A precise hypocenter determination method using network correlation coefficients and its application to deep low-frequency earthquakes

Abstract

A knowledge of the precise locations of deep low-frequency earthquakes (LFEs) along subduction zones is essential to be able to constrain the spatial extent of various slow earthquakes and the underlying physical processes. We have developed a hypocenter determination method that utilizes the summed cross-correlation coefficient over many stations, denoted a network correlation coefficient (NCC). The method consists of two parts: (1) an estimation of relative hypocenter locations for every pair of events by a grid search, and (2) a linear least squares inversion for self-consistent relative hypocenter locations for the initial centroid. We have applied this method to ten LFEs in the Tokai region, Japan. Statistically significant values of NCC indicate the relative locations for many pairs, which in turn determine the self-consistent locations. While the catalog depths are widely distributed, the relocated hypocenters fall within a 2-km depth range, which implies that LFEs in the Tokai region occur on the plate interface, similar to LFEs in western Shikoku.

Non-volcanic Tremor: A Window into the Roots of Fault Zones

“…A similar result has been reported for Mexico (Payero et al, 2008). Although previous studies have argued that tremor is distributed in depth in Japan (e.g., Nugraha and Mori, 2006), these findings and those from other subduction zones contrast with recent results showing that tremor in Japan is concentrated in depth at the plate interface Ohta and Ide, 2008). Does this difference represent a real variation?…”

New Frontiers in Integrated Solid Earth Sciences

“…A broad distribution of tremor has been reported by several sources for Cascadia Kao et al, 2005;. Although previous studies have argued that tremor is distributed in depth in Japan (e.g., Nugraha and Mori, 2006), these findings and those from other subduction zones contrast with recent results showing that tremor in Japan is concentrated in depth at the plate interface Ohta and Ide, 2008). Although previous studies have argued that tremor is distributed in depth in Japan (e.g., Nugraha and Mori, 2006), these findings and those from other subduction zones contrast with recent results showing that tremor in Japan is concentrated in depth at the plate interface Ohta and Ide, 2008).…”

Precise hypocenter distribution of deep low-frequency earthquakes and its relationship to the local geometry of the subducting plate in the Nankai subduction zone, Japan

“…Applying this procedure to every pair of N events yields N × ( N − 1) relative locations, Δ x 12 NCC ,…Δ x ij NCC ,…Δ x NN −1 NCC . Assuming that these values have Gaussian errors, we determine the hypocenter locations of N events x m ( m = 1, …, N ), minimizing where w ij is a weighting factor associated with the maximum NCC [ Ohta and Ide , 2008], and a is a hyperparameter, controlling the weight of prior information of the initial absolute location x i 0 given by the JMA catalog. We select the optimum value of a by minimizing Akaike's Bayesian information criterion (ABIC) [ Akaike , 1980; Yabuki and Matsu ' ura , 1992].…”

“…For such noisy records, a more robust location method is necessary. In a previous study, we developed a new accurate hypocenter determination method using the summation of cross‐correlation coefficients for network stations, termed network correlation coefficient (NCC) [ Gibbons and Ringdal , 2006], and demonstrated its applicability to LFEs [ Ohta and Ide , 2008]. Here, we slightly modify this network correlation method (NCC method) to relocate LFE hypocenters along the whole region of the Nankai subduction zone.…”