Simultaneous GPS and teleseismic monitoring of glacial stick-slip of Whillans Ice Stream, West Antarctica.

The TIDES experiment is a coordinated effort to measure and model the Siple Coast ice streams using GPS instruments, local and teleseismic seismographs, radio echo sounding data, and active-source seismic data. The experiment grew out of the discovery that the flow speed and the local seismicity of these ice streams (Whillans, Kamb, Bindschadler, and MacAyeal, originally ice streams B, C, D, and E, respectively) are controlled or modulated by the local ocean tide beneath the Ross Ice Shelf. Arrays of GPS and seismic instruments were deployed on the ice streams. Radio echo sounding of the subglacial environment, allied to numerical modeling of the base as a frictional material, have resulted in a new view of the stability of these important conduits of West Antarctic ice to the ocean. In addition, analysis of broadband seismic data shows that changes in the motion of the Whillans Ice Stream (ice stream B) can be detected at great distances. The ice streams of the Siple Coast are grossly similar in size, surface slope, and average flow speed, but remarkably different in detail. The downstream part of ice stream B flows primarily by stick slip motion, with the times and magnitude of slip related to the tides at the grounding line. The partitioning between rapid slip and slow inter-slip deformation is variable between different parts of the ice stream. The slip timing and to a lesser extent the slip magnitude is strongly correlated to the tide cycle, with a complex relationship to the tide height. Ice stream D and E do not undergo stick-slip, however there is significant modulation of the flow speed of these glaciers with the tide. In all these ice streams, the tidal effect can be detected far inland behind the grounding line. The slip of ice stream B can be detected at seismographs around the Antarctic continent, and even in Australia. Alhough each slip is ~25 minutes long, we detect three main packets of seismic energy of durations of ~100 seconds at periods of 20-100 seconds. Time correlation with the GPS observations identifies the first packet with the initial rupture nucleation and the final packet with rupture termination at the grounding line. The seismic amplitudes are correlated with the spring-to-neap tidal cycle and with rupture velocity. We suggest that the slip events can be monitored using permanent seismic stations, allowing us to detect changes over a longer time period than is possible with in-situ measurements. More generally, other glaciers and ice streams can be remotely monitored for fast glacial slip using seismic detection techniques.