Vertical Crustal Deformation Due To Viscoelastic Earthquake Cycles at Subduction Zones: Implications for Nankai and Cascadia.

Despite significant progress in studying subduction earthquake cycles, the vertical deformation is still not well understood. Here, we use a generic viscoelastic earthquake‐cycle model that has recently been validated by horizontal observations to explore the dynamics of vertical earthquake‐cycle deformation. Conditioned on two dimensionless parameters (i.e., the ratio of earthquake recurrence interval T to mantle Maxwell relaxation time tM (T/tM) and the ratio of downdip seismogenic depth D to elastic upper‐plate thickness Hc (D/Hc)), the modeled viscoelastic deformation exhibits significant spatiotemporal deviations from the simple time‐independent elastic solution. Caution thus should be exercised in interpreting fault kinematics with vertical observations if ignoring Earth's viscoelasticity. By systematically exploring these two parameters, we further investigate three metrics that characterize the predicted vertical deformation: the coastal pivot line (CPL), the uplift zone (UZ) landward above the downdip seismogenic extent, and the secondary subsidence zone (SSZ) in the back‐arc region. We find that these metrics can all be time‐dependent, subject to D/Hc and T/tM. The CPL location and the UZ width are mainly controlled by D/Hc and T/tM, respectively. The presence of the SSZ is prevalent during the interseismic phase due to viscous mantle flow driven by ongoing plate convergence. Contemporary vertical deformation in Nankai and Cascadia is largely consistent with the model predictions and features differences mainly related to contrasting D/Hc values in the two margins. These findings suggest that vertical crustal deformation bears fruitful information about subduction‐zone dynamics and is potentially useful for inversions of key subduction‐zone parameters, deserving properly designed monitoring. Plain Language Summary: The community studying subduction earthquake deformation usually focuses on the horizontal component due to its strong signals and low observation errors. However, over the past decade, it has increasingly been realized that the vertical deformation is often more diagnostic of megathrust faulting and rheological structure. Additionally, communities studying glacial isostatic adjustment, sea‐level rise, surface and mantle processes and beyond are recently working hard to refine their models by teasing out the earthquake component, which is thought to contaminate their vertical observations. To address these demands, we use computer models validated by the horizontal observations to simulate the vertical movement of the land during the period between great subduction zone earthquakes. We find that the earthquake and subsequent fault locking can cause vertical movement of the surface—subsidence near the coast, uplift landward and subsidence again in the far field. These three features can vary in deformation rates and location/area depending on how deep the fault is locked, how weak the mantle is, and how much time has elapsed since the earthquake. Therefore, previous vertical observations and newly deployed instruments designed to detect key signals are essential for understanding earthquakes, tsunamis and the dynamics of the subduction system. Key Points: On earthquake (EQ)‐cycle time scale, coastal pivot line (CPL), uplift zone (UZ) and secondary subsidence zone (SSZ) are key vertical featuresLocation of the CPL is indicative to downdip seismogenic extent and may migrate landward or seaward depending mainly on the mantle rheologyThe SSZ, as observed in Nankai and Cascadia, is due to interseismic viscous mantle flow and are affected by EQ elapsed and recurrence time [ABSTRACT FROM AUTHOR]