Quantizing the Complexity of the Western United States Fault System with Geodetically and Geologically Constrained Block Models

Monday, 15 December 2014: 4:30 PM
Eileen L Evans, Harvard University, Earth and Planetary Sciences, Cambridge, MA, United States and Brendan J Meade, Harvard University, Cambridge, MA, United States
Geodetic observations of interseismic deformation provide constraints on miroplate rotations, earthquake cycle processes, slip partitioning, and the geometric complexity of the Pacific-North America plate boundary. Paleoseismological observations in the western United States provide a complimentary dataset of Quaternary fault slip rate estimates. These measurements may be integrated and interpreted using block models, in which the upper crust is divided into microplates bounded by mapped faults, with slip rates defined by the differential relative motions of adjacent microplates. The number and geometry of microplates are typically defined with boundaries representing a limited sub-set of the large number of potentially seismogenic faults. An alternative approach is to include large number of potentially active faults in a dense array of microplates, and then deterministically estimate the boundaries at which strain is localized, while simultaneously satisfying interseismic geodetic and geologic observations. This approach is possible through the application of total variation regularization (TVR) which simultaneously minimizes the L2 norm of data residuals and the L1 norm of the variation in the estimated state vector. Applied to three-dimensional spherical block models, TVR reduces the total variation between estimated rotation vectors, creating groups of microplates that rotate together as larger blocks, and localizing fault slip on the boundaries of these larger blocks. Here we consider a suite of block models containing 3-137 microplates, where active block boundaries have been determined by TVR optimization constrained by both interseismic GPS velocities and geologic slip rate estimates.