Influence of Inhomogeneous Structure in Recovering the Source Parameters of the September 22, 1993 Non Proliferation Experiment
Abstract
In this study, we analyzed local waveforms from the September 22, 1993 Non-Proliferation Experiment (NPE) recorded by 30 stations within 3 km of the shot point to estimate the NPE source parameters with a variety of empirical and numerical models. The best source parameters were compared against the known values (chemical Y=1.1 Kt; h=393m). The derived results were thus used in establishing the level of accuracy in the estimated values. We considered the segment of waveforms free from the path effect and concentrated only on the spectral data of the initial P waves (0.6 to 2.0s time window starting at the P-wave onsets). As data were collected from different agencies, we normalized the spectra relative to 1Hz amplitude which eliminated any problem with the instrument's gain. Next, we predicted normalized spectral amplitude for a range of depth and yield of the event using the Mueller-Murphy (MM) source and tried to fit the normalized spectral amplitude of the individual stations, as well as the normalized spectra averaged over all stations. We found that we could recover the yield within 0.1-0.4 units, and the depth within a factor of 2 of the announced value. We also examined the influence of source region 3D and 1D geology, and of the topography, separately using the fall off of the spectra beyond the corner frequency. We found the spectral fall off in the recorded data is best represented by the 3D model. Our 3D model was developed based on the available geology and near source structure published in the DoE proceedings. To examine the sensitivity of the P wave spectra, several sets of finite-difference seismograms were simulated for our selected 30 stations for six depths ranging between 100 and 600 m, and convolved with a MM source of various yields. These waveforms were treated as observations and processed accordingly. For each depth simulation, average normalized spectra were computed and source parameters were determined. We found that both the depth and yield estimates from the simulated data exhibited the same level of uncertainty as determined by processing the recorded spectral spectra of the NPE.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.S43A2220S
- Keywords:
-
- 7200 SEISMOLOGY;
- 7219 SEISMOLOGY / Seismic monitoring and test-ban treaty verification