Progressive Multi-Channel Correlation Research Articles

A Single Array Approach for Infrasound Signal Discrimination from Quarry Blasts via Machine Learning

Since various phenomena produce infrasound, including both man-made and natural sources, the ever-growing dataflow demands automatic processes via machine learning for signal classification. In this study, we demonstrate a single array approach at the Piszkés-tető (PSZI) infrasound array. Our dataset contains nearly 14,000 manually categorized infrasound detections, processed with the progressive multi channel correlation (PMCC) algorithm from three different sources, such as quarry blasts, storms and signals from a power plant. The dataset was split into a training, a validation and a test subset. Time and frequency domain features as well as PMCC-related features were extracted. Three additional PMCC-related features were constructed in a way to measure the similarity between detections and to be used in single array monitoring. Two different classifiers, support vector machine and random forest, were used for training. Training was performed with three-fold cross validation with grid search. The classifiers were tuned on the training and validation set using the f1 metric (harmonic mean of positive predictive value and true positive rate). Training, validation and testing were performed with and without our three new features that measure similarity between the detections in order to assess their importance in single array monitoring. The selected classifiers reached f1 scores between 0.88 and 0.93. Our results show a promising step toward automatic infrasound event classification.

Megameter propagation and correlation of T-waves from Kermadec Trench and Islands

On 18 June 2020 and 4 March 2021, very energetic low-frequency underwater T-wave signals (2 to 25 Hz) were recorded at the Comprehensive Nuclear-Test-Ban Treaty (CTBT) International Monitoring System (IMS) hydrophone stations in the Pacific Ocean (Stations HA11 and HA03) and the South Atlantic Ocean (Station HA10). This work investigates the long-range (megameters) propagation of these T-waves. Their sources were three powerful submarine earthquakes in the Kermadec Trench and Islands, located at approximately 6000, 8800, and 15100 km from Stations HA11, HA03, and HA10, respectively. Arrival time and back azimuth of the recorded T-waves were estimated using the Progressive Multi-Channel Correlation algorithm installed on the CTBT Organization (CTBTO) virtual Data Exploitation Centre (vDEC). Different arrivals within the duration of the earthquake signals were identified, and their correlations were also analyzed. The data analysis at HA03 and HA10 revealed intriguing T-wave propagation paths reflecting, refracting, or even transmitting through continents, as well as T-wave excitation along a chain of seamounts. The analysis also showed much higher transmission loss (TL) in the propagation paths to HA11 than to HA03 and HA10. Moreover, strong discrepancies between expected and measured back azimuths were observed for HA11, and a three-dimensional (3D) parabolic equation model was utilized to identify the cause of these differences. Numerical results revealed the importance of 3D effects induced by the Kermadec Ridge, Fiji archipelago, and Marshall Islands on T-wave propagation to HA11. This analysis can guide future improvements in underwater event localization using the CTBT-IMS hydroacoustic sensor network.

The multichannel maximum-likelihood (MCML) method: a new approach for infrasound detection and wave parameter estimation

SUMMARY We are presenting a new and novel approach to the detection and parameter estimation of infrasonic signals. Our approach is based on the likelihood function derived from a multisensor stochastic model expressed in different frequency channels. Using the likelihood function, we determine, for the detection problem, the generalized likelihood ratio (GLR) and for the estimation of the slowness vector, the maximum likelihood estimation (MLE). We establish new asymptotic results (i) for the GLR under the null hypothesis leading to the computation of the corresponding p-value and (ii) for the MLE by focusing on the two wave parameters: backazimuth and horizontal trace velocity. The multichannel maximum-likelihood (MCML) detection and estimation method is implemented in the time–frequency domain in order to avoid the presence of interfering signals. Extensive simulations with synthetic signals show that MCML outperforms the state-of-the-art multichannel correlation detector algorithms like the progressive multichannel correlation in terms of detection probability and false alarm rate in poor signal-to-noise ratio scenarios. We also illustrate the use of the MCML on real data from the International Monitoring System and show how the improved performances of this new method lead to a refined analysis of events in accordance with expert knowledge.

