Infrasound is radiated by a variety of geophysical processes including earthquakes, severe weather, volcanic activity, geomagnetic activity, ocean waves, avalanches, turbulence aloft, and meteors. The general properties of these signals are described in the context of the measurement challenges presented in detecting them. A brief history provides background concerning the evolution of infrasonic detection technology. Recent improvements in both hardware and processing software have made passive detection and identification of infrasonic sources on a continuous basis practical and should lead to valuable operational applications. These hardware and software advances will be described. The detection of meteors, meteorites, and space debris is an area reviewed to indicate the capabilities and uses of infrasonic observing systems. The fact that infrasonic systems together with seismic, hydroacoustic, and radionuclide systems are planned for the International Monitoring System offers wide opportunities for future synergistic research and some of these are indicated. Finally, potentially valuable geophysical applications are summarized.

1. INTRODUCTION AND DEFINITION OF TERMS
Infrasound is radiated from many geophysical processes and from some civilization sources. Figure 1 shows typical signal sound pressure levels as a function of frequency, using the threshold of human hearing as a reference. The vertical scales are in dB relative to .0002 microbar , the threshold of human hearing (left scale) and also the absolute pressure in microbars (right scale). The threshold of hearing and the threshold of feeling cross over at a frequency of about 20 Hz, which means that frequencies below this point are felt rather than heard. Other reference points on this plot include the levels and frequencies of physiological noise and typical hydrostatic pressure changes produced by the small altitude changes involved with running or walking. This figure is intended to provide a reference point for understanding the range of minute pressure changes usually occurring for atmospheric infrasound.

Figure 1.

Figure 2 summarizes the amplitudes of a variety of infrasonic signal types as a function of period range in seconds over which they usually occur. This figure also presents suggested definitions to describe the different ranges. For example, periods between .05 to 1 second (20 to 1 Hz) are logically called near-infrasound (an analog to near-infrared). Periods between 1 and about 100 seconds
(1 to .01 Hz) define the range of infrasound (well below the human hearing), and signals with periods greater than 100 seconds can be designated acoustic/gravity waves to indicate that the restoring force of gravity as well as compressibility is determining signal properties.

Figure 2. Typical pressure amplitudes of infrasonic signals as a function of period in seconds with proposed definitions for the various frequency ranges indicated.

Figure 3 is a plot of wavelength as a function of signal period showing several important types of pressure disturbances that can be distinguished by spatial scale. The diagonal line marked acoustic velocity shows wavelength as a function of period for acoustic phase speeds with a variety of infrasonic signal types indicated. The upper signal class is for earthquake Rayleigh waves showing large wavelengths (typically 100 kilometers) compared with infrasound wavelengths for the same frequency range (typically 5 to 10 kilometers). For the lower speed waves on the plot (with a disturbance speed of 10 meters/second assumed ) the wavelength is much shorter than an acoustic wave at the same frequency (10's of meters compared with kilometers). This explains why spatial filtering can be used to advantage for wind and gravity wave noise reduction, improving the detectability of infrasonic signals.

Figure 3. Wavelengths or scales of pressure signals as a function of period, emphasizing three regimes: seismic, acoustic, and gravity wave/turbulence.

2. INSTRUMENTATION
Infrasonic instrumentation has been reviewed by Cook and Bedard¹ (1972). Sensors designed up until the 1970's provided sensitivity capabilities (noise levels < .001 microbar) well in excess of those required for signal detection. This is because wind noise or sounds from infrasonic sources invariably determine detection threshold limits. Higher quality electronics available today permit the use of simpler pressure sensing elements than applied in the past (e.g. Cordero et al.² (1957)). Thus, efforts directed toward improving spatial filters provide the most potential payoff in terms of achieving all weather operation in the presence of winds and turbulence. The spatial filtering approach devised by Daniels³ (1959) consisted of long (1000 foot ) lengths of tapered pipe with flow resistors at intervals to provide areal sampling of pressure signals using a single sensor. Spatial filtering approaches are described further by Bedard⁴ (1977) and an example of a recent space filter design is shown in Figure 4. More recently, porous irrigation hose has replaced the metal pipes which were relatively expensive to construct and maintain. Wind noise reduction remains an area where improvements can be made.

Another important design factor is the fact that pressure sensing elements, because of dynamic range considerations, are used with high-pass pneumatic filters, which can introduce amplitude and phase changes on detected signals. Thus, because an infrasonic observatory applies arrays of microphones to beamstear and determine properties of propagating acoustic waves, it is important to closely match the time constants of all array elements. Pressure calibration techniques can insure that proper amplitude and phase matching is achieved.

Figure 4. Example of a space filter used with infrasonic microphones for noise reduction.

3. PROCESSING TECHNIQUES
Because typical infrasonic signals show little or no change in either amplitude or frequency relative to the background noise, there is a need to perform array processing. This processing usually presumes a plane wave front and that the acoustic waves remain coherent as they propagate over the array. Typical array dimensions are about a quarter of the acoustic wavelength of interest. Thus, spatial correlation is exploited to identify infrasonic