Articles from UAF infrasonics group in the on-line journal Inframatics

………………………………………………………………….

Inframatics Number 13 March 2006

Infrasonic Array Observations at I53US of the 2006 Augustine Volcano Eruptions

By:  C. R. Wilson, John V. Olson, Curt A. L. Szuberla, Steve McNutt, Guy Tytagt, and Douglas P. Drob

Abstract

The recent January 2006 Augustine eruptions, from the 11th to the 28th, have produced a series of 12 infrasonic signals that were observed at the I53US array at UAF. The eruption times for the signals were provided by the Alaska Volcanic Observatory at UAF using seismic sensors and a Chaparral microphone that are installed on Augustine Island. The bearing and distance of Augustine from I53US are, respectively, 207.8 degrees and 675 km. The analysis of the signals is done with a least-squares detector/estimator that calculates, from the 28 different sensor-pairs in the array, the mean of the cross-correlation maxima (MCCM), the horizontal trace-velocity and the azimuth of arrival of the signal using a sliding-window of 2000 data points. The data were bandpass filtered from 0.03 to 0.10 Hz. The data are digitized at rate of 20 Hz. The average values of the signal parameters for all 12 Augustine signals are as follows: MCCM = 0.85 (STD 0.14), Trace-velocity = 0.346 (STD 0.016) km/sec, Azimuth = 209 (STD 2) deg. The celerity for each signal was calculated using the range 675 km and the individual travel times to I53US. The average celerity for all ten eruption signals was 0.27 (STD 0.02) km/sec. Ray tracing studies, using mean values of the wind speed and temperature profiles (along the path) from  NRL, have shown that there was propagation to I53US by both stratospheric and thermospheric ray paths from the volcano.

Inframatics Number 12, December 2005 pages 4 to 7

Infrasound Associated with Mt. Steller Avalanche

Kenneth M. Arnoult Jr., Charles R. Wilson, John V. Olson and Curt A. L. Szuberla

Abstract

Infrasound signals arising from a large avalanche were observed at the I53US array. The avalanche occurred on Alaska’s Mt. Steller on September 14, 2005. Some50 million cubic meters of rock and ice fell roughly 2600 m vertically and spread 9500m from the base. The mountain is located 540 km southeast of the Fairbanks array. .We analyzed several signals from the event and characterizes them in terms of spectral content, trace velocity, planarity, and elevation angle.

Inframatics Number 10 June 2005 pp 1 to 13

Infrasound from Auroral Electrojet Motions at I53US

Charles R. Wilson

Abstract

Geomagnetic data from the College International Geophysical Observatory (CIGO)in Fairbanks is used to illustrate the relationship of lower ionosphereic currents during times of auroral substorms and the resultant observation of auroral infrasonic waves(AIW) at the I53US infrasonic array in Fairbanks, Alaska. Auroral electrojet currents associated with auroral arcs produce perturbations in the geomagnetic field that are recorded at CIGO. The direction and magnitude of the total horizontal disturbance vector (THDV) perturbation from the auroral currents are compared with the direction of propagation of large AIW signals observed at I53US during the auroral substorm. If the magnetic field of the auroral currents, as measured at the surface, are approximated using an infinite line-current model, then it is possible to show that the THDV perturbation, at a time 5 to 7 minutes prior to the observation of the associated AIW, is approximately parallel to the direction of propagation of the AIW bow wave. The delay time of 5 to 7minutes for propagation of the AIW to the surface from 110 km altitude of the auroral arc is referenced to the time at which the THDV is a maximum. It has previously been shown that the AIW observed at I53US are bow waves generated by the lateral supersonic motion of auroral electrojet arcs.

