Non-double-couple components of fluid-related earthquakes

It has been reported that non-double-couple (non-DC) components in seismic sources are contained for earthquakes involving fluid, such as earthquakes occurring in volcanic regions or induced earthquakes associated with deep well injection in oil, gas and geothermal fields (e.g., Martínez-Garzón et al., 2017). To understand the underlying physical process of these earthquakes, it is important to estimate moment tensor solutions with high precision. However, their reliable estimates are difficult because of poor knowledge of the underground structure, which may lead to large artificial non-DC components. In order to overcome this problem, we proposed a method that iteratively applies the relative moment tensor inversion (RMTI) (Dahm, 1996) to source clusters improving each moment tensor as well as their relative accuracy. This method successfully solves the problem of RMTI that errors in the moment tensor of reference events lead to biased solutions for other events, while taking advantage of RMTI that the moment tensors can be determined without a computation of Green's function. The procedure is summarized as follows:

(1) Sample co-located multiple earthquakes with source mechanisms, as initial solutions, determined by an ordinary method.

(2) Apply the RMTI to estimate the moment tensor of each event relative to those of the other events.

(3) Repeat the step 2 for the modified moment tensors until the reduction of total residual converges.

We focused on the supraslab earthquake cluster beneath Kanto, Japan, which is induced by fluid injection due to the drainage from the Philippine Sea plate during episodic slow slip (Nakajima and Uchida, 2018). We analyzed 13 earthquakes that occurred between 2010 and 2013, which had magnitude larger than 2.0. The application of the present method revealed that most of events predominate in shear components, but they also have a small non-DC component due to fault compaction with sufficient accuracy. A possible mechanism is that fluid pressure lowers the frictional strength of the fault, so that the shear failure occurs and the fault zone with highly porous material is compacted at the same time. The sense of fault compaction seems to be natural consequence considering the high confining pressure of the depth where these earthquakes occurred (∼30 km).