Small-event Yield and Source Characterization Using Local P and S-wave Coda Source Spectra

The mission of the Air Force Technical Applications Center (AFTAC) requires accurate yield estimation for nuclear explosions. Historically, the focus has been on larger yield events (mb > ~4.5) using teleseismic body wave magnitudes and applying test-site-specific corrections for yield estimates. The regional coda methodology provides unprecedented stability and avoids test site bias because it is based upon absolute source spectra. Increasingly, however, there is interest in monitoring smaller events both for yield and source characterization. Unfortunately, these events may only be recorded with adequate signal-to-noise ratio at local distances from one station. The project goals were to extend the well-established regional coda methodology to local distances using S and P-wave codas in regions of little-to-no calibration data and/or regions of high attenuation and lateral complexity. Previous studies show that local coda has a unique property of homogenizing its energy over a volume of the Earth’s crust such that path corrections for distances less than ~200 km are not necessary, or minimal at worst. Our plan this year was to use existing datasets from a variety of active tectonic settings and source types with the aim of assessing performance under the assumption of little to no calibration data. We have compared S-wave coda path attenuation curves from a variety of regions to look for similarities and differences that could be correlated to the degree of tectonic activity. In general, we can make some preliminary statements. First, we find that central Italy exhibits the strongest attenuation (0.5 to 2.0 Hz), followed by Taiwan and western Iran for distances ranging between ~20 and 300 km. On the other hand, upstate New York, South Africa, and the Korean Peninsula appear to have the lowest attenuation, which is reassuring since both regions are the most tectonically stable. The San Francisco Bay Area is somewhere between the two. This special feature may make it easier to define an a priori set of coda-calibration parameters that can be transported to new but geophysically similar regions. For example, it appears that tectonically similar regions have similar a coda path, envelope shape, and peak envelope velocity, which will allow us to derive average “local background” models then apply them to other regions for testing and evaluation, including cases to mimic an uncalibrated, single-station deployment. 2008 Monitoring Research Review: Ground-Based Nuclear Explosion Monitoring Technologies