HR: 1340h
AN: S43B-1306    [Abstracts]
TI: New Codes for Ambient Seismic Noise Analysis
AU: * Duret, F
EM: fduret@usgs.gov
AF: Polytech Paris UPMC, Bat. Esclangon 4 place Jussieu Case Courrier 135, Paris, 75005, France
AU: * Duret, F
EM: fduret@usgs.gov
AF: US Geological Survey, 345 Middlefield Rd. MS 977, Menlo Park, CA 94025, United States
AU: Mooney, W D
EM: mooney@usgs.gov
AF: US Geological Survey, 345 Middlefield Rd. MS 977, Menlo Park, CA 94025, United States
AU: Detweiler, S
EM: shane@usgs.gov
AF: US Geological Survey, 345 Middlefield Rd. MS 977, Menlo Park, CA 94025, United States
AB: In order to determine a velocity model of the crust, scientists generally use earthquakes recorded by seismic stations. However earthquakes do not occur continuously and most are too weak to be useful. When no event is recorded, a waveform is generally considered to be noise. This noise, however, is not useless and carries a wealth of information. Thus, ambient seismic noise analysis is an inverse method of investigating the Earth's interior. Until recently, this technique was quite difficult to apply, as it requires significant computing capacities. In early 2007, however, a team led by Gregory Benson and Mike Ritzwoller from UC Boulder published a paper describing a new method for extracting group and phase velocities from those waveforms. The analysis consisting of recovering Green functions between a pair of stations, is composed of four steps: 1) single station data preparation, 2) cross-correlation and stacking, 3) quality control and data selection and 4) dispersion measurements. At the USGS, we developed a set of ready-to-use computing codes for analyzing waveforms to run the ambient noise analysis of Benson et al. (2007). Our main contribution to the analysis technique was to fully automate the process. The computation codes were written in Fortran 90 and the automation scripts were written in Perl. Furthermore, some operations were run with SAC. Our choices of programming language offer an opportunity to adapt our codes to the major platforms. The codes were developed under Linux but are meant to be adapted to Mac OS X and Windows platforms. The codes have been tested on Southern California data and our results compare nicely with those from the UC Boulder team. Next, we plan to apply our codes to Indonesian data, so that we might take advantage of newly upgraded seismic stations in that region.
DE: 7223 Earthquake interaction, forecasting, and prediction (1217, 1242)
DE: 7230 Seismicity and tectonics (1207, 1217, 1240, 1242)
DE: 7294 Seismic instruments and networks (0935, 3025)
DE: 8120 Dynamics of lithosphere and mantle: general (1213)
DE: 8124 Earth's interior: composition and state (1212, 7207, 7208, 8105)
SC: Seismology [S]
MN: 2007 Fall Meeting