Inframatics Number 9 March 2005 pages 27 to 30

Frequency Domain Coherence between High Trace-Velocity Infrasonic Signals at 153US and Video Data from Pulsating Aurora

Charles R. Wilson and John V. Olson

Abstract

This brief article on infrasound signals observed at I53US during an aurora event over Fairbanks investigates the frequency domain coherence between the luminous intensity of pulsating aurora and the high trace-velocity infrasound received at the surface. Video data of the aurora display from an All-Sky video camera on the night of December 5, 2003 was compared with the pressure waveform data from the eight microphone infrasonic array at Fairbanks. A propagation delay time for the infrasound sample, with respect to the video data sample, of five minutes was used. The frequency domain coherence was found to be high between the pulsating aurora intensity above the array and the high trace-velocity infrasound signals at the surface in the passband from 0.05 to 0.075 Hz. Ray-tracing studies indicate that such high trace-velocity infrasound signals are from a source within 35 Km of the zenith above the array for a source height of 110 km. The presence of pulsating aurora patches within this area during the time interval of received infrasound, and the high coherence between the video and infrasound data lead us to believe that periodic heating of the atmosphere by pulsating aurora is the actual source of the observed high trace-velocity infrasound.

Inframatics Number 9 March 2005 pages 31 to 35

I53US and I55US Signals from Manam Volcano

Charles R. Wilson and John V. Olson

Abstract

We have searched the infrasonic records from I53US array in Fairbanks, Alaska for infrasound from the Manam Volcano eruption in Papua New Guinea that occurred on January 27, 2005. A strong signal of 0.3 Pa p-t-p amplitude was found in the passband from 0.03 to 0.10 Hz on the 27th from 23:12 to 23:51 UT. The trace velocity of the signal was 0.330 ± 0.007 km/sec the back azimuth was 252.5 ± 1.3 degrees, the mean of the cross-correlation maxima was 0.860. These estimates were made using a plane wave model for the wave front and the small residual errors in the least-squares fit validate the model. Assuming that the eruption time at Manam Island was 14:00:00 UT the celerity for the Manam eruption signal received at I53US was 0.282 km/sec.

Inframatics Number 7 September 2004 pages 15 to 21

The Application of the Pure-State Filter to Infrasound Array Data

John V. Olson

Introduction

The problem of separation of signal from noise is of universal interest. In multivariate data sets and especially those that arise from arrays of sensors the unwanted noise and clutter signals may not be spatially uniform. In such cases adaptive processes are employed that recognize the noise content of the data stream and then perform signal detection and estimation in some efficient way.

The process we describe here is based upon the identification of pure states of information in the frequency domain through the analysis of the spectral matrices at each frequency. Fano [1] has defined pure states as “States of maximal information...” that are “...characterized by the existence of an experiment that gives a result predictable with certainty when performed on a system in that state...”. Operationally we identify pure states as those whose spectral matrix is characterized by a single eigenvalue. In vector processes these states are referred to as polarized and early versions of Samson’s work [3][4] in this field used this term to identify pure states.

In this manuscript we describe the construction of a measure of the degree of polarization (pure state) developed by Samson. Samson’s degree of polarization is constructed from invariants of the spectral matrix and serves as a measure of generalized coherence that is unaffected by the sensor or array geometry. It is applicable to both scalar and vector processes. Here we show its efficacy in the scalar data that result from arrays of infrasound sensors.

The application of the pure-state concept to a data set proceeds as follows. Data from an array of sensors is Fourier transformed and the spectral matrix is formed at each frequency estimate. Using Samson’s pure state estimator a value of the polarization, P2, is determined from the spectral matrices and is, therefore, a function of frequency. Interpreted as a filter, the Fourier transforms of the original data are multiplied by P2 and the modified transforms are inverted into the time domain producing a filtered version in which frequency components that represent coherent signals survive while those that represent noise are reduced.

Inframatics Number 7 September 2004 pages 22 to 25

A Study of the Effects of Snow Cover on Infrasonic Signal Levels: 1. Variations in the RMS levels.

