In recent decades, advancements in geodetic coverage and technology have significantly enhanced our ability to document surface deformation in subduction zones, during and between large earthquakes. Crucially, analyzing interseismic displacements allows us to identify locked regions along the subduction zone interface. These areas are more prone to rupture in subsequent megathrust events, the most devastating earthquakes on Earth. Despite these important improvements, one limitation is that geodetic data is constrained by its short temporal span, usually covering only decades, that is: a small fraction of the duration of seismic cycles. The question of whether the spatial pattern of megathrust locking persists or evolves over multiple seismic cycles thus remains unanswered.

Subduction zone landscapes, on the other hand, record much longer timescales, providing a unique opportunity to examine deformation over hundreds of thousands of years, far exceeding the duration of individual seismic cycles. Interestingly, the patterns of interseismic surface uplift recorded over decades resemble the field of rock uplift that have shaped forearc landscapes over hundreds of thousands of years. These long-term signals are constrained by a growing body of geomorphological work, including the edges of continental shelves, the distribution of marine terraces, profiles of coastal rivers, fluvial incision rates and others. The correlation between the short- and long-term uplift fields suggests that locking patterns not only influence earthquake behavior but also leave a lasting imprint on Earth's topography. However, we often characterize off­-fault deformation associated with repeating seismic cycles as purely elastic, implying that aside from slip on the megathrust, no permanent strain—and consequently, no long-­term rock uplift is expected. This raises the question of how short-term locking patterns become encoded in the permanent landscape features we observe today, a connection that remains poorly understood.

Our research proposes that variations in the degree of megathrust locking induce not only elastic strain but also increments of non-recoverable brittle deformation in the overriding plate, manifested primarily through interseismic upper plate seismicity. We compute the accumulated deformation imparted by this distributed seismicity over hundreds of earthquake cycles, and demonstrate it results in subtle but discernible patterns of uplift resembling interseismic displacements. The upper panel in the figure below illustrates how the concentration of  diffuse upper plate seismicity and long-term uplift are greatest where the gradient in slip deficit is sharpest, resulting in long-term permanent surface displacements similar to the interseismic field of uplift. The lower panel in the figure shows how upper plate off-fault deformation involves a small degree of nonrecoverable interseismic deformation, superimposed on the dominant elastic component. Our findings thus suggest that megathrust locking is stable over multiple earthquake cycles, highlighting the role geomorphology can play in constraining Earth’s greatest source of seismic hazard.

Bar Oryan et al., Megathrust locking encoded in subduction landscapes.Sci. Adv.10,eadl4286(2024).DOI:10.1126/sciadv.adl4286