International Monitoring System infrasound data products for atmospheric studies and civilian applications

Abstract. The International Monitoring System (IMS) was established in the late 1990s for verification of the Comprehensive Nuclear-Test-Ban Treaty (CTBT). Upon completion, 60 infrasound stations distributed over the globe will monitor the Earth's atmosphere for low-frequency pressure waves. In this study, we present advanced infrasound data products of the 53 currently certified IMS infrasound stations for atmospheric studies and civilian applications. For this purpose, 18 years of raw IMS infrasound waveform data (2003–2020) were reprocessed using the Progressive Multi-Channel Correlation (PMCC) method. A one-third octave frequency band configuration between 0.01 and 4 Hz was chosen to run this array-processing algorithm which detects coherent infrasound waves within the background noise. From the comprehensive detection lists, four products were derived for each of the certified 53 IMS infrasound stations. The four products cover different frequency ranges and are provided at the following different temporal resolutions: a very low-frequency set (0.02–0.07 Hz, 30 min; https://doi.org/10.25928/bgrseis_bblf-ifsd, Hupe et al., 2021a), two so-called microbarom frequency sets – covering both the lower (0.15–0.35 Hz, 15 min; https://doi.org/10.25928/bgrseis_mblf-ifsd, Hupe et al., 2021b) and a higher (0.45–0.65 Hz, 15 min; https://doi.org/10.25928/bgrseis_mbhf-ifsd, Hupe et al., 2021c) part – named after the dominant ambient noise of interacting ocean waves that are quasi-continuously detected at IMS stations, and observations with center frequencies of 1 to 3 Hz (5 min), called the high-frequency product (https://doi.org/10.25928/bgrseis_bbhf-ifsd, Hupe et al., 2021d). Within these frequency ranges and time windows, the dominant repetitive signal directions are summarized. Along with several detection parameters, calculated quantities for assessing the relative quality of the products are provided. The validity of the data products is demonstrated through example case studies of recent events that produced infrasound detected at IMS infrasound stations and through a global assessment and summary of the products. The four infrasound data products cover globally repeating infrasound sources such as ocean ambient noise or persistently active volcanoes, which have previously been suggested as sources for probing the winds in the middle atmosphere. Therefore, our infrasound data products open up the IMS observations also to user groups who do not have unconstrained access to IMS data or who are unfamiliar with infrasound data processing using the PMCC method. These types of data products could potentially serve as a basis for volcanic eruption early warning systems in the future.

Updated Global Reference Models of Broadband Coherent Infrasound Signals for Atmospheric Studies and Civilian Applications

AbstractThe International Monitoring System (IMS) infrasound network has been established to detect nuclear explosions and other signals of interest embedded in the station‐specific ambient noise. The ambient noise can be separated into coherent infrasound (e.g., real infrasonic signals) and incoherent noise (such as that caused by wind turbulence). Previous work statistically and systematically characterized coherent infrasound recorded by the IMS. This paper expands on this analysis of the coherent ambient infrasound by including updated IMS data sets with data up to the end of 2020 for all 53 of the currently certified IMS infrasound stations using an updated configuration of the Progressive Multi‐Channel Correlation (PMCC) method. This paper presents monthly station‐dependent reference curves for the back azimuth, trace velocity, and root mean squared amplitude, which provides a means to determine the deviation from the nominal monthly behavior. In addition, a daily Ambient Noise Stationarity (ANS) factor based on deviations from the reference curves is determined for a quick reference to the coherent signal quality compared to the nominal situations. Newly presented histograms provide a higher resolution spectrum, including the observations of the microbarom peak, as well as additional peaks reflecting station‐dependent environmental noise. The aim of these reference curves is to identify periods of suboptimal operation (e.g., nonoperational sensor) or instances of strong abnormal signals of interest.

