[S01] PM-1

[S01-13] Where have all the isotropic components gone?

*Hitoshi KAWAKATSU^1,2 (1. Earthquake Research Institute, The University of Tokyo, 2. Institute of Earth Sciences, Academia Sinica)

As the earth's lithosphere that hosts earthquakes consists of seismically anisotropic fabrics, the seismic moment tensors naturally posses the isotropic components in addition to the deviatoric non-double couple components (Aki and Richards, 1980; Kawasaki and Tanimoto, 1981). The "zero-trace" constraint in the moment tensor inversion, often employed by the standard catalogues, such as GCMT, USGS, etc., has obscured this problem of the isotropic component. Kawakatsu (2023, SSJ; 2024, BSSA, in press) showed how the isotropic component might be converted to the vertical CLVD component in the GCMT solutions of shallow earthquakes to infer the anisotropic structure (VTI) of the lithosphere. Whether or not such an approach is applicable to deeper earthquakes in the subducting lithosphere seems to be a natural extension as the previous research on this problem did not consider the consequence of the presence of the isotropic component carefully. Vavrycuk (2004), in his pioneering work to connect NDCs to the Tonga deep slab anisotropy, confirmed that his model did not produce significant isotropic components in the moment tensors, indicating his approach is internally consistent; on the other hand, such care is absent in the work of Li, Zheng et al. (2018) who reported the presence of huge anisotropy (~25% anisotropy of S-wave in TTI) in the subducting lithosphere of the globe. Some of the models may produce >20% isotropic NDCs in moment tensors, which appears inconsistent with the relatively small (<10%) isotropic components of full-moment tensor inversion results for deep earthquakes (Kawakatsu, 1991, 1996). A preliminary modeling of events in the Tonga slab, employing the methodology developed by Kawakatsu (2024) but still assuming that GCMT solutions represent the deviatoric part of the full moment tensors as done in the previous research, indicates much weaker anisotropy may explain the NDCs of the reported moment tensors. However, until the issue where the isotropic components reside in GCMT solutions is fully resolved, the answer might be blowing in the wind.