| Summary: | This project is providing ground-truth information on explosions conducted at the principal mines within 500 kilometers of the ARCES station, and is assembling a seismic waveform database for these events from local and regional stations. The principal mines of interest are in northwest Russia (Khibiny Massif, Olenogorsk, Zapolyamy, and Kovdor groups) and Sweden (Malmberget, Kiruna). These mines form a natural laboratory for examining the variation of mining explosion observations with source type, since they include colocated surface and underground mines and mines conducting a variety of different shot types. In September 2002 we deployed two lines of temporary stations from the Khibiny Massif through and to the north of the ARCES station. This deployment is producing data that will allow researchers to examine the variation of discriminants caused by varying source-receiver distance and the diversity of explosion types. To date, we have collected ground-truth information on 1,118 explosions in the Kola Peninsula, and have assembled waveform data for approximately 700 of these. The database includes waveforms from instruments temporarily deployed in the Khibiny Massif mines, from the Apatity network just outside of the Massif, from LVZ, KEV and ARCES, and from the stations deployed along the two lines into northern Norway. In this paper we present representative waveforms for several types of shots recorded at various regional distances. We have conducted a preliminary study of the variation of phase ratios as a function of source type. This study shows significant differences in Pd/Sn and Pd/Lg ratios for two types of mining explosions: surface ripple-fired explosions and compact underground explosions. Compact explosions are, typically, underground explosions of a few tons with only one or two short delays, and are the closest approximation to single, well-tamped explosions available in the Khibiny mines. The surface shots typically are much larger (ranging up to hundreds of tons), with many delays. The surface mine that we present results for typically also conducts several distinct shots across the mine nearly simultaneously (with a few seconds or tens of seconds). Measured phase ratios are more consistent for compact underground explosions. This consistency is an expected result given the smaller scope for shot variation in these smaller events. In addition, Pd/Lg ratios appear more stable than Pd/Sn ratios for both types of events. The most interesting result is that the compact underground explosions are richer in shear energy (i.e. having smaller P/S ratios) than their surface ripple-fired counterparts. We continue to work on an approach for identifying the principal mines to be targeted for screening at a particular station. Often, routine industrial blasts constitute a large proportion of events detected by monitoring stations close to major mining districts. Many mines may be present, and it may be a problem to determine which subset of mines is responsible for the majority of the events, and should be prime candidates for the deployment of ground-truth collection resources. Our solution to this problem entails several steps. The first is to find geographic clusters of events that may correspond to major groups of mines. For this step, we use event density maps generated from existing network catalogs. This year we examined some of the tradeoffs in generating event density maps: use of automated bulletins to produce maps vs. analyst-reviewed bulletins, and the amount of time required to produce stables maps which can be used to identify significant mines.
|
|---|