Okal, E. A., and C. R. Bina, On the cessation of seismicity at 660 to 690 km, Eos, Transactions of the American Geophysical Union, 78, Spring Supplement, S215, 1997.
S32A-7
We update the results of Stark and Frohlich [1985] and Rees and Okal [1987] concerning the depth of the deepest subduction zone earthquakes. We provide a more complete and precise dataset by (i) including recent PDE and ISC entries to the end of 1995; (ii) using BCIS listings of historical events not available to Rees and Okal; and (iii) providing realistic error bars on the hypocentral depths (especially of historical earthquakes) by a Monte Carlo procedure consisting of injecting Gaussian noise into the data. The various subduction zones are found to fall into three categories: (1) In the Tonga-Kermadec area, seismicity extends down to 685-690 km; (2) In a large group of subduction zones, seismicity ceases at 650-660 km. This includes Sangihe, Marianas, Banda Sea, Java, Peru, Bolivia and Kuriles as well as the recumbent Vityaz group, and the isolated events in Colombia and Spain; (3) In other subduction zones, including Argentina and Solomon Islands, seismicity does not reach deeper than 615 km.
This dataset strongly suggests that the depth of maximum seismicity is controlled by the second mantle transition, which would be expected to be significantly deflected downwards from 660 km only in the case of the Tonga slab, due to its very fast sinking rate. A preliminary estimate of the required thermal anomaly in the Tonga slab, would be an extra -300 to -400 K, which is in general agreement with thermal models.
Finally, we address the problem of the aseismicity of the lower mantle. Under the assumption that there may exist a mechanism of seismogenesis in the lower mantle (such as metastability during further phase transformations), we use extrapolation of frequency-moment relations (beta-values) to quantify the maximum level of seismicity in the lower mantle compatible with the failure to observe a single event below 690 km since the development of adequate instrumentation. We conclude that the lower mantle is at least 3 orders of magnitude less seismic than the transition zone.
Copyright © 1997 American Geophysical Union