Modeling Distributed Acoustic Sensing Signals from an Explosion

Session: Explosion Seismology Advances
Type: Oral
Date: 4/29/2020
Time: 03:00 PM
Room: 240
Description: 

Photonic sensors, such as distributed acoustic sensors (DAS), are providing new data streams for analysis. These sensors use pulses of light travelling along an optical fiber to record transient strain produced by incident seismic waves. Using data from the Source Physics Experiment that includes both borehole and surface fiber observations of a buried explosion, we explore various ways of modeling the fiber observations and understanding the response. A straight fiber is primarily sensitive to longitudinal strain parallel to the fiber and the response decreases as the incidence angle increases by a factor of cos². However, some sensitivity may occur even with waves arriving perpendicular to the fiber. Amplitudes depends in part on the coupling and cable construction. A helically wound fiber should be sensitive to all components of strain. Frequency response is dependent in part on the gauge length. We use two algorithms to model the signal: a 2D wavenumber algorithm and a 3D finite difference code. Both codes model only linear elastic wave propagation. The wavenumber algorithm is fast and can handle high frequencies, but our current implementation only calculates one strain component. The 3D finite difference code generates the strain components directly but requires more computational time. We also generate velocity synthetics at co-located surface geophones. Forward models using the known explosion and receiver locations indicate that a purely isotropic explosion source cannot match the amplitude and timing of the observed waveforms and that an anisotropic component is required to create the observed S waveforms. However, differences are apparent between the synthetics and observations and we are working to understand the causes of the misfit. Some of the differences may be due to non-linear effects, such as spall, which are not included in the current model. LLNL-ABS-800728. Prepared by LLNL under Contract DE-AC52-07NA27344. SNL is managed and operated by NTESS under DOE NNSA contract DE-NA000352.

Presenting Author: Robert J. Mellors

Authors