Adaptive Techniques for Estimating and Locating Small Events Using Regional Waveform Data
Abstract
Attempts at locating and identifying small events could benefit greatly from a better appreciation of regional seismograms. Essentially, we need more information from the few regional seismograms available than just arrival times and amplitude ratios. Here we discuss a method of estimating source parameters using as few as one broadband station in conjunction with travel-time and polarities from at least one more station. The method employs an adaptive grid-search of matching three-component waveform records against synthetics to establish source location and depth. The basic matching procedure contains a trade-off of source mechanism with location. The better the constraint on mechanism and depth, the better the location. It appears that refined depths and origin times can be obtained for GT5 events if the first-motion polarities are available to constrain the focal plane. The locations of GT10 and GT25 and EHB (Engdahl, Kennett, and Buland) events can also be improved if CMT's and other independent depth estimates are available. Preliminary results applied to the PASSCAl data recorded in Pakistan and Tibet (China) and TriNet (CAL) proved highly effective when locating events located approximately between two stations. If both stations have usable three-component data, we can determine depth and source mechanism and refine the location. If we assume the mechanism and depth are known (Master Reference Event), we can relocate the event using just one broadband station and the P-wave arrival time pick from the other, or we can refine the crustal model. We find good agreement when testing these results against those from the entire array.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2000
- Accession Number
- ADA530005
Entities
People
- Brian Savage
- Chen Ji
- Don V. Helmberger
Organizations
- California Institute of Technology