Spatiotemporal Resolution and Applications of Back-Projection in Laboratory Fore-Shock Fault Mechanics

The large gap in current understanding around earthquake source physics and the kinematics of rupture across a single asperity or contact patch limits our ability to compare seismic events across field and laboratory scales. Seismic imaging methods including back-projection and kinematic source inversion are ever more widely applied in field studies to develop this understanding, but are relatively rarely used in laboratory environments. Compared to the field, laboratory earthquake experiments generally have fewer sensors available but the acoustic emissions (AE) data recorded is often higher fidelity, higher frequency, more broadband, less attenuated, and more repeatable. By improving full-waveform analysis of laboratory AE events we seek to reveal the complexities of co- and aseismic slip down to single asperities and help bridge the gap to field-scale observations. Back-projection is typically applied to imaging the source kinematics of megathrust events using teleseismic data from large, regular seismometer arrays and limited assumptions about wave propagation. We adapt the method to a laboratory array of sixteen AE sensors with wide local coverage and irregular spacing. The controlled laboratory environment provides additional advantages such as a homogeneous velocity model, analytical Green's functions and absolutely calibrated displacement sensors. We conduct a preliminary method assessment using synthetic data and station locations to determine the theoretical resolution of different array configurations and frequency bands. These results are verified against experimental data from calibration force sources with varied frequency content. Finally, the method is applied to fore-shock AE events, occurring prior to full fault rupture in a direct shear experiment. Local slip along the fault boundaries is recorded by non-contact displacement sensors, providing a fault-level aseismic slip background to contextualize the back-projected asperity-level slip results.