Variations in pore-fluid pressure along subduction megathrusts are often invoked to explain differences in fault slip behavior between margins. However, many other parameters vary between subduction zones, making it difficult to isolate the causes of elevated pore fluid pressures and their effects on megathrust behavior. Here we show evidence from pre-stack depth migrations of multichannel seismic reflection data along the subduction zone off the Alaska Peninsula for significant, systematic along-strike variations in the thickness, continuity, P-wave velocity, and estimated pore-fluid pressure of the subducting sediment layer that lines the shallow plate interface within 25 km of the trench. These variations appear to correlate with changes in seismicity, locking, and earthquake rupture history. The currently locked and seismically quiet Semidi segment has a continuous, thick (600–900 m) subducted sediment layer that is characterized by low seismic velocities and elevated pore pressure (λ* ≈ 0.4–0.8). The subducted sediment in the neighboring Shumagin Gap, a region with low geodetic coupling and abundant small earthquakes, is thinner, irregular and has lower pore pressure (λ* < 0.2–0.3). We suggest that the thicker and weakly faulted sediment layer entering the trench at Semidi is associated with a continuous and overpressured sediment layer lining the shallow plate interface, but forms a large coherent asperity as it dewaters and consolidates at greater depths, thus favoring large earthquakes. In contrast, thinner incoming sediment disrupted by outer rise faulting at Shumagin results in a plate interface without elevated pore-fluid pressures, but which is heterogeneous and complex at all depths, contributing to creep and frequent small earthquakes.

You do not currently have access to this article.