The International Data Centre infrasound processing system, a 25 years travel

In 2001, when the first data from an International Monitoring System infrasound station started to arrive in near real-time at the International Data Centre (IDC), its infrasound processing system was in a premature state. The IDC embarked for a multi-year design and development of its dedicated processing system, which led to operational IDC automatic processing and interactive analysis systems in 2010. In the next ten years the IDC produced over 40,000 infrasound events reviewed by expert analysts. In an effort to continue advancing its methods, improving its automatic system and providing software packages to Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO) users, the IDC focused on several projects. First, the automatic system for the identification of valid signals was redesigned with the development of DTK-(G)PMCC (Progressive Multi-Channel Correlation), which is made available to CTBTO users within NDC-in-a-Box. And second, an infrasound model was developed for automatic waveform network processing software NET-VISA with an emphasis on the optimization of the network detection threshold by identifying ways to refine signal characterization methodology and association criteria. Ongoing and future improvements of the IDC processing system are planned to further reduce analyst workload that includes atmospheric propagation modeling and enhancements of the automatic pipeline components.

Infrasonic Earthquake Detectability Investigated in Southern Part of Japan, 2019.

The Kochi University of Technology (KUT) Infrasound Sensor Network contains 30 infrasound sensors which are distributed all over Japan especially in Shikoku Island. At all infrasound stations installed with three-axis accelerometers to measure the peak ground acceleration (PGA). Many earthquakes were detected by our system after establishing of the network since 2016. In this study we will focus on all the possibilities for infrasound detection generated from earthquakes using KUT sensor network and International Monitoring system (IMS) stations for the earthquakes which were detected in southern part of Japan during 2019. As for earthquakes with strike-slip mechanisms the P-waves could not be detected by our sensors. In addition, The conversion from seismic to acoustic waves can be happened through the generating of the T-phase from oceanic earthquakes. On 9 May 2019, progressive multi-channel cross correlation (PMCC) method applied infrasound and hydroacoustic waves from two earthquakes happened in west of Kyushu Island as the T-phase was well-recorded at H11N station near Wake Island. Moreover, infrasound propagation modeling is applied to the reconstructed atmosphere profile by Ground to Space Model (AVO-G2S) to confirm the infrasound arrivals, furthermore the 3D ray tracing process and the calculations by using the transmission loss equation with normal modes and parabolic equation methods are investigated. The study confirmed the infrasound generation scenario from the T-phase of oceanic propagation.

Analysis of Hydroacoustic Signals Associated to the Loss of the Argentinian ARA San Juan Submarine

On November 15, 2017, an event related to the disappearance of the Argentine military submarine ARA San Juan was detected by two hydrophone triplet stations of the IMS network, established to enforce the nuclear test ban treaty (CTBT). From the two direct hydroacoustic arrivals recorded at 6000 and 8000 km from the localized wreckage, calculated location based on hydroacoustic data only is poorly constrained, and the associated uncertainties are large. In an attempt to interpret the recorded signals, an air dropped calibration grenade was conducted by the Argentine Navy two weeks later, on December 1, 2017, near the last known position of the submarine. From the comparison of temporal and spectral features of both events, we confirm the impulsive nature of the San Juan event. Array processing was performed with a progressive multi-channel correlation method (PMCC). Fine propagation details of direct arrivals are very well resolved in time-frequency space and thirteen secondary arrivals are revealed for the San Juan event, within the fifteen minutes following direct arrivals. The detections presented in this paper were calculated with DTK-PMCC software embedded in the NDC-In-A-Box virtual machine, and can be reproduced by any CTBTO principal user (Member State user which can access raw waveform data and data bulletins). All the identified late arrivals are associated to reflections or refractions from seamounts, islands and the South American continental Slope. The accurate identification of all the reflectors allows to significantly improve the source location accuracy: 95% confidence ellipse area has been reduced by a factor of 100 compared to location obtained from direct arrivals only, and the estimated location is 3.5 km from the known location of the wreckage. The originality of the relocation method is that it is based on the joint inversion of both San Juan and calibration events unknown parameters, and from the selection of only a well-chosen subset of secondary arrivals. Its calculation did not require either the need of advanced oceanographic specifications, or sophisticated methods requiring heavy computational means. Finally, a detailed cepstral analysis of the direct and secondary arrivals has allowed to detect the existence of a second impulse (doublet) in the signals associated to both San Juan and calibration events. Unlike the calibration event, the anisotropic character of the delays measured from the San Juan cepstra suggests that the 15 November signal was generated by two impulsive acoustic sources closely separated in space and time over scales comparable to the size of the submarine. This study demonstrates the capability of the hydroacoustic component of the IMS network to accomplish its mission of Comprehensive Nuclear-Test-Ban Treaty monitoring.