Vanessa T. Santana and John V. Olson

Introduction

Infrasound sites have been operated successfully in climates where winter snow cover is common. Nevertheless, we are not aware of any studies of the effects of snow cover on the sensitivity of infrasonic arrays. We have begun a study of infrasound data collected at I53US, the CTBT-IMS station at Fairbanks, Alaska to see if we can detect changes in the infrasonic signal levels observed there as a function of the depth of the snow pack. We have surveyed a calendar year’s worth of data that includes approximately four months for which there was continuous snow cover. We find no significant change in the mean RMS levels of the infrasound wave field or the variance in the RMS levels over that year. In the next phase of this program we plan to study the effects on spectral content of the infrasound as a function of snow depth.

Inframatics Number 6 June 2004 pages 1 to 7

Infrasound Signal Detection using the Fisher F-Statistic

John V. Olson

Introduction

Fisher’s F-statistic has become an important method used in the detection of signals in infrasound array data. A detector based upon the Fisher F-statistic was first proposed by Melton and Bailey [5] and implemented for seismic arrays by several authors (see Blandford [2] for further discussion and references). Melton and Bailey noted the similarity between the ratios of variances to be tested in a series of statistical trials to the ratio of signal to noise power in signal processing.

The utility of the F-statistic as a detector in the frequency domain was shown by Shumway [8].

Inframatics Number 6  June 2004 pages 8 to 12

The Least Squares Estimation of the Azimuth and Velocity of Plane Waves

John V. Olson and Curt A. L. Szuberla

Abstract

We describe a procedure based upon least-square estimation to obtain the azimuth of arrival and the apparent speed of plane waves passing across a planar array. The method is based upon estimation of the signal time delays between sensor pairs and the best fit of a plane wave to that set of time delays. The method also allows errors in the estimates of velocity and azimuth to be made. 

Inframatics number 5 March 2004 page 20

Geophysical Institute Acquires Infrasound Sensor Technology

Daniel Osborne

The University of Alaska is now in the process of completing the purchase of the assets of Chaparral Physics Consultants of NM, Inc. An agreement was signed three months ago to transfer the technology, knowledge and assets of Chaparral, along with exclusive rights to all of the company’s products and name to the Geophysical Institute (GI) at the University of Alaska Fairbanks. The GI was selected as the successful candidate among many interested institutions seeking to acquire Chaparral’s assets.

Inframatics Number 4  December 2003 pages 1 to 8

Infrasound from Erebus Volcano at I55US in Antarctica

Charles R. Wilson, John V. Olson, Daniel L. Osborne, and Alexis Le Pichon

With the installation in 2001 of the I55US CTBT/IMS infrasonic array on the Ross Ice Shelf it has been possible to observe two types of infrasound associated with Mt. Erebus volcano, namely impulsive events and continuous events of long duration. The active crater on Mt. Erebus is about 25 kilometers to the west- northwest of the I55US 8 sensor infrasonic array at an azimuth of 337º. A survey of the infrasonic data at I55US for the period April 2002 to March 2003 was made to document further impulsive signals from the small Strombolian eruptions in the lava lake that frequently occur on Erebus. During this 12 month period 182 Strombolian type impulsive signals were observed at I55US from Erebus. The average back azimuth and trace velocity for these signals were respectively: 336.5 ± 0.39 deg and 320.5 ± 6.3 m/sec. The impulsive signals were found to be coherent over the entire 1.7 km aperture of the 8 sensor array. In the passband from 0.7 to 10 Hz the amplitude of the impulsive signals reached maximum amplitude of 0.5 Pascals. A typical Strombolian infrasonic signal is shown in Figure 1.

Inframatics Number 3 September 2003 pages 6 to 10

Mountain Associated Waves at 153US and 155US in Alaska and Antarctica in the Frequency Passband from 0.015 to 0.10 Hz

