Characterization of an Explosion Source in a Complex Medium by Modeling and Wavelet Domain Inversion
Abstract
Explosions are often conducted in complexes with chambers, tunnels, and shafts used for access and instrumentation. These structures and sharp topographic features can act as strong scatterers of seismic waves and complicate the radiation patterns from explosions. In extreme cases they could affect discrimination between earthquakes and explosions. The objectives of this research are (1) to study the effects of these near-source scatterers on seismic waves radiated from explosions, and (2) to use a wavelet domain based moment-tensor inversion scheme to determine "explosive" and "multi-couple" components of the source. To calculate seismograms from an explosion near strong scatterers we use a new finite difference algorithm that accomplishes variable griding by coordinate transformation or "stretching." This method provides excellent numerical stability while increasing computational efficiency. It is capable of 3-D calculations for sources near strong scatterers in heterogeneous media. Seismograms are calculated to determine effects of various scatterers on seismic radiation patterns. The code is developed for realistic earth models including (1) free surface, (2) layered structure, (3) surface topography, and (4) seismic attenuation. In addition, a perfectly matched layer (PML) was incorporated into the finite-difference code to improve the absorption at the boundaries and for saving memory. Forward modelings using the 3-13 finite-difference code are conducted with an explosive source and tunnel in a full space and a layered half-space. Calculations are carried out for each case: (1) a reference model without a tunnel, and (2) a finite length horizontal tunnel included. We show the capacity of our code in modeling wave scattering due to topographical features. The method can be used to analyze data from earthquakes and explosions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2006
- Accession Number
- ADA455323
Entities
People
- M. Nafi Toksoez
Organizations
- Massachusetts Institute of Technology