Abstract
Seismic yield estimation at regional distances will become increasingly important in monitoring a Low-Yield Threshold Test Ban (LYTTB). In order to investigate seismic yield estimation using regional data, we have examined regional mb(Pn) values for a large population of Nevada Test Site (NTS) underground nuclear explosions that occurred in widely varying geologic media with a range of yields from less than 1 kt to 300 kt. Magnitude-yield relationships were developed using data from a set of pre-1985 events (calibration set). To obtain an independent estimate of the uncertainty in estimating yields of events at NTS using mb(Pn) alone, the calibrated relationships were applied to a set of post-1984 events (prediction set). We found that the average uncertainty in the yield estimate, based on a weighted combination of the yield estimates from the four Livermore NTS Network (LNN) seismic stations, is about 1.8 (at the 2σ level). If data are available from all four LNN stations, the uncertainty is about 1.7 (at the 2σ level).
By including a term for the gas-filled porosity of the rock near the working point in the magnitude-yield relationship, we were able to develop a single relationship valid for events both above and below the water table, as well as those in alluvium and tuff. We have found, however, systematic differences in the relationships for events in Yucca Flat, Pahute Mesa, and Rainier Mesa.