Seismo-Acoustic Studies in the European Arctic S. J. Gibbons, Frode Ringdal and Tormod Kværna NORSAR P.O. Box 53 N-2027 Kjeller, Norway (tormod@norsar.no) Infrasound Technology Workshop, Tunisia, 18 22 October 2010
Infrasound Signals at Seismic Sensors We first became aware of infrasound signals on the seismic instruments of the ARCES array in northern Norway when colleagues at the Kola Regional Seismological Center in Russia installed a microbarograph subarray at Apatity. Acoustic signals at Apatity were associated with seismic signals at ARCES from a repeating source. On closer examination, the seismic signals at ARCES were followed after ~9-11 minutes by acoustic signals. These had always been missed often detected, they were just classified as noise since they did not propagate with seismic velocities.
Array Analysis We can perform f-k analysis on all parts of the wavetrain. The acoustic signal indicates a similar direction to the seismic signals but travelling with the speed of sound in air. The resolution for azimuth and velocity estimates is much higher for this slowly propagating and highly coherent acoustic signal than for the more scattered seismic phases.
Variability of the Seismic and Acoustic Signals Essentially identical seismic signals from event to event. Implies same source region and similar source functions. The large variation in the acoustic signals can then only be the result of atmospheric properties. The events can be detected readily using a correlation detector on the seismic signals.
Variability of the Seismic and Acoustic Signals Display the coherence as a function of time (calculated using the appropriate time-shifts between the array sensors). Distance 175 km Observations at ARCES seismic sensors 2001 2006 (Aug Sep) Some arrivals between 500 and 570 seconds presumed tropospheric. Arrivals between 600 and 700 seconds presumed stratospheric. Shadow-zone arrivals for basically all events.
Experimental Microbarograph Array at ARCES Explosions in the fall of 2008 were observed by an experimental infrasound array at three of the sites at ARCES. In addition to the phases observed usually, for 11 out of 36 events, small amplitude, relatively impulsive signals were observed after 14-15 minutes. The times are consistent with thermospheric phases although such phases have very rarely been observed at these distances.
Array Analysis Performing array analysis on the late arrivals indicates a far higher apparent velocity or a steeper angle of incidence. This is also consistent with the hypothesis that these phases are thermospheric. It is unclear as to why these phases are observed some times and not at others. This may be a signal-tonoise issue the signals always have smaller amplitudes than the tropospheric/stratospheric signals. 625 seconds 346 m/s 928 seconds 566 m/s
Infrasound Observations at Several Stations in the Region Clear infrasonic signals detected across the Fennoscandia/NW Russia infrasound network. The seismic signals detected out to a distance of ~300km. The infrasound signals are detected out to distances greater than 1000 km a good reason for combining seismic and infrasonic monitoring! Using the ARCES seismic array we can examine direction estimates using many different subarray configurations.
Find Event Clusters by Waveform Correlation (seismic) Large numbers of seismoacoustic events from a given source region can be collected using a waveform correlation detector. This allows us to use sites even when no Ground Truth information is provided by the operators. Even when there is significant dissimilarity between signals from subsequent events (e.g. due to different source-time functions) correlation detectors on arrays can usually identify almost all events by examining the alignment of single-channel cross-correlation traces.
Infrasound Observations during 2002-2008
A new source of well-constrained seismo-acoustic events. The Suurikuusikko gold mine, operated by Agnico Eagles limited, is located approximately 10 km west of ammunition demolition site of the Finnish military. Operations began in 2006. The photo shows the volume excavated so far and the final extent of the open pit: Many years of Ground Truth seismo-acoustic events to come!!!
The Suurikuusikko mine is situated with an almost optimal coverage of infrasound stations at regional distances A common problem in studies of infrasound is the sparsity of stations. The six stations in the Fennoscandia/NW Russia network are distributed to give an almost optimal coverage. ARCES and Kiruna are on the edge of the shadow zone to the north and west. Apatity is almost due east at over 300 km. Three stations are to the south at different distances.
Current Status of Known Seismo-Acoustic Sources and Stations Several new sources of seismo-acoustic events have been identified only in the last few months. The seismic signals are often weak and below the threshold that warrants their inclusion in the reviewed event bulletin. Weak seismic signals can still pinpoint a source accurately in time and space using correlation and matched-field methods.
Current Status of Known Seismo-Acoustic Sources and Stations The source-to-receiver distances in the region range from local (~10 km) up to several hundred kilometers. Many sources occur allyear-round and, with stations to the east and west, will allow us to assess seasonal variability of observations. An interesting profile of sources is found at 175 km, 200 km, 250 km, and 270 km to the west of ARCES: just inside and outside of the shadow zone.
Summary A wealth of seismo-acoustic sources is found in northern Fennoscandia and north-western Russia. Several new sources have been identified recently. ARCES seismic array data goes back to 1987 over 20 years of observations! Correlation detectors can scan the data archives for repeating sources. The network of stations provides a broad coverage of distances and directions at which to observe the regional infrasound signals. Many are at distances at which the propagation of infrasound is poorly understood (e.g. within the so-called zone of silence) Large numbers of well-constrained events allow us to assess the variability of the seismo-acoustic wavefield: * the range of times and direction estimates that can be anticipated * the circumstances under which acoustic signals are observed at all
9 December 2009, Northern Norway, 07:50 local time
Infrasound signals from White Sea Missile Launch
ARCI observation 9 December 2009
15 July 2009 vs 9 December 2009 Δ_ARCES ~ 730 km T_ARCES ~ 42 min Δ_Apatity ~ 360 km T_Apatity ~ 20 min
Optical Observatory, Svalbard Fish-eye lense, Nikon Camera East West 06:51:41 UTC 06:52:31 UTC 06:53:20 UTC
100 km above Novaya Zemlya / Kara Sea