Charles R. Wilson and John V. Olson

Introduction

Atmospheric turbulence generated by mountain ranges that interrupt the tropospheric wind flow can produce aerodynamic infrasound that propagates thousands of kilometers from the source regions. These mountain associated infrasonic waves (MAW) have been observed for years by infrasonic arrays operated by the University of Alaska in Antarctica and in interior Alaska. Now that the CTBTO/IMS infrasonic arrays I53US at Fairbanks and I55US at Windless Bight, Antarctica have been installed, we have begun searching the data for  coherent infrasonic signals in two passbands: long period from 0.015 to 0.10 Hz, and short period from 0.10 to 10 Hz. The detection algorithm that we use in searching for coherent infrasonic waves that propagate across the sensor array is based upon the mean of each of the maxima of all the inter-microphone cross-correlations. That is, the normalized cross-correlation function is computed for each microphone pair and its maximum is identified. Next the mean of all of these maxima (mean (maximum (Cij)) is then defined to be the output value of the detection algorithm (MCCM) for that data window. For search analysis the data are segmented into small windows a few minutes long and the detection algorithm is applied to each window resulting in a series of detection values of the following parameters: (1) C = the mean(maximum(Cij)); (2) Vel = the trace velocity in Km/sec; and (3) Az = the azimuth of arrival in degrees for each data window in the time series of data. This set of values of: C, Vel and Az is then the final output of the MCCM algorithm. For the MAW signal search we scan an entire 24 hour record with a sliding window that is 10000 points or 500 seconds in length. We have found that the choice of the length of the data window is not of highest importance as long as it wide enough to contain a few cycles of the frequency of interest. MAW are long period waves therefore all the I53US and I55U data were first passband filtered from 0.015 to 0.10 Hz before application of the detection algorithm for signal search.

Inframatics Number 3  September 2003 pages 11 to 14

Infrasound Observed at 153US from Large Alaskan Earthquakes in 2002

Charles R. Wilson

Infrasonic waves that can be associated with earthquakes were observed with the microphone array at I53US in Fairbanks, Alaska on October 23 at 11:27:19 UTC and November 3 at 22:12:41 UTC. The magnitude of the 10/23 foreshock earthquake was 6.7 with an epicenter at 63.51N, 147.91W. The major earthquake on 11/03 had a magnitude was 7.9 with an epicenter at 63.52N, 147.53W, was just to the east of the 10/23 foreshock. The epicenters of both of these very large earthquakes were on the Denali Fault that runs through the Alaska Range about 90 miles to the south of Fairbanks.

Inframatics Number 2 June 2003 pages 10 to 18

Auroral Infrasound at Fairbanks, Alaska as Observed at CTBT Infrasonics Array 153US

Charles R. Wilson

Introduction

Infrasound signals that are associated with auroral activity have been observed at Fairbanks over the past 30 years with infrasonic microphone arrays operated with both analogue and digital data acquisition systems. The installation  greatly increased quality of the infrasonic data with which to study natural sources of infrasound. The factors that have contributed to the higher quality data from I53US are: (1) new Chaparral microphones of much greater sensitivity, (2) an extended passband from 0.015 Hz to 10 Hz, (3) an increase in the digital data acquisition rate from one Hz to 20 Hz., (4) the improved detection capability by using an eight-sensor pentagon-triangle array, (5) better signal detection and characterization algorithms that have been developed at UAF for I53US.

In the historic data set at Fairbanks all the AIW signals that were detected were found to be bow waves that are generated by the supersonic motion of auroral arcs that contained strong electrojet currents. This radiated infrasound was highly anisotropic propagating as a bow wave moving in the same direction as that of the auroral arc. The trace velocity of the AIW bow wave across the microphone array was the same as the velocity of the moving auroral arc. These bow wave AIW are usually impulsive in waveform and were relatively infrequent. Recently we have begun to detect a new type of AIW signal that appears to be associated with pulsating aurora displays. Pulsating auroras occur predominantly after magnetic midnight, (10:00 UT at Fairbanks). Pulsating auroras are a part of the recovery phase of the auroral substorm whenever it occurs (Akasofu, S.-I., Polar and Magnetospheric Substorms, D. Reidel, Hingham Mass., 1968). Visually pulsating auroras appear as irregular patches of luminosity turning on and off in a periodic fashion with periods from 2 sec to 20 sec. They are produced by energetic electrons precipitating into the atmosphere.

=================================================================================

Infrasound Articles in Published Journals

………………………………………………………………….

