SUBDUCTION CHANNEL MODEL FOR ACCRETION, SEDIMENT SUBDUCTION, AND TECTONIC EROSION AT CONVERGENT PLATE MARGINS

Convergent plate margins display a wide range in tectonic behavior. For example, margins such as Lesser Antilles and Oregon-Washington are sites where broad accretionary prisms have accumulated since the mid-Tertiary. Mariana and Guatemala, on the other hand, are sites where there has been extensive, if not complete sediment subduction and probably even significant subduction erosion during the same time period. Deep-sea drilling and other studies have led to the conclusion that the primary processes that operate at convergent margins are 1) subduction accretion by offscraping at the front of the overriding block and underplating to its base, 2) sediment subduction, and 3) subduction erosion.

The subduction channel model of Shreve and Cloos (1986) is based upon the premise that the descending plate and overriding block at margins where convergence is faster than a few cm/yr are analogous to the guide surface and slide block in a slipper bearing and that subducting sediment is analogous to the lubricant. Complex and varied behavior occurs because the overriding block is not rigid but slowly deforms, the buoyancy of subducting sediment affects the flow, sediment (or upwelled melange) can underplate onto the hanging wall, the sediment supply can vary widely relative to the capacity of the system, and other factors. Quantitative predictions bearing on a wide variety of geological and geophysical observations include the profile of thickness of the subduction channel shear zone, the pattern and speeds of flow in the channel, the stresses exerted on the channel walls, the amount of sediment offscraped and underplated, the zones where subduction erosion is most likely, the amount of sediment subducted to the volcanic arc, the kinematic patterns of flow near the inlet to the shear zone, the formation of subduction melange and other features of convergent plate margins. The subduction channel model does NOT address the same questions as the critical wedge theory of Davis, Dahlen, Suppe and others. Wedge theory and nearly all other quantitative models for convergent margin tectonics focus on the deformation of sediment AFTER its incorporation into an accretionary prism. The subduction channel models focuses on the flow of subducting sediment BEFORE incorporation into a prism or subduction to the depths of arc magmagenesis.

At all margins, some sediment is subducted to the depths of arc magmagenesis and a zone of bulldozer-like action, known as the "zone of compression" is present at the base of the inner trench slope. There are five possible kinematic patterns in the zone of compression (Types A-E), but only three are common. The decollement imaged in seismic reflection profiles is just the lowermost fault in the downward expanding zone of shear. Only fortuitously does it correspond to the boundary separating materials that will become accreted from that which becomes deeply subducted. The amount of bulldozed sediment accreted by is the difference between the amount of incoming sediment (thickness times subduction speed less compaction) compared to the capacity of the margin for dragging sediment past the inlet. The pattern of underplating depends upon the shear stress distribution along the hanging wall, the sediment supply along the channel, and the permeability of the overriding block. Subduction erosion occurs where the incoming sediment supply is sufficiently small that high shear stresses develop on the base of the overriding block and pieces of the block become detached. Whether the state of stress in the overriding block (crystalline overriding plate ± accretionary prisms) is at the point of incipient Coulomb failure for shortening or extension depends upon the magnitude of the shear stress on the hanging wall, the surface slope, and the mechanical properties of the overriding block.