CHARACTERIZATION OF MINING EXPLOSIONS AT REGIONAL DISTANCES: IMPLICATIONS WITH THE INTERNATIONAL MONITORING SYSTEM

Abstract

The International Monitoring System (IMS) being constructed in support of the Comprehensive Nuclear‐Test‐Ban Treaty introduces new opportunities to nuclear test monitoring by providing open access to global data from seismic, hydroacoustic, infrasonic, and radionuclide sensors. These sensors will detect myriad natural and man‐made events and can be used to identify those that have explosive characteristics and therefore might be clandestine nuclear tests. Detection and identification of seismic events must be conducted at a lower magnitude threshold (mb = 3.5 and lower) than has been previously considered. Concomitant with the lower monitoring threshold will be an increased number of events that must be scrutinized. This collection will be largely composed of regional observations in which the seismic waves have traversed complex geological structures. High‐fidelity regional, geophysical models will be needed to support accurate location and source identification. Source identification will not be limited to the separation of single‐fired nuclear explosions from earthquakes as in previous testing treaties. The lower‐magnitude threshold and increased reliance on regional observations assures that mining explosions will be detected by the monitoring system. It is important that the signals from mining explosions are properly identified to avoid false alarms of the monitoring system. Cooperation with the mining industry, including deployment of close‐in instrumentation and extensive documentation of the explosions, provides critical information for interpreting the performance of regional discriminants. Linkage of these observations to appropriate physical models of the blasting process is also enhanced through this cooperative research effort. A number of discriminants for characterizing mining explosions have been identified including P to Lg ratios at high and low frequencies, surface wave to high‐frequency body wave amplitudes, Rg at short distances, high‐ and low‐frequency time‐independent spectral modulations, signal correlation, and temporal clustering. Variable performance of individual discriminants results from such factors as mine‐specific blasting practices and the complexity of the regional wave propagation. This variability attests to the need for a suite of region‐specific discriminants. Regional calibration with modest‐size, single‐fired contained chemical explosions can provide a basis for developing region‐specific procedures. Broadband data provide the basis for the most robust set of discriminants. The inclusion of infrasonic data as part of the IMS introduces a potential for the combined use of seismic and infrasonic data for the identification of near‐surface explosions. The generation and, to a greater extent, the propagation of mining explosion infrasonic signals is not well understood, but empirical data attest to its future utility. Evidence for the accidental, near‐simultaneous detonation of a large amount of explosives during standard delay‐fired explosions is presented. Events such as these have single‐fired characteristics and may prove to be problematic in discrimination analysis.