On the Secondary Zone of Interseismic Subsidence and the Mechanics of Subduction

Vertical deformation is the most discriminating observable of the seismic cycle in subduction zones, but its complex measurement and contamination by multiple controlling processes have limited its exploitation. While horizontal displacements primarily reflect interplate coupling, vertical signals encode the competition among elastic megathrust loading, plate tectonic forcing, mantle flow and relaxation, and long-wavelength mass redistribution within the upper plate and forearc. Advances in satellite geodesy now allow vertical deformation rates to be resolved with sufficient precision and spatial coherence to provide new constraints on long-standing conceptual models of interseismic deformation. Recent observations reveal the existence of a secondary zone of interseismic subsidence (SZIS) in Cascadia, Nankai, Japan Trench, and Southern Chile. We use new data and models to assess the persistence of a SZIS and quantify and unravel its physical controls.

Using multi-track InSAR rate maps in northern Chile, we identify an inter-seismic secondary zone of subsidence landward of the primary coastal uplift belt. The presence of this SZIS supports the existence of a persistent secondary zone of interseismic subsidence. However, within the classical backslip framework, the elastic half-space predicts a slow monotonic transition from coastal uplift to inland subsidence. We show that it cannot reproduce the observed secondary trough without invoking unphysical coupling distributions or implausible fault geometries. The discrepancy is therefore not parametric, but conceptual. Interestingly, the existence of a SZIS was first predicted by our numerical seismotectonic models (van Dinther et al., PAGEO, 2019). We use these cross-scale visco-elasto-plastic models to demonstrate the critical role of a visco-elastic lower crust, which allows for an elastic upper crust that is thin enough to bulge under compression transferred across a coupled megathrust. We find that this mechanism is important, but it is not the only relevant mechanism. To quantify and detangle the physical mechanisms in more detail, we build a data-driven model of Northern Chile and aim to explain lateral variations along our observed segment. We integrate high-resolution earthquake catalogue, seismic tomography, and gravity anomaly observations to constrain slab geometry, forearc rheology, density structure, and seismogenic zone dimensions. Our fully dynamic visco-elasto-plastic earthquake cycle model with invariant rate-and-state friction resolves sequences of quasi-periodic earthquakes and can build topography over them. Through that, we aim to explain the presence of a SZIS also in our early geomorphic and geological interpretations of upper-plate deformation. Those multi-scale observations support that vertical surface displacements are not only governed by elastic rebound of megathrust faulting but also include a long-term long-wavelength deformation signal possibly related to position-dependent buckling of the upper plate.

We argue that, together with the existence of a persistent secondary zone of coseismic uplift of the largest earthquakes, such a secondary zone of deformation is a persistent and characteristic feature of seismic cycle deformation in subduction zones. This primary diagnostic will allow for a reinterpretation of the mechanics of subduction through an extension of the canonical backslip surface-deformation model.