Practical Issues in the Location of Small Events Under a CTBT: Poor Station Coverage and Poorly Known Velocity Structure,
Abstract
In regions that do not contain good station coverage, a reasonably well known velocity structure and moderate or large magnitude events (the latter needed for calibration using teleseismic constraints on the location of a master event), it is very difficult to obtain accurate regional event locations. In this situation standard error ellipses generated during event locations may give the impression that an event location is better deterinined than it really is. False interpretations of the data may be due to a three dimensional velocity structure bias unaccounted for by the method, or simply to phase misidentification. Forward waveform modeling does help constrain the structure, however this approach contains a number of pitfalls. Modeling works best if ground truth is available at several epicentral distances and azimuths. Otherwise an inappropriate model might be constructed by matching synthetic waveforms and arrival times to observed data assumed to be at the wrong location. That model might thereafter be used to locate additional events. Regional arrays, or arrays originally constructed for teleseismic monitoring, can sometimes be used to identify regional phases using frequency-wavenumber (F-K) measurements of phase velocity in short (3-5 second) windows on the seismograms, however this only gives a solution if the signal to noise ratio is high. The F-K method is sensitive to the chosen center frequency and bandwidth, and to thee dimensional heterogeneities surrounding the array. Some of these practical issues are discussed in scenarios involving events located within 500 km of a recording station or array. We model seismograms using reflectivity and linear finite difference techniques, assuming flat isotropic layers for simplicity. Phases that can be identified and modeled at these distances include P and S wave reflections and refractions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 14, 1995
- Accession Number
- ADP204508
Entities
People
- Anton Dainty
- John Cipar
- Katharine Kadinsky-cade
- Rong-song Jih
Organizations
- Phillips Laboratory