HR: 1340h
AN: G53B-0878 INVITED     [Abstracts]
TI: A series of transient slip events on Kilauea volcano, Hawaii.
AU: * Desmarais, E K
EM: emilyd@stanford.edu
AF: Stanford University, 397 Panama Mall Dept. of Geophysics, Stanford, CA 94305 United States
AU: Segall, P
EM: segall@stanford.edu
AF: Stanford University, 397 Panama Mall Dept. of Geophysics, Stanford, CA 94305 United States
AU: Miklius, A
EM: asta@usgs.gov
AF: U.S.G.S., Hawaiian Volcano Observatory Hawaii National Park, Volcano, HI 96718 United States
AU: Cervelli, P
EM: pfcervelli@usgs.gov
AF: U.S.G.S., Alaska Volcano Observatory, Anchorage, AK 99508 United States
AB: Deformation on Kilauea volcano, Hawaii is monitored by a network of continuously recording GPS stations, among other methds. Since its installation in 1996, the GPS network has detected four spatially coherent accelerations on Kilauea's south flank that are not caused by either intrusions or earthquakes. These events, each lasting several hours to two days, occurred in September 1998, November 2000, July 2003, and January 2005. Previously, Cervelli et al., (Nature, 2002) interpreted the 2000 event as a silent earthquake due to slip on a sub-horizontal fault beneath Kilauea's south flank. We inverted the cumulative displacements ( less than 2 cm) using a simulated annealing algorithm for each event and found similarly sized, near horizontal, uniform slip source locations for all four events at depths of ~6 km. The estimated slip magnitudes are between 9 and 15 cm, with the upper block moving seaward. The 2005 event is the largest detected to date. Volcano-tectonic (VT) earthquakes on the south flank of Kilauea are typically restricted to the volume between the East Rift Zone and the Hilina and Poliokeawe Palis. Seismicity in this volume increased significantly during the silent events at depths of 5-10 km. However, all of the VT earthquakes were small ( less than M3) and their cumulative moment does not account for the moment released during the silent slip events. We are currently examining seismic waveform data for evidence of other signals, such as non-volcanic tremor, that might be associated with the slip events. To determine the exact onset and duration of the silent earthquakes, we invert for slip as a function of time directly from raw GPS phase and pseudorange observations. The November 2000 silent earthquake was preceded 9 days earlier by nearly 1 m of rainfall, which was speculated in Cervelli et al., (Nature, 2002) to have reduced fault stability through surface loading or pore pressure increase. In contrast, both the 2003 and 2005 events occurred without anomalous rainfall.
DE: 8419 Volcano monitoring (7280)
SC: Geodesy [G]
MN: Fall Meeting 2005