J. Acous. Soc. Am. Vol. 115 No. 1, pages 253 to 258, January 2004

Uncertainties associated with parameter estimation in atmospheric infrasound arrays

Curt A. L. Szuberla and John V. Olson

Abstract

This study describes a method for determining the statistical confidence in estimates of direction-of-arrival and trace velocity stemming from signals present in atmospheric infrasound data. It is assumed that the signal source is far enough removed from the infrasound sensor array that a plane-wave approximation holds, and that multipath and multiple source effects are not present Propagation path and medium inhomogeneities are assumed not to be known at the time of signal detection, but the ensemble of time delays of signal arrivals between array sensor pairs is estimable and corrupted by uncorrelated Gaussian noise. The method results in a set of practical uncertainties that lend themselves to a geometric interpretation. Although quite general, this method is intended for use by analysts interpreting data from atmospheric acoustic arrays, or those interested in designing and deploying them. The method is applied to infrasound arrays typical of those deployed as a part of the International Monitoring System of the Comprehensive Nuclear-Test-Ban Treaty Organization.

Geophysical Research Letters, Vol. 32, L14810, pages 1 to 4, July 2005

High trace-velocity infrasound from pulsating auroras at Fairbanks, Alaska

Charles R. Wilson, John V. Olson and Hans C. Stenbaek-Nielsen

Abstract

We report here on infrasound signals observed using the infrasound microphone array at Fairbanks, Alaska that are apparently associated with the presence of pulsating auroras near the zenith. We have investigated the frequency domain coherence between the luminous intensity of the pulsating auroras and the high trace-velocity infrasound received at the earth’s surface. Video data of the aurora display from an All-Sky video camera on the night of December 5, 2003 was compared with the pressure waveform data from the eight-microphone infrasonic array at Fairbanks. Enhanced coherence between the two signals was observed when a propagation delay time for the infrasound sample with respect to the video data sample was used. The frequency domain coherence was found to be high between the pulsating aurora intensity above the array and the high trace-velocity infrasound signals at the surface in the pass band from 0.03 to 0.08 Hz. Ray-tracing studies indicate that such high trace-velocity infrasound signals originate from sources within 35 km of the zenith above the array for a source height of 110 km. The presence of pulsating aurora patches within this area during the time interval of received infrasound and the high coherence between the video and infrasound data lead us to believe that periodic heating of the atmosphere by pulsating aurora is the actual source of the observed high trace-velocity infrasound.

J. Acous. Soc. Am. Vol 117, No. 3, pages 1032-1037, March 2005

Distribution of wave packet sizes in microbarom wave trains observed in Alaska

John V. Olson and Curt A. L. Szuberla

Abstract

This work reports on a study of the distribution of wave packet sizes contained in intervals of continuous microbarom activity. Microbaroms are a class of atmospheric infrasound that is characterized by narrow-band, nearly sinusoidal, waveforms with periods near 5 s. They are known to be generated by marine storms, presumably through a nonlinear interaction of surface waves, however the detailed analysis of the process is still incomplete. The data analyzed were obtained using the University of Alaska infrasound array of four microphones located in central Alaska. Because of the narrow-band feature of the microbarom signals, the Hilbert transom is applicable as a method for finding phase breaks in the signal. The phase breaks are interpreted as the demarcation of the boundaries of wave packets. When applied to long sequences of microbaroms a broad distribution of packet lengths is found that diminishes monotonically with length and has a mean near 10 cycles and a variance nearly as large. The distribution function decreases exponentially with packet length. The distribution of packet sizes is influenced by the presence of multiple sources and multiple propagation paths between the sources and the sensor array. Identification of individual packets should open the way to a more detailed analysis of microbarom wave trains. After separating the wave train into individual wavelets the inter-microphone correlation is estimated as a function of microphone separation. As has been observed in earlier microbarom studies, a decrease in correlation was observed for microphone pairs orthogonal to the direction of propagation when compared to correlations between microphones spaced along the direction of wave propagation.