Long-term coastal uplift due to non-recoverable forearc deformation during the interseismic phase of the subduction earthquake cycle

Simple elastic dislocation models have been widely used to describe the surface displacements associated with subduction zone earthquake cycles. To first order, these assume a portion of the plate interface is locked during the interseismic period, inducing subsidence in the offshore domain and uplift in the onshore region. In contrast, megathrust earthquakes will impart the opposite surface displacement with offshore uplift and onshore subsidence. Such a purely elastic description of the earthquake cycle implies that interseismic deformation should be entirely compensated by large megathrust earthquakes, amounting to effectively zero deformation over numerous cycles. Recent studies however propose that spatial patterns of interseismic (short-term) deformation are reflected in long-term trends of coastal uplift (Jolivet et al., 2020), as well as in the morphology of subduction margins, which is shaped over 100s of kyrs by the interaction of tectonic and surface processes (Malatesta et al., 2021). This suggests that the repetition of seemingly elastic cycles somehow leads to non-recoverable long-term deformation.

We postulate that a small increment of inelastic deformation accumulates during each interseismic phase, leading to a long-term unbalance of co-, post- and interseismic strain. To test this hypothesis, we evaluate the variations in upper plate stress imparted by down-dip gradients in megathrust locking during the interseismic period in the Chile and Cascadia subduction zones. We add these changes to the estimated background stress state of the upper plate, and assess the extent of frictional yielding within the forearc as a function of interseismic slip deficit and upper plate strength. We find that the onset of yielding in the late interseismic phase coincides with observed areas of microseismicity at these subduction margins, typically located above the downdip end of the locked zone.

We then estimate the permanent surface uplift imparted by this upper plate yielding employing a statistical approach. We model frictional yielding of the forearc as incremental slip on a population of small faults whose spatial distribution reflects the fraction of the interseismic phase duration spent at yield. We further assume that the temporal distribution of these slip follows a Gutenberg-Richter distribution of parameters consistent with the observed microseismicity. Upon summing the displacements due to each of these dislocations, we estimate the irreversible surface displacement field associated with multiple seismic cycle.  This ultimately amounts to permanent uplift concentrated above the transition from freely slipping to fully coupled megathrust, and is consistent with the geometry and rates of long-term uplift recorded in Chile. We also demonstrate how our model can explain the recently reported correlation between location of downdip locking limit and shelf break in many active margins.