A viscoelastic and afterslip postseismic deformation model for the 1964 Alaska earthquake

We developed a 3-D viscoelastic model, in concert with an afterslip model, to describe the postseismic deformation following the 1964 Alaska earthquake. Our model incorporates a realistic geometry including an elastic slab with very low dip angle. These geometric factors are important and require a reanalysis of the 1964 coseismic model. Our coseismic model differs from previous models in that the Montague Island splay fault extends farther along the Kenai Peninsula coast, and as a result, slip on the megathrust in that region is smaller. We computed postseismic deformation models using a range of mantle viscosities with Maxwell relaxation time tau of 1 to 50 years, with a best estimate of 20 years. The viscoelastic model explains most of the trenchward motion observed in the present velocity field but has little impact on cumulative 1965 to present uplifts. After removing the viscoelastic response, the largest residual cumulative uplifts were located downdip from areas of largest coseismic slip area, and we explain them using an afterslip model, constrained in space by the observed cumulative postseismic uplifts and in time by tide gauge records. No single mechanism can explain both the 30 years of cumulative uplift and the present velocities, but a combination of viscoelastic relaxation, afterslip, interseismic elastic deformation, and motion of southern Alaska relative to North America explains the first-order features of the observations. Forty years after the earthquake, the present-day velocities contain a significant component of postseismic deformation, so very long lived postseismic deformation plays an important role in the subduction zone earthquake cycle for great earthquakes.