CHANGES IN SHALLOW VT SEISMICITY AND STRESS BETWEEN BACKGROUND AND PRE-ERUPTIVE PHASES (1999-2004), MOUNT ST. HELENS, WASHINGTON

The reawakening of Mount St. Helens began on September 23, 2004, with a swarm of shallow (<1 km below the summit) volcano tectonic (VT) earthquakes. VT seismicity gradually gave way to low frequency and hybrid events towards the end of the vent-clearing phase, which culminated in a series of explosions between October 1 and 5. Systematic changes in the orientation of the deep (4-6 km) stress field measured by fault-plane solutions (FPS) were observed at Mount St. Helens during past activity, motivating a detailed analysis of shallow VT seismicity recorded prior to and during the 2004 eruption. VT earthquakes recorded during a background period between January 1999 and July 2004 and between September 23 and 29, 2004 were repicked and high-quality locations and FPS were determined. FPS for the background period show primarily normal faulting, which continued through September 23. Normal faulting ceased on September 24, when the dominant faulting was strike-slip and reverse, but reappeared on September 25 in addition to continued strike-slip and reverse faulting which continued through September 29. The orientation of maximum compressive stress, estimated from FPS p-axis orientations was found to be WNW-ESE on September 24, heterogeneous on September 25, and WNW-ESE again between September 26 and 29. Results of FPS stress tensor inversions also show an increase in the relative heterogeneity of the stress field on September 25. We suggest a model for the pre-eruption seismic sequence as follows: upward magma migration produced normal faulting and a steady increase in the pressure in the dike on the 23rd. By the 24th the pressure in the dike reached failure and magma broke through the bottom of a seismogenic volume located ~1-2 km beneath the summit, inducing strike slip and reverse faulting and rotating the local stress field to a WNW-ESE orientation. On the 25th the pressure again increased as magma migrated upward inducing a mixture of normal, reverse, and strike slip events and a more heterogeneous local stress field. The pressure approached failure again towards the 29th as magma neared the upper boundary of the seismogenic volume producing a mixture of strike slip and reverse faulting and another rotation of the local stress field to WNW-ESE. The pressure reached failure on October 1 and produced the first explosion of the 2004 eruption.