Continuous Seismic Threshold Monitoring of the Northern Novaya Zemlya Test Site: Long-Term Operational Characteristics

Abstract : In this work we demonstrate the practical capability of the Continuous Seismic Threshold Monitoring method to monitor the northern Novaya Zemlya test site at a very low threshold over an extended time period, using data from the Fennoscandian array network (NORESS, ARCESS and FINESA). We show that during February 1992 the network based magnitude threshold, at the 90% confidence level, stays below msubb = 2.50,99.72% of the total time. We further explain all peaks in the network magnitude thresholds exceeding msubb = 2.6 as resulting from interfering signals from an identified seismic event (teleseismic or regional), or a short outage of the most important array (ARCESS). We also argue that this implies that at the given confidence level, there has been no seismic event of msubb > or = 2.6 at the northern Novaya Zemlya test site during February 1992. During normal conditions, i.e., when the network threshold is low, ARCESS is clearly the most important array, followed by NORESS and FINESA. But during time periods when the ARCESS noise level is high, or when there are interfering events, the relative contribution of NORESS and FINESA increases significantly. The redundancy resulting from the use of several arrays is also essential during outages of one or more of the arrays. The threshold magnitudes for each array during background noise conditions are close to normally distributed, at least within shorter time intervals. Small deviations from the normal distribution occur because of long-term fluctuations in the background noise level. The average magnitude thresholds at FINESA exhibit strong weekly and diurnal variations. The latter are particularly significant on workdays.