Blood falls, Taylor Glacier, Antarctica: subglacially-sourced outflow at the surface of a cold polar glacier as recorded by time-lapse photography, seismic data, and historical observations

Abstract

Blood Falls forms when iron-rich, hypersaline, subglacially-sourced brine flows from a crack in the surface of Taylor Glacier, Antarctica. If air temperatures are low enough, the brine freezes to form a fan-shaped icing deposit. In chapter two, historical observations (including photos, oral histories, written descriptions, and field sketches) are evaluated using a confidence assessment framework to compile a history of brine icing deposit presence or absence during summer field seasons between 1903-1904 and 1993-1994. Additionally, an alternative explanation for a small, localized advance of a portion of the terminus is proposed: rather than temperature-driven ice viscosity changes, rising lake level drove temporary, localized basal sliding which induced advance, thinning, and collapse of a part of the terminus previously grounded on a proglacial moraine. In chapter three, time-lapse imagery is used to document a 2014 wintertime brine release that occurred in the absence of surface melt. This suggests that meltwater-driven fracture propagation of surface crevasses downward into the glacier was not a likely factor in this brine release event, as has been previously proposed. Further, there is no evidence for an increase in Rayleigh-wave activity prior to or during the brine release that would be characteristic of shallow seismic sources. Together, this suggests that sufficient pressure is built in the subglacial system to trigger basal crevassing and fracture propagation upward to allow brine release at the surface. In chapter four, two different seismic detectors that use ratios of short-term to long-term seismic energy variance to identify seismic events are compared. The detectors use different statistical distributions to determine what constitutes a large enough ratio to trigger an event detection. Differences between what the two detectors identify as events rather than background noise are interpreted as environmental microseismicity with a distinct diurnal and seasonal occurrence. Minimum detectable event sizes over 3-day time windows are compared. Together, these studies provide context for the history of brine release events, wintertime brine release characteristics, and descriptions of the local seismic environment at Taylor Glacier.