintertime Brine Discharge at the Surface of a Cold Polar Glacier and the Unexpected Absence of Associated Seismicity

Abstract

Episodic discharge of subglacially sourced, iron-rich brine at the terminus of Taylor Glacier, Antarctica, forms the feature named Blood Falls (Figure 1). The brine discharges over weeks to months during releases that occur several times per decade, and can occur during any season. Brine deposits have primarily been observed at two subaerial locations: the Blood Falls site at the glacier surface and, less frequently, at a lateral site where brine icings have been observed in the ice-marginal stream bed at the northern terminus margin. Compilations of brine release activity and brine deposit observations include Black (1969), Carr (2021, Keys (1979), andLawrence (2017). In this paper, we focus on the Blood Falls site; therefore, phrases like "brine discharge" and "brine release event" refer to discharge at the Blood Falls site unless otherwise specified.An unresolved question is what triggers the episodic brine release. Carmichael et al. (2012) hypothesize that meltwater-driven fracturing during the summer melt season could propagate deep enough into the glacier to trigger brine outflow. However, springtime observations of brine icing superimposed on lake ice (e.g., Black, 1969;Keys, 1979) indirectly suggest that brine releases can occur during the wintertime in the absence of surface Abstract A subglacial groundwater system beneath Taylor Glacier, Antarctica, discharges hypersaline, iron-rich brine episodically at the glacier surface to create Blood Falls. However, the triggering mechanism for these brine release events is not yet understood. Identifying which fracture processes are observed seismically can help us better characterize the hydrological system at Taylor Glacier, and more generally, provide us with a broader understanding of englacial hydrologic activity in cold glaciers. We document wintertime brine discharge using time-lapse photography. Subfreezing air temperatures during the brine discharge indicate that surface melt-induced hydrofracture is an unlikely trigger for brine release. Further, we analyze local seismic data to test a hypothesis that fracturing generates elevated surface wave energy preceding and/or coinciding with brine release events. Our results show no discernible elevated Rayleigh wave activity prior to or during Blood Falls brine release. Instead, we find a pattern of seismic events dominated by a seasonal signal, with more Rayleigh events occurring in the summer than the winter from the Blood Falls source area. We calculate that the volumetric opening of cracks that would generate Rayleigh waves at our detection limits are of similar size to myriad cracks in glacier ice, lake ice, and frozen sediment in the terminus area. We therefore propose that any fracturing coincident with brine release activity likely consists of a series of smaller opening events that are masked by other seismicity in the local environment.Plain Language Summary Blood Falls is a reddish feature that forms at the terminus of Taylor Glacier in Antarctica when hypersaline, iron-rich brine flows from sediment...