Understanding the Split Characteristics of the Tropical Mesoscale Convective System (MCS) of April 9, 2018, in Northern Ghana Using Infrasound Data

The split characteristics of the tropical Mesoscale Convective System (MCS) of April 9, 2018, in northern Ghana were studied using infrasound data measured by the mobile array (I68CI) which was deployed by C?te d’Ivoire National Data Center (NDC) in collaboration with the Comprehensive Nuclear-Test-Ban Treaty (CTBT). These infrasound measurements were made during a measurement campaign from January 1st, 2018 to December 31, 2018, in northeast Cote d’Ivoire, precisely in Comoe National Park. Graphic Progressive Multi-Channel Correlation (GPMCC) method based on a progressive study of the correlation functions was used to analyze and visualize data. The infrasound detection from this MCS shows clearly a division of the MCS structure into 2 distinct subsystems under the effect of internal and external constraints not well known but related to convection; a smaller subsystem in the north, associated with an area of intense rainfall of about 30 mm/hour and located at 9.5°N - 2°E with an azimuth of 70° and, a large subsystem in the south, associated with a zone of high rainfall of about 96 mm/hour and located at 8.8°N - 1.4°E with an azimuth of 90°. These two subsystems were located 200 km and 260 km from the I68CI station with frequencies of 2.3 Hz and 1 Hz respectively. The mesoscale convective systems in this region are moving from East to West and including several storm cells.

The 12 December 2017 Baumgarten Gas Hub Explosion: A Case Study on Understanding the Occurrence of a Large Infrasound Azimuth Residual and a Lack of Seismic Observations

The Baumgarten explosion occurred on 12 December 2017 at a gas storage site about 30 km east of Vienna, Austria. Acoustic arrivals from this accidental surface explosion were detected at dozens of stations of the AlpArray seismic network to distances up to 150 km, mainly in easterly directions. Thus it was expected that the Hungarian infrasound array PSZI located about 230 km to the east-southeast of Baumgarten would detect this acoustic wave as well. Standard progressive multichannel correlation processing and frequency-wavenumber analysis identified a signal emerging at 7:57:55 UTC from an azimuth of 296°–300° and with trace-velocity > 400 m/s. The extraordinarily high trace-velocity and excessive backazimuth residual, relative to the explosion site direction of 282°, however cast strong doubts on the arrival’s connection to the Baumgarten event. Accounting for the effect of non-planar geometry of the infrasound array results in a reduction of the azimuth residual by half. Additionally, 2D and 3D raytracing methods are used including the European Centre for Medium-Range Weather Forecasts (ECMWF) atmospheric model to further explain the remaining azimuth residual as well as to elucidate the large trace velocity estimates. The prevailing stratospheric winds in excess of 150 m/s are identified as the underlying cause. Including both factors the initial azimuth residual of up to 18° decreases to ~ 4°, allowing to associate the infrasound signal at PSZI with the Baumgarten event. Finally, the data from a seismic station at 30 km range is re-investigated for magnitude estimation. The local magnitude of ML < 1.0 explains well the scarcity of seismic observations within 50 km range, where three or four stations show signals, mainly consisting of Rg-type surface waves, but no body waves.

