International Monitoring System Arrays Research Articles

Identifying suitable time periods for infrasound measurement system response estimation using across-array coherence

SUMMARY Microbarometers deployed to measure atmospheric infrasound are often connected to, or housed within, a wind noise reduction system (WNRS). At infrasound arrays of the International Monitoring System (IMS), being deployed as part of Comprehensive Nuclear-Test-Ban Treaty verification measures, the WNRS typically comprises an 18 m diameter pipe array. Over the past decade an in situ method has been developed to estimate the measurement system (sensor + WNRS) response characteristics, by comparing its recordings with those made on a colocated reference sensor with known response and no WNRS. The method relies upon the identification of time periods for which the reference sensor and measurement system are subject to the same input pressure field. It has proven difficult to reliably identify such time periods at frequencies $\lt 0.1\,$ Hz using recordings at a single location, resulting in a negative bias in estimated measurement system gain values (the ‘dip artefact’) in the 0.02–0.1 Hz passband. The IMS is deploying arrays of microbarometers, and we show that a measure of across-array coherence can be used to identify time periods associated with acoustic signal propagation. Amplitude response estimates, using 1 yr of data from four IMS arrays, indicate that the dip artefact can be removed by retaining for analysis only those time periods that exhibit high across-array coherence. Moreover, our analysis confirms the hypothesis that the dip artefact is associated with time periods during which wind-generated pressure fluctuations dominate, leading to partial suppression of noise with length scales less than the extent of the WNRS. At two arrays within continental forests accurate amplitude responses are estimated across the 0.02–4 Hz passband, as acoustic signals at all frequencies can be identified. At two oceanic island arrays, the low numbers of time windows with above-noise acoustic signal in the 0.02–0.1 Hz passband make reliable response estimation at these frequencies difficult or impossible. It is recommended that the methodology for estimating the response of an infrasound measurement system at an array should incorporate a multichannel coherence measure; data centres may already routinely compute such measures in their signal detection algorithms.

Threshold Magnitudes for a Multichannel Correlation Detector in Background Seismicity

Colocated explosive sources often produce correlated seismic waveforms. Multichannel correlation detectors identify these signals by scanning template waveforms recorded from known reference events against data to find similar waveforms. This screening problem is challenged at thresholds required to monitor smaller explosions, often because non-target signals falsely trigger such detectors. Therefore, it is generally unclear what thresholds will reliably identify a target explosion while screening non-target background seismicity. Here, we estimate threshold magnitudes for hypothetical explosions located at the North Korean nuclear test site over six months of 2010, by processing International Monitoring System (IMS) array data with a multichannel waveform correlation detector. Our method (1) accounts for low amplitude background seismicity that falsely triggers correlation detectors but is unidentifiable with conventional power beams, (2) adapts to diurnally variable noise levels and (3) uses source-receiver reciprocity concepts to estimate thresholds for explosions spatially separated from the template source. Furthermore, we find that underground explosions with body wave magnitudes mb = 1.66 are detectable at the IMS array USRK with probability 0.99, when using template waveforms consisting only of P -waves, without false alarms. We conservatively find that these thresholds also increase by up to a magnitude unit for sources locatedmore » 4 km or more from the Feb.12, 2013 announced nuclear test.« less

International Monitoring System Correlation Detection at the North Korean Nuclear Test Site at Punggye‐ri with Insights from the Source Physics Experiment

Online Material: Figures of array correlation detection with an alternative embedded event, correlation score as a function of time‐bandwidth product, and correlation score using alternative time periods. Seismic waveform correlation offers the prospect of greatly reducing event detection thresholds when compared with more conventional processing methods. Correlation is applicable for seismic events that in some sense repeat, that is they have very similar waveforms (e.g., Gibbons and Ringdal, 2006). A number of recent studies have shown that correlated seismic signals may form a significant fraction of seismicity at regional distances (e.g., Schaff and Richards, 2011; Dodge and Walter, 2015; Slinkard et al. , 2015). For the particular case of multiple nuclear explosions at the same test site, regional distance correlation also allows very precise relative location measurements (e.g., Waldhauser et al. , 2004; Wen and Long, 2010) and could offer the potential to lower thresholds when multiple events exist (e.g., National Research Council, 2012). Using the Comprehensive Nuclear‐Test‐Ban Treaty (CTBT) International Monitoring System (IMS) seismic array at Matsushiro, Japan (MJAR), Gibbons and Ringdal (2012) were able to create a multichannel correlation detector with a very low false alarm rate and a threshold below magnitude 3.0. They did this using the 2006 or 2009 Democratic People’s Republic of Korea (DPRK) nuclear explosion as a template to search through a data stream from the same station to find a match via waveform correlation. In this article, we extend the work of Gibbons and Ringdal (2012) and measure the correlation detection threshold at several other IMS arrays. We use this to address three main points. First, we show the IMS array station at Mina, Nevada (NVAR), which is closest to the Nevada National Security Site (NNSS), is able to detect a chemical explosion that is well under 1 ton with the right template. …

