Frequency-size distributions of Wadati-Benioff zone and near-boundary intraplate earthquakes: Implications for intermediate and deep seismicity

We conduct a systematic analysis of the frequency-moment distribution of earthquake populations in WadatiBenioff zones, motivated by a three-fold increase in the available GlobalCMT database since the previous study by Okal and Kirby (1995), in which some datasets may have been undersampled. We use 100-km depths bins, processed with both least-squares and maximum-likelihood algorithms, and apply an F-test methodology to assess the robustness of our results, in particular regarding multi-segment behavior within individual depth bins. Our results generally support Okal and Kirby's (1995), and the new enlarged dataset suggests the resolution of the third segment predicted by these authors for the 500-600 km bin in Tonga, in the framework of their threetier model for two-step source saturation of seismicity generated by transformational faulting within a wedge of metastable olivine. In the 200-300 and 300-400 km depth bins, we find comparable behaviors, revealing the artificial nature of the traditional 300-km boundary between intermediate and deep earthquakes. Our view is that it may be deeper than 300 km, and fundamentally variable between distinct subduction zones. We extend our study by considering frequency-moment distributions for intraplate earthquakes in oceanic lithosphere before it subducts at trenches. Those results support the interpretation of seismicity in the shallowest portions of slabs as involving the reactivation of existing faults under a process of dehydration embrittlement. However, outer rise events occurring immediately before subduction feature different population statistics, which are probably related to the release of bending stresses of a different origin.

Automated summary

A systematic analysis of earthquake populations in WadatiBenioff zones was conducted, revealing multi-segment behavior and supporting previous research findings. The study suggests that the traditional 300-km depth boundary between intermediate and deep earthquakes may be artificial and depth boundaries could vary between different subduction zones.