Improved infrasound array processing with robust estimators

SUMMARY Infrasound array data are commonly used to detect and characterize infrasonic signals from a variety of natural and anthropogenic sources. Here we examine the effectiveness of robust regression estimators (L1-norm regression, M-estimators and least trimmed squares) for infrasound array processing, and compare them against standard array processing algorithms (least-squares estimation, frequency–wavenumber analysis and progressive multi-channel correlation) using a combination of real and synthetic data. Of particular interest is how each algorithm performs when one of the array elements produces data outliers. Synthetic tests on elements containing a clock error, constant values or only pink noise are performed, and we analyse the relative ability of the estimators to recover plane wave parameters. The L1-norm regression, M-estimate, frequency–wavenumber analysis and least trimmed squares estimates provided superior results than conventional least-squares estimation. Evaluation of least trimmed squares weights consistently identified the element with the simulated error, providing additional information on array performance. Least trimmed squares processing consistently identified an element with reversed polarity for Alaska Volcano Observatory array ADKI. International Monitoring System stations IS57 and IS55 were likewise processed. Data from an element of IS57, which had lower cross-correlation values than the remaining elements, were consistently identified as having outliers in array processing. An element with a timing error was identified in the analysis of IS55 data. These results suggest robust regression methods, in particular least trimmed squares, improve upon standard methods and should be used more widely, as they can provide robust array processing results and insight into array performance. Further, robust regression methods are not limited to infrasound array processing applications, and it is likely that they would also be effective for seismic array data.

Global Monitoring and Characterization of Infrasound Signatures by Large Fireballs

Large meteoroids can be registered in infrasound recordings during their entry into the Earth’s atmosphere. A comprehensive study of 10 large fireball events of the years 2018 and 2019 highlights their detection and characterization using global infrasound arrays of the International Monitoring System (IMS) of the Comprehensive Nuclear-Test-Ban Treaty (CTBT). The study focuses on the observation and event analysis of the fireballs to estimate their respective location, yield, trajectory, and entry behavior. Signal characteristics are derived by applying the Progressive Multi-Channel Correlation method as an array technique. The comparison of the events with a NASA reference database as well as the application of atmospheric propagation modeling allows to draw conclusions about infrasound-based detection capability, localization accuracy, yield estimation, and source characterization. The infrasound technique provides a time- and location-independent remote monitoring opportunity of impacting near-Earth objects (NEOs), either independent or complementary to other fireball observation methods. Additionally, insights about the detection and localization capability of IMS infrasound stations can be gained from using large fireballs as reference events, being of importance for the continuous monitoring and verification of atmospheric explosions in a CTBT context.

How Can the International Monitoring System Infrasound Network Contribute to Gravity Wave Measurements?

Gravity waves (GWs) propagate horizontally and vertically in the atmosphere. They transport energy and momentum, and therefore GWs can affect the atmospheric circulation at different altitude layers when dissipating. Thus knowledge about the occurrence of GWs is essential for Numerical Weather Prediction (NWP). However, uniform networks for covering GW measurements globally are rare, especially in the troposphere. It has been shown that an infrasound station of the International Monitoring System (IMS) infrasound network is capable of measuring GWs at the Earth’s surface. The IMS was deployed for monitoring the atmosphere to verify compliance with the Comprehensive Nuclear-Test-Ban-Treaty. In this study, the Progressive Multi-Channel Correlation Method (PMCC) is used for re-processing up to 20 years of IMS infrasound recordings in order to derive GW detections. For this purpose, two alternative PMCC configurations are discussed, covering GW frequencies equivalent to periods of between 5 min and 150 min. These detections mainly reflect sources of deep convection, particularly in the tropics. At mid-latitudes, coherent wind noise more often produces spurious detections. Combining the results of both configurations provides a global dataset of ground-based GW measurements, which enables the calculation of GW parameters. These can be used for improving NWP models.

