### Regional Seismic Event Location in Asia

## Preliminary P_{n} Back-Projection Results

To improve our ability to predict PPCs in regions where data is sparse or non-existent, we have begun to look at using tomography to develop a regional velocity model. We use 6033Pn residuals from 55 stations for earthquakes satisfying the 50/90 rule (hopefully, less than about 20km error). Teleseismic station corrections are calculated from data in the USGS EDRs and applied prior to backprojection.
For each pixel,, rays were binned into eight back-azimuth groups. Pixel weights (*W*_{p}) were determined based on the number of rays in each bin (n_{1...8}) and the total number of rays crossing the pixel (n_{T}). If all n_{i}, i=1,8>=8, then *W*_{p} = 1. If not then *W*_{p} = (n_{T}/8)^{.5} * [1/(8-n_{1}) +1/(8-n_{2})...1/(8-n_{8})].

The result shown below is for one iteration of backprojection. These results are consistent with results from an earlier study by McNamara et. al. (1997) whose data we have obtained by not yet incorporated into our study.

Introduction | Seismicity
of Asia | Station Variances | Example
PPC Surfaces

Regionalized Lop Nor Relocations |
Grid Search Locations |
Preliminary P_{n} Back-Projection Results |
Conclusions