Extracting the elastodynamic 2D Green's tensor from noise correlations: a numerical study

Abstract

The correlation of the coda of seismic records and the measurement of seismic noise are nowadays important as they can be used to obtain the physical characteristics of the medium. The extraction of Green's tensor between two points is possible by averaging the cross correlation of the records at those points. We explore numerically Green's tensor retrieval in the time domain using canonical 2D elastic cases in order to evince the recovery of body waves under controlled circumstances. To this end, parametric analyses are performed using up to 5000 random sources. The isotropy of the illumination and attenuation are modified with noticeable effects on the retrieved signals. Several source distributions are used and reveal the importance of isotropy of illumination on the retrieved Green's tensor.