No abstract

No abstract

No abstract

In the recent publication (Rössler et al., 2007) we reported on small, but significant, non-shear components for one M L 3.1 earthquake in West Bohemia on 15 October 2000, which persisted even if an anisotropic crust was considered. For data analysis we used P waves recorded at temporary and permantent stations of the WEBNET of the Institute of Geophysics AS CR and of other networks. Unfortunately, only after publication of the results, we were informed that the vertical components of three stations, LBC, STC and VAC, were distorted by clipping (J. Horálek, personal communication) and thus should not be used for source retrieval. Clipping of P-wave forms was invisible at STC and VAC and difficult to detect at LBC. It could be found from an analysis of S waves, which we, unfortunately, did not study. Neither did we know the clipping level. We have therefore repeated calculations ignoring the three clipped traces. Despite omitted traces, the newly obtained source parameters did not differ much from the previous ones (see Table 1). The main differences with results published by Rössler et al., (2007) consist in the reduction of data residual by 0.01, in a small rotation of the rupture plane, and in a small decrease of non-shear components. Nevertheless, the non-shear components remain non-negligible, indicating existence of a crack opening during rupture process. Following our previous argumentation, jackknife (see Fig. 1) and bootstrap (not Table 1. Data residual R, slip inclination δ as well as mean slip inclination δ and standard deviation σ(δ) (the latter two result from bootstrap tests), rupture plane orientation, seismic moment M T (M T = M × 10 14 Nm), and moment-tensor components retrieved by using velocity Model II. Upper line: results from Rössler et al., (2007), lower line: results with the three clipped traces omitted during inversions. Slip Inclination [°] Rupture Plane [°] Moment Tensor [%] Model R δ δ σ(δ) Strike Dip Rake M DC ISO CLVD II 0.06 86.8 86.9 1.2 166 76 −40 3.6 82.0 +4.9 +13.1 II (this paper) 0.05 88.2 88.6 1.1 169 76 −39 3.5 91.4 +5.7 +2.8

SUMMARY Anisotropic material properties are commonly neglected during moment tensor inversion. On the other hand, anisotropy is a widely observed rock property. We show that anisotropy may greatly influence characteristics of moment tensors. For the inversion we apply a method based on amplitude spectra of waveforms in isotropic media. We investigate effects of anisotropy on seismic moment, moment‐tensor components, and apparent slip inclination of dislocation point sources. The direct calculation of moment tensors for shear sources in anisotropic regions shows spurious non‐double‐couple components that may be mistaken as an indication of (apparent) opening or closing of the fracture plane. On the other hand real volumetric components may be increased but also hidden in the presence of anisotropy. These effects as well as the seismic moment depend on the orientation of the elastic tensor relative to the fault plane and the slip direction. If anisotropy is present near the source but isotropy is assumed during inversion, the properties of the moment tensor can still be obtained in a good approximation. In the case where anisotropy extends to the medium along the ray path, only the fault orientation can be successfully retrieved by inverting qP waves to derive the deviatoric moment tensor. The inversions show that retrieved moment tensors can deviate systematically from moment tensors of shear and tensile sources expected in isotropic media. Further complications may arise when qS waves are included in the inversion process. We account for near‐source anisotropy to re‐interpret moment tensors derived for two events at the KTB super deep drill hole, SE‐Germany. The obtained source mechanisms are close to shear faulting although the moment tensors comprise non‐double‐couple components. We interpret the volumetric moment‐tensor components partly as a result of anisotropy. This indicates that for detailed studies of volumetric source components anisotropy should be considered during inversion. In addition, we show that for shear sources in anisotropic media the elastic properties near the source can also be derived from inverted moment tensors in media where anisotropy is restricted to the source region.