SPEAKER: Thomas Bodin (ENS Lyon)
DATE: 9:00 AM, Thursday, June 11, 2020
TITLE: Transdimensional Estimation of Surface Strain Rates from GPS Measurements: Application to California
ABSTRACT:
Investigating actively deforming areas often relies on strain rate calculations using geodetic data such as GPS velocities. Producing a continuous strain rate map from discrete data is an inverse problem traditionally tackled with standard interpolation schemes. However, most algorithms rely on arbitrarily user-defined regression parameters that directly determine the smoothness of the recovered velocity field, and hence the amplitude of its spatial derivatives, resulting in biased estimates of the strain rates and their uncertainties.
Here we propose a transdimensional Bayesian method to estimate surface strain rates from GPS measurements. We parametrize the velocity field with a variable number of Delaunay triangles, and use a Markov chain Monte-Carlo algorithm to sample the probability distribution for the surface velocity and its derivatives. We illustrate this Bayesian approach on synthetic tests mimicking real geophysical settings and GPS data distributions. Finally, using GPS velocities from the MIDAS dataset [Blewitt et al. 2016], we apply this method to the western US and estimate the probabilistic strain rates along the main fault systems, including the San Andreas system.