Earthquake Seismic

A 70-element broadband array will be deployed for two years, in a 50-km-radius pattern surrounding the volcano (Fig. 1). This array will be used for a variety of analyses, both of Earth structure and seismicity. The 10-15 km station spacing is designed to optimize imaging from the middle crust to slab, and to provide a substantial boost in our ability to detect small earthquakes. During the last two years the PNSN located 768 earthquakes with M > 0 within the array, 92 of which were deeper than 20 km. These should record on most of the array producing 50,000 P arrivals, and a comparable number of S arrival times, with sufficient crossing paths to obtain good resolution to depths of 20 km or more. We will use a code that we developed to simultaneously invert active-source and earthquake travel times of moho- reflected and first-arriving phases for earthquake location, 3-D structure and reflector geometry (Preston et al. 2003), which is based on the 3-D finite difference travel-time codes pioneered by Vidale (1990) and Hole and Zelt (1995). We will also explore the use of double-difference tomography, which is known to increase resolution on volcanoes (e.g Pesicek et al., 2008). Knowledge of the extremely heterogeneous upper crust from the active-source experiment dramatically increases the resolving power at depth from the earthquake data. Recording both P and S waves for identical source receiver pairs allows for robust estimates of 3-D Vp, Vs and Vp/Vs which is critical for addressing questions about partial melt (e.g. De Natale et al. 2004). Attenuation tomography on volcanoes has great potential for illuminating regions of partial melt (Tsumura et al. 2000). We will conduct an experiment with an unprecedented density of stations and sources to increase resolution over previous studies. See Lees 2007 for a thorough review of past experiences in seismic imaging of volcanoes worldwide.