HR: 0800h
AN: S31A-0273    [Abstracts]
TI: Influence of the seismic noise characteristics on noise correlations
AU: Krüger, F
EM: kruegerf@geo.uni-potsdam.de
AF: University of Potsdam, Institut fur Geowissenschaften Postfach 60 15 53, Potsdam, DEU D-14415
AU: * Pedersen, H A
EM: Helle.Pedersen@obs.ujf-grenoble.fr
AF: University of Potsdam, Institut fur Geowissenschaften Postfach 60 15 53, Potsdam, DEU D-14415
AU: * Pedersen, H A
EM: Helle.Pedersen@obs.ujf-grenoble.fr
AF: GeoForschungsZentrum Potsdam, Section 2.4 Telegrafenberg, Potsdam, D-14473 Germany
AU: * Pedersen, H A
EM: Helle.Pedersen@obs.ujf-grenoble.fr
AF: University of Grenoble, LGIT BP 53, Grenoble, 38041 France
AB: Cross-correlating seismic noise to obtain the Green function between two seismic sensors is a promising new technique for crustal tomography. We use data from 38 temporary broadband sensors in Finland to study how the nature of the seismic noise influences noise correlations. The study area is particularly well adapted for this study as the lateral variations of surface wave velocities are very small. We apply a processing technique which makes it possible to extract broadband signals between 2s and 35s period without applying any signal clipping. The extracted Green's functions are symmetrical at low and high-frequencies, but are strongly direction dependent around the two microseismic peaks (at approximately 15s and 8s). For these periods the signal amplitude varies by a factor of 90 between different azimuths. The maximum amplitudes are obtained for east-west profiles for the first microseismic peak and for east-west and north-south profiles for the second microseismic peak. The phase velocities when defined are however correct for all azimuths when all the traces in a given direction are used for slant stack. The group velocities calculated on individual traces do on the contrary vary strongly with azimuth, with a 1/cos dependence. We also apply f-k analysis on data from the NORSAR (Norway) and Gräfenberg (Germany) arrays as well as a small-aperture array in Finland to constrain the arrival directions of the coherent part of the noise. The microseismic noise generation is located west of the array for the first microseismic peak and west and north of the array for the second microseismic peak, in excellent agreement with the noise correlations. The f-k analysis also confirms that the noise generators are more randomly distributed for the long period (25s-50s) and short period (2s-4s) part of the noise. We conclude that in some frequency bands the noise in the study area is dominated by plane energy wavefronts, with strong phase perturbations. The profile direction must be carefully selected in these frequency bands to obtain valid group velocities for tomography.
DE: 7200 SEISMOLOGY
DE: 7255 Surface waves and free oscillations
DE: 7270 Tomography (6982, 8180)
SC: Seismology [S]
MN: Fall Meeting 2005