HR: 16:15h
AN: T44A-02    [Abstracts]
TI: A 3-D P wave velocity model of the crust and uppermost mantle of the Mariana volcanic arc
AU: * Calvert, A J
EM: acalvert@sfu.ca
AF: Simon Fraser Universitty, Dept of Earth Sciences , Burnaby, BC V5A 1S6 Canada
AU: Klemperer, S L
EM: sklemp@pangea.stanford.edu
AF: Stanford University, Dept. of Geophysics, Stanford, CA 94305 United States
AU: Takahashi, N
EM: narumi@jamstec.go.jp
AF: Japan Marine Science and Technology Centre, 3173-25, Showa-machi, Kanazawa-ku, Yokohama, 236-0001 Japan
AU: Kerr, B C
EM: bckerr@pangea.stanford.edu
AF: Stanford University, Dept. of Geophysics, Stanford, CA 94305 United States
AB: As part of the Izu-Bonin-Mariana subduction factory project, a 3-D seismic refraction survey was acquired in 2002 over the Mariana volcanic arc between 14.5 and 18.5 degrees N using a combination of ocean-bottom seismometers, Reftek receivers on land, and large airgun shots from the R/V Maurice Ewing. First arrival travel times have been combined with similar data from an approximately east-west 2-D airgun refraction profile across the arc acquired by the JAMSTEC R/V Kaiyo using ocean-bottom seismometers spaced at intervals of approximately 5 km. We inverted these 277,000 travel times for a 3-D subsurface P wave velocity model using isotropic iterative first arrival seismic tomography (FAST). Forward travel times were calculated through a 500 m model grid, and the inverted model velocities were recovered in 1000 m cubic cells. The starting velocity model did not vary laterally beneath a rugged seafloor interface, which was fixed in subsequent iterations, and a 0-1250 m-thick sediment layer that was introduced in areas where the water depth exceeded 1000 m. The sediment layer simulated well deposits both on the flanks of the volcanoes and in the deep ocean basin, was updated in subsequent iterations of the inversion, and served to remove unrealistic small-scale structure in the final model. After nine iterations of the non-linear inversion, the RMS traveltime residual was reduced from 0.672 s to 0.127 s, equivalent to a normalised chi-squared value of 1.0. First arrival tomography does not typically recover sharp velocity contrasts at deep interfaces such as the Moho. Therefore in our preliminary interpretations we have employed the 7.6 km/s isovelocity contour as a proxy for the location of the Moho. The thickness of the igneous forearc crust decreases from 14 km in the north of the survey area to 9 km in the south. The Eocene arc, which is no longer active, exhibits an igneous crustal thickness of 21-24 km with much of this variability associated with the topography of the volcanic edifices. An 18-22 km thick igneous crust characterizes the 3-4 Ma active Mariana arc, which is located approximately 40 km west of the inactive Eocene arc. P wave velocities within the upper crust of the active arc appear to be systematically lower than in the inactive arc, by approximately 450 m/s. At a depth of 15 km, in contrast, velocities are around 350 m/s higher in the active arc. These results suggest an evolution of arc structure with increasing age: closure of fractures and porosity in the upper crust through hydrothermal circulation and a reduction in the mafic character of the mid-lower crust, presumably as a result of some degree of crustal differentiation.
DE: 7240 Subduction zones (1207, 1219, 1240)
DE: 8170 Subduction zone processes (1031, 3060, 3613, 8413)
DE: 8180 Tomography (6982, 7270)
DE: 8185 Volcanic arcs
SC: Tectonophysics [T]
MN: Fall Meeting 2005