Crustal Cracks in Areas of Active Deformation: Correlation of GPS and Seismic Anisotropy

Seismic anisotropy in the upper crust can be observed from shear-wave split- ting. It is closely related to crack distribution, orientation and density via the orientation of fast polarisation and the delay time between the two perpendic- ular components of the original shear-wave. Since stress variations can a ect crustal cracks, they should change shear-wave splitting. Another observable result of variations in stress can be deformation measured by GPS (Global Positioning System). Although there are as yet few publications linking these di erent methods, some have suggested an alignment of fast direction with maximum horizontal compressive stress and maximum horizontal compressive strain. We examine whether we can observe this or a di erent relation of seis- mic anisotropy to strain and stress changes in three di erent settings. We performed shear-wave splitting analyses of local earthquakes and baseline and strain calculations around Taupo caldera (New Zealand), Aso caldera (Japan), and around an area on the Raukumara peninsula (New Zealand) associated with slow slip on the Hikurangi subduction interface. Both anisotropy and deformation vary with time in all three regions, but the time variations do not strongly correlate with each other. We suggest that a strong deformation signal observed at Taupo caldera might have a regional, non-volcanic source, and small variations in shear-wave splitting cannot be linked to variations in GPS time series or baselines. At Aso caldera strong deformation seems to be directly related to volcanic activity. Seismic anisotropy also shows a change, but at a slightly di erent time than the GPS signal. On the Raukumara peninsula, the strong deformation associated with slow slip does not show up as a variation in seismic anisotropy, although variations in shear-wave splitting do exist in this area. Overall, we observe an alignment of fast direction with either maximum horizontal compressive strain or stress or both for a subset of time periods and stations. In addition, there is a contribution of structure to the observed anisotropy. We conclude that deformation and seismic anisotropy cannot always be linked in a straightforward way. Instead, shear-wave splitting may be connected to smaller scale processes than can be detected by the current densities of the GPS networks.