Elastic Slab in Viscoelastic Mantle: Effects on Determining Megathrust Slip and Mantle Viscosity During Postseismic and Interseismic Phases

Abstract

AbstractEarthquake‐cycle deformation at subduction zones is not only induced by cyclical faulting processes but also impacted by the Earth structure and rheology. Oversimplifying the first‐order Earth structure in models leads to biased estimations of fault kinematics and Earth rheology and hence misinterpretation of subduction‐zone dynamics. Here we first use synthetic models to investigate the impacts of an elastic slab on viscoelastic postseismic and interseismic deformation and on determining afterslip, locking distribution, and mantle viscosity. We confirm that the presence of a slab causes anchored viscous flow in the mantle wedge and less extensive viscoelastic deformation in the overriding plate than the depth‐dependent layered models without a slab during the interseismic and postseismic phases. By assuming the predicted viscoelastic deformation from the model with a slab as the synthetic true observations, layered models obtain a higher viscosity and more slip on the downdip of the seismogenic zone than the true model during both phases. These biased estimates are found varying with the displacement component, spatial and temporal coverages of the used observations. We then conduct two real‐world case studies at the Japan subduction zone and find that the layered model obtains the biased fault kinematics and mantle viscosity that are similar to the synthetic modeling results, indicating that the slab signature may have been observed in the geodetic data. Our findings highlight the importance of including a slab in modeling the earthquake‐cycle deformation and thoroughly exploring the sensitivities of surface deformation, especially the verticals, on the deep Earth structures.