Location of high‐frequency P wave microseismic noise in the Pacific Ocean using multiple small aperture arrays

AbstractWe investigate source locations of P wave microseisms within a narrow frequency band (0.67–1.33 Hz) that is significantly higher than the classic microseism band (~0.05–0.3 Hz). Employing a backprojection method, we analyze data recorded during January 2010 from five International Monitoring System arrays that border the Pacific Ocean. We develop a ranking scheme that allows us to combine beam power from multiple arrays to obtain robust locations of the microseisms. Some individual arrays exhibit a strong regional component, but results from the combination of all arrays show high‐frequency P wave energy emanating from the North Pacific basin, in general agreement with previous observations in the double‐frequency (DF) microseism band (~0.1–0.3 Hz). This suggests that the North Pacific source of ambient P noise covers a broad range of frequencies and that the wave‐wave interaction model is likely valid at shorter periods.

Probing the edge of the West African Craton: A first seismic glimpse from Niger

AbstractConstraints on crustal and mantle structure of the Eastern part of the West African Craton have to date been scarce. Here we present results of P receiver function and SK(K)S wave splitting analyses of data recorded at International Monitoring System array TORD in SW Niger. Despite lacking in lateral coverage, our measurements sharply constrain crustal thickness (∼41 km), VP/VS ratio (1.69 ± 0.03), mantle transition zone (MTZ) thickness (∼247 km), and a midlithospheric discontinuity at ∼67 km depth. Splitting delay times are low with an average of 0.63 ± 0.01 s. Fast directions follow the regional surface geological trend with an average of 57 ± 1°. We suggest that splitting is due to fossil anisotropic fabrics in the crust and lithosphere, incurred during the Paleoproterozoic Eburnean Orogeny, with possible contributions from the later Pan‐African Orogeny and present‐day mantle flow. The MTZ appears to be unperturbed, despite the proximity of the sampled region to the deep cratonic root.

The spatial coherence structure of infrasonic waves: analysis of data from International Monitoring System arrays

The spatial coherence structure of infrasonic waves: analysis of data from International Monitoring System arrays

Improvements to Seismic Monitoring of the European Arctic Using Three-Component Array Processing at SPITS

Abstract The detectability of low magnitude seismic events in the European Arctic is determined primarily by the small-aperture International Monitoring System arrays ARCES and SPITS. In August 2004, the SPITS array was upgraded to a broadband array with an increase in the sampling rate from 40 to 80 Hz. Most important, however, for the detection and location of small-magnitude seismic events was the deployment of three-component instruments at six of the nine sites. Detection and correct classification of secondary phases are of paramount importance for events observed by only a small number of stations at regional distances; and, in the absence of the strong Lg phases typically observed for continental propagation paths, multiple three-component stations were deemed necessary to exploit the higher S -phase amplitudes anticipated on the horizontal sensors. We demonstrate improved signal-to-noise ratios (SNRs) for S phases on horizontal beams for several events close to Novaya Zemlya. Horizontal component f - k analysis improves direction estimates and phase classification for low-SNR signals. We demonstrate secondary phases that are misidentified by vertical-only f - k analysis but which are correctly classified by three-component array processing. A significant problem with array processing at SPITS is the overlap in slowness space of regional P and S phases. Phase identification is improved greatly by comparing the coherence between vertical traces with the coherence between horizontal traces. Considerations in the routine array processing of SPITS data are reviewed, including the need for elevation corrections in slowness estimation and the need to take into account azimuth-dependent variation of apparent velocity estimates for regional phases.

Southeast Indian Ocean-Ridge earthquake sequences from cross-correlation analysis of hydroacoustic data

SUMMARY Parameters of earthquake sequences, for instance location and timing of foreshocks and aftershocks, are critical for understanding dynamics of mid-ocean ridge and transform faults. Whole sequences including small earthquakes in the ocean cannot be well recorded by land-based seismometers due to large epicentral distances. Recent hydroacoustic studies have demonstrated that T waves are very effective in detecting small submarine earthquakes because of little energy loss during propagation in Sound Fixing and Ranging channel. For example, an MW 6.2 (2006 March 6, 40.11 ◦ S/78.49 ◦ E) transform-fault earthquake occurred at the Southeastern Indian Ocean Ridge, but National Earthquake Information Center only reported three aftershocks in the first following week. We applied cross-correlation method to hydroacoustic data from the International Monitoring System arrays in the Indian Ocean to examine the whole earthquake sequence. We detected 14 aftershocks and none foreshock for the earthquake and locations of these aftershocks show an irregular pattern. From the observation, we suggest that the feature could be caused by complicated transcurrent plate-boundary dynamics between two overlapped spreading ridges that is possibly explained by the bookshelf faulting model.

