P-WAVE ARRAY POLARIZATION ANALYSIS AND EFFECTIVE ANISOTROPY OF THE BRITTLE CRUST

P-wave polarization constrains local anisotropy in the vicinity of the receivers. Using three-component and array data from the regional GERESS array in south-eastern Germany, we measure polarization a and propagation vectors s for P phases of 120 events. Angular deviations a-s between these normalized vectors often approach 10-degrees, rendering them easily measurable. The effect of anisotropy can be distinguished from remote effects, since all remote effects, such as source mislocation, distant lateral heterogeneity or distant anisotropy, affect polarization and propagation vectors simultaneously. Averaging removes sensitivity to near-receiver heterogeneity, and local anisotropy is left as the sole cause of the effects in a-s. This method hence gives local effective anisotropy in the near-receiver crust averaging over a depth interval of a wavelength (almost-equal-to 6 km). We resolve strike and dip of the symmetry plane and also two dimensionless numbers eta and tau which give constraints on four of the elastic parameters. The optimum model (variance reduction 44 per cent) has symmetry plane orientation of strike 113-degrees and dip 49-degrees to the north, which corresponds closely to the consistently observed gneiss foliation direction in the area (120-degrees, 50-degrees-60-degrees). Comparing eta and tau with predictions from different physical models we find that the data are fit by a gneiss model assuming that the anisotropy is dominated by the mica, if 3-8 per cent of the mica are well aligned. This suggests that anisotropy in the region studied is dominated by the effect of local foliation rather than the regional stress field.