Linking Paleogeodesy Observations with Geodynamic Models of Megathrust Earthquakes: Implications for Cascadia and Other Megathrust Zones

The use of geologic records of crustal motions (Paleogeodesy) is becoming a more important component in unravelling details of plate boundary coupling. This requires (a) an understanding of the conditions under which the geologic record is produced and preserved, and (b) how the expected pattern of vertical motions vary in time and space within the upper plate of subduction zones - the former is the domain of the paleo-geodesist, the latter falls to geodynamics. Recent models of crustal deformation in response to the earthquake cycle in megathrusts show that small variations in the location of the observing site and the time scale recorded by observations will have a large effect on the signal and its fidelity in recording megathrust processes such as coupling and rupture. The earthquake cycle produces distinctive patterns of crustal displacements that reflect details of the pattern of plate coupling and rupture behavior. Patterns of horizontal deformation are relatively simple, and have been successfully used to define current coupling patterns. The patterns of vertical motions (which are recorded in the paleo-geodetic data) are more complex, with the direction of motion (up/down), its amplitude, and rate varying in space and time through the earthquake cycle. This makes observations of vertical motions complicated but potentially more diagnostic of parameters such as coupling and rupture. Vertical motions preserved in the geologic record reflect some combination of coseismic and short-term post-seismic motions (including after-slip and early viscous relaxation), both of which vary significantly in the vicinity of the down-dip limit of coupling and rupture. Because of the substantial signal from immediate post-earthquake processes (after-slip?), models utilizing only elastic models of co-seismic motions are likely to mis-estimate fault slip or the extent of fault rupture. Here we incorporate the effects of visco-elastic and other non-elastic processes to explore the implications of these effects, calibrated by observations after the 2004 Banda Aceh and 2011 Tohoku great earthquakes, to improve our understanding of the rupture patterns of great earthquake events such as 1700 Cascadia.