Volatiles that are transported by subducting lithospheric plates to depths greater than 100 km are thought to induce partial melting in the overlying mantle wedge, resulting in arc magmatism and the addition of significant quantities of material to the overlying lithosphere. Asthenospheric flow and upwelling within the wedge produce increased lithospheric temperatures in this back-arc region, but the forearc mantle (in the corner of the wedge) is thought to be significantly cooler. Here we explore the structure of the mantle wedge in the southern Cascadia subduction zone using scattered teleseismic waves recorded on a dense portable array of broadband seismometers. We find very low shear-wave velocities in the cold forearc mantle indicated by the exceptional occurrence of an 'inverted' continental Moho, which reverts to normal polarity seaward of the Cascade arc. This observation provides compelling evidence for a highly hydrated and serpentinized forearc region, consistent with thermal and petrological models of the forearc mantle wedge. This serpentinized material is thought to have low strength and may therefore control the down-dip rupture limit of great thrust earthquakes, as well as the nature of large-scale flow in the mantle wedge.
Abstract. This study examines thermal and structural controls of the updip and downdip rupture limits of great subduction thrust earthquakes. Data on past great earthquake seismic limits have been compiled for four continental subduction zones, Cascadia, SW Japan, south Alaska, and Chile. These limits have then been compared to the predictions of several models for what constrains great earthquake rupture. Temperatures on the subduction thrusts have been estimated by finite element numerical models. The landward limits of the observed updip aseismic zones correspond to the position where the thrust temperature reaches about 100øC, that is, depths of about 2 to 10 km for the subduction zones studied. This temperature agrees with the dehydration of stable sliding smectite clays to illite-chlorite. The temperatures in this region are controlled mainly by the thickness of sediment on the incoming crust and by the crustal age and thus heat flow. The downdip limits correspond to the depth on the thrust where either (1) the temperature reaches about 350øC, which corresponds to thermally activated stable-sliding behavior for crustal rocks (with a transition to 450øC), or (2) about 40 km depth if 350øC is reached at greater depth. Depths of about 40 km approximately correspond to the intersection of the thrust with the continental forearc Moho, and this downdip limit may be a consequence of stable-sliding serpentinite or talc and other hydrated forearc mantle rocks. The primary temperature controls on the downdip region are the age of the subducting oceanic plate and the thrust dip profile. Secondary control comes from the thickness of incoming sediment, the convergence rate, and the radioactive heat generation in the overlying forearc. The 100øC updip limit occurs near the trench for young subducting plates with a thick sediment section such as Cascadia (6-8 Ma), and up to 80 km landward for older oceanic crust such as south Alaska (-50 Ma). The 350øC downdip thermal limit is applicable for young oceanic plates (e.g., Cascadia and Nankai), whereas the forearc mantle limit applies for older plates (e.g., south Alaska and Chile except near the Chile Rise). For the margins studied that have experienced great earthquakes, there is generally good agreement between the postulated thermal and Moho limits and the rupture or seismogenic zone as defined by the distribution of aftershocks and by waveform, tsunami and dislocation modeling. The downdip limit of the interseismic locked zone from dislocation modeling is also in agreement with these limits.
Subduction thrust faults generate earthquakes over a limited depth range. They are aseismic in their seaward updip portions and landward downdip of a critical point. The seaward shallow aseismic zone, commonly beneath accreted sediments, may be a consequence of unconsolidated sediments, especially stable-sliding smectite clays. Such clays are dehydrated and the fault may become seismogenic where the temperature reaches 100-150°C, that is, at a 5-15 km depth. Two factors may determine the downdip seismogenic limit. For subduction of young hot oceanic lithosphere beneath large accretionary sedimentary prisms and beneath continental crust, the transition to aseismic stable sliding is temperature controlled. The maximum temperature for seismic behavior in crustal rocks is -350°C, regardless of the presence of water. In addition, great earthquake ruptures initiated at less than this temperature may propagate with decreasing slip to where the temperature is -450°C. For subduction beneath thin island arc crust and beneath continental crust in some areas, the forearc mantle is reached by the thrust shallower than the 350°C temperature. The forearc upper mantle probably is aseismic because of stable-sliding serpentinite hydrated by water from the underthrusting oceanic crust and sediments. For many subduction zones the downdip seismogenic width defined by these limits is much less than previously assumed. Within the narrowly defined seismic zone, most of the convergence may occur in earthquakes. Numerical thermal models have been employed to estimate temperatures on the subduction thrust planes of four continental subduction zones. For Cascadia and Southwest Japan where very young and hot plates are subducting, the downdip seismogenic limit on the subduction thrust is thermally controlled and is shallow. For Alaska and most of Chile, the forearc mantle is reached before the critical temperature, and mantle serpentinite provides the limit. In all four regions, the seismogenic zones so defined agree with estimates of the extent of great earthquake rupture, and with the downdip extent of the interseismic locked zone.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.