Seismic detection of transient changes beneath Black Rapids Glacier, Alaska

Abstract

To gain new insight into the mechanisms of basal motion, I have demonstrated the feasibility of an active seismic technique to measure temporal changes in basal conditions on sub-hourly time-scales. Using changes observed in the summer of 1993 on Black Rapids Glacier, I have determined part of the basal morphology and the mechanisms of seismic change there. One region of the glacier's bed was monitored daily using seismic reflections, for a period of 45 days. The majority of these reflections were nearly identical. However, the englacial drainage of two ice-marginal lakes and one supra-glacial pothole upglacier of the study site each caused significant anomalies in the daily reflections, as well as dramatic increases in basal motion. Two of these seismic anomalies were nearly identical despite the fact that their drainage events occurred at different locations. Further, these two seismic anomalies were followed by records identical to the non-anomalous state, showing that the changes were seismically reversible. In one of these events, two records taken 36 minutes apart revealed that the transition between the anomalous and normal states occurred completely within this short interval. Reflection arrival times during the anomalies require that a basal layer at least 5 m thick was either created or changed in situ. Reflection amplitudes indicate that such a layer could be either water or a basal till, but water layers of such thickness are not physically reasonable. Published values of wave speeds and densities of till are then compared to those constrained by the observed reflection coefficients. Only a decrease in till saturation can produce the observed changes in reflection amplitudes in the time required. Because the transition from anomalous to normal states can occur in as little as 36 minutes, any mechanisms involving the diffusion of water through a thick till layer are ruled out, such as a change in porosity or pore-water (or effective) pressure. We therefore interpret the cause of the seismic anomalies as due to a temporary decrease in saturation, and propose that such a change may occur quickly and reversibly following a lake drainage by a redistribution of the overburden pressure.