Calibrating the Chiang Mai Seismic Array (CMAR) for Improved Event Location

Reliable single-array location is important for nuclear monitoring and other Earth science applications ( e.g. , Lindquist et al. 2005; Guilbert et al. 2005; Rost and Garnero 2006). Effective use of small-to-medium-aperture (20–25 km or less) seismic arrays requires calibration to account for local and distant propagation effects. Array calibration reduces bias in bearing estimation and event location at regional and teleseismic distances. There is a need for calibration at the International Monitoring System array CMAR (Chiang Mai, Thailand). First mentioned by Koch and Kraundofer (1997) and Bondar et al. (1999), then by Guilbert et al. (2005) and Koper and Fatehi (2008), significant effort was devoted to study-specific, empirical corrections of bias in back-azimuth observed at CMAR. However, even though dipping Mohorovicic discontinuity beneath CMAR was mentioned by Bondar et al. , 1999 as a possible explanation of the observed parameter residuals, no further research was devoted to solving this hypothesis. We observe cosine-shape back-azimuth and horizontal-velocity residuals at CMAR and propose a fast and effective formula for first-order corrections for a dipping Mohorovicic (Moho) discontinuity beneath the array, valid for near-regional to teleseismic distances. Empirical slowness (as defined by Aki and Richards 1980) and back-azimuth corrections are currently used for event location using International Monitoring System arrays and single stations (Bondar et al. 1999). The corrections by Bondar et al. (1999) are estimated on automatically analyzed datasets using frequency-wavenumber analysis (Kvaerna and Dornbos 1986) for stations at teleseismic (>20 ° ) epicentral distance. These corrections are also smoothed over a polar grid. A different approach is to tie the observed systematic residuals to a crustal structure beneath the array ( e.g. , dipping Moho discontinuity beneath the array, differences in elevation of the instruments, or complicated geology at the site) and to find a physical model that explains these residuals. …

Global Detection of Infrasonic Signals from Three Large Bolides

We present the infrasonic observations of three large bolides that were observed at numerous International Monitoring System (IMS) infrasound arrays on a global scale. First, a simple procedure for the global association of infrasound detections from large infrasound events is outlined. Infrasound signals are associated with large events based on arrival time, backazimuth and uniqueness at a given IMS array. Next, we apply the algorithm to three bolides and investigate some of the factors affecting the detectability of infrasound from large events. Our findings suggest that site-noise effects significantly degrade the capability of the IMS infrasound network, suggesting that more effort is required to reduce ambient site noise. These results have implications for the use of infrasound measurements (in particular those from IMS stations) as a tool for evaluating the global flux of near-Earth objects.

The detectability of infrasound in The Netherlands from the Italian volcano Mt. Etna

Infrasound from the eruptions of Mt. Etna was detected in The Netherlands at the Deelen Infrasound Array (DIA) during the summer of 2001. Zonal cross winds lead to a deviation in the observed back azimuth. An average azimuthal deviation of 2 . 6 ∘ is explained by westward stratospheric winds along the infrasonic trajectories through the atmosphere. The signals have signal-to-noise power ratios lower than 0.6. The detectability of Mt. Etna's infrasound is tested on sub-arrays of the 13 elements of DIA. Doing so, sparse arrays used in the International Monitoring System (IMS) are mimicked. Both small aperture sub-arrays of 800 m and large aperture sub-arrays of 1400 m are applied. The results are evaluated on the basis of the number of detections made. The number of detections strongly increases with the number of elements. Small aperture sub-arrays perform significantly better than large aperture sub-arrays. With a factor of two reduction in the aperture, two instruments less can be used to obtain the same number of detections. Larger aperture arrays perform less well because of the loss of spatial coherence of the infrasound. The average accuracy of the detections slightly increased, when the number of elements is increased. The result of the use of more elements is that lower signal-to-noise power ratio events could be detected. The trade-off between more detections and the larger average deviation in back azimuth and apparent sound speed causes the slight increase accuracy. The results were comparable to the detections obtained at IMS array IS26 in Germany. It is concluded that successfully applying infrasound as monitoring technique strongly depends on the array configuration. The frequency versus spatial coherence of signals will play a decisive role in detecting low signal-to-noise ratio events.

A Scheme for Initial Beam Deployment for the International Monitoring System Arrays

A Scheme for Initial Beam Deployment for the International Monitoring System Arrays