Improved infrasound array processing with robust estimators

Accurate infrasound source and path characterization rely on high-quality array processing parameter estimates. Physical and statistical assumptions underlying conventional array processing techniques sometimes fail in practice due to propagation effects or station degradation. Unlike conventional least squares regression, robust regression estimators are relatively insensitive to data that deviate from the assumed planar model. We compare two such estimators, M-estimators, and least-trimmed squares (LTS), to conventional array processing methods (frequency-wavenumber beamforming, progressive multi-channel cross correlation, L1 regression, and ordinary least squares) using synthetic and real infrasound data. Synthetic testing suggests that robust estimators are resistant to timing errors and noise contamination. We also present case studies from both International Monitoring System and the Alaska Volcano Observatory infrasound data that demonstrate how these techniques have produced accurate array processing results despite an element polarity reversal, timing error due to the loss of GPS lock, and a deviation from the plane wave assumption. We also evaluate the effectiveness of these techniques to arrays with differing geometries and number of elements, and note that the examination of LTS residuals enables outlying inter-element differential times to be flagged automatically, providing a data quality tool.

Progresses with the International Data Centre Infrasound System

The International Data Centre (IDC) advances its methods and continuously improves its automatic system for infrasound technology. The IDC focuses on enhancing the automatic system for the identification of valid signals and the optimization of the network detection threshold by identifying ways to refine signal characterization methodology and association criteria. An objective of this study is to reduce the number of associated infrasound arrivals that are rejected from the automatic bulletins when generating the reviewed event bulletins (REB). Progresses related to several ongoing projects at the IDC will be review:—improving the detection accuracy at the station processing stage by introducing the infrasound signal detection and interactive review software DTK-(G)PMCC (Progressive Multi-Channel Correlation) and by evaluating the performances of detection software;—development of the new generation of automatic waveform network processing software NET-VISA to pursue a lower ratio of false alarms over G...

Assessment of Infrasound Detectors Based on Analyst Review, Environmental Effects, and Detection Characteristics

To assess infrasound detector performance, automated detections by the progressive multichannel correlation method (Cansi, 1995) and the adaptive F ‐detector (AFD; Arrowsmith et al. , 2009) are compared with signals identified by five independent analysts. Each detector was applied to a 4‐hr time sequence recorded by the Korean seismoacoustic array, CHNAR, composed of small (<100 m) and large (∼1000 m) aperture subarrays. Detector effectiveness was estimated for a selection of array elements and detection thresholds under low‐ and high‐noise conditions. Estimated receiver operating characteristic based on events identified by analysts evaluates the change in detection probability ( P d ) and false‐alarm probability ( P f ) for various detector parameters. This empirical study documents that the use of smaller aperture subarrays by both detectors increases P d with smaller p ‐values recommended for AFD to minimize P f . P d is impacted most by noise level, as shown by an increase in detections for average root mean square amplitudes from 1.2 to 3.2 MPa. Critical to this assessment is the identification of the source of the noise, constrained by signal characteristics, complementary seismic observations, and realistic atmospheric modeling. Based on signal characteristics (correlation value, phase velocity, and detection azimuth) and raytracing using global and local weather datasets, we conclude that during low‐noise conditions some detections from local distances (10–50 km) are affected by surface wind direction, and a second set is affected by tropospheric winds. This illustrates the role that surface and higher‐atmosphere winds play in array performance when assessing signals from regional infrasound sources in which local detections may be considered as noise or clutter. [Electronic Supplement:][1] Figures showing summary of detection results and polar plots of correlation estimates and phase velocity with respective to azimuth. [1]: http://www.bssaonline.org/lookup/suppl/doi:10.1785/0120160125/-/DC1

Characterizing infrasonic ocean ambient noise in Northwest Pacific

The ability of the International Monitoring System (IMS) infrasound network to detect atmospheric explosions and events of interest strongly depends on station specific ambient noise which includes both incoherent wind noise and real coherent infrasonic waves. To characterize the coherent ambient noise, a broadband array processing was performed with IMS continuous waveform archive from 2007 to 2016 using the Progressive Multi-Channel Correlation algorithm (PMCC). The processing parameters include a new implementation of adaptive frequency dependent window length and a logarithmic band spacing. Such configuration allows to better discriminate between interfering signals with improved accuracy in the wave parameters estimations. Multi-year comparisons between the observed and modeled directional microbarom amplitude variations at several IMS stations using two-dimensional wave energy spectrum ocean wave products are performed to build of a reference database of infrasound oceanic sources in Northwest Pacific. The expected benefits of such studies concern the use of multi-year complementary data to finely characterize coupling mechanisms at the ocean-atmosphere interface. In return, a better knowledge of the source of the ambient ocean noise opens new perspectives by providing additional integrated constraints on the dynamics of the middle atmosphere and its disturbances where data coverage is sparse.

Detection and classification of seismic events with progressive multi-channel correlation and hidden Markov models

The progressive multi-channel correlation (PMCC) detector and a statistical time series classifier based on hidden Markov models (HMMs) are associated in order to detect and classify seismic events at an array station. We classify the signal patterns that have triggered PMCC detections thanks to HMMs. The signal is transformed into sequences of feature vectors, which are used to train HMMs. The features include outputs from the PMCC detector describing the seismic wave propagation. Our method is applied to the automatic classification of regional seismic event, teleseismic event, and noise. The experimental data set comprises records from the Songino array station in Mongolia. Results on test events show an average classification performance of 84%.

Optimizing the PMCC Algorithm for Infrasound and Seismic Nuclear Treaty Monitoring

We introduce novel methods to determine optimum detection thresholds for the Progressive Multi-Channel Correlation (PMCC) algorithm used by the International Data Centre (IDC) to perform infrasound and seismic station-level nuclear-event detection. Receiver Operating Characteristic (ROC) curve analysis is used with real ground truth data to determine the trade-off between the probability of detection (PD) and the false alarm rate (FAR) at various detection thresholds. Further, statistical detection theory via maximum a posteriori and Bayes cost approaches is used to determine station-level optimum “family” size thresholds before detections should be considered for network-level processing. These threshold-determining methods are extensible for family-characterizing statistics other than “size,” such as a family’s collective F-statistic or signal-to-noise ratio (SNR). Therefore, the reliability of analysts’ decisions as to whether families should be preserved for network-level processing can only benefit from access to multiple, independent, optimum decision thresholds based upon size, F-statistic, SNR, etc.

Coherent ambient infrasound recorded by the International Monitoring System

Ambient noise recorded by the International Monitoring System (IMS) infrasound network includes incoherent wind noise and coherent infrasonic signals; both affect detection of signals of interest. We present summary statistics of coherent infrasound recorded by the IMS network. We have performed systematic broadband (0.01-5 Hz) array processing of the IMS historical dataset (39 stations from 2005 to 2010) using an implementation of the Progressive Multi-Channel Correlation (PMCC) algorithm in log-frequency space. From these results, we estimate multi-year 10th, 50th, and 90th percentiles of the rms pressure of coherent signals in 15 frequency bands for each station. We compare the resulting coherent noise models in the 15 frequency bands with raw power spectral density noise models, which inherently include both incoherent and coherent noise. We show that IMS arrays consistently record coherent ambient infrasound across the broad frequency range from 0.01 to 5 Hz when wind-noise levels permit. Multi-year averaging of PMCC detection bulletins emphasizes continuous signals such as oceanic microbaroms, as well as persistent transient signals such as repetitive volcanic, surf, or anthropogenic activity (e.g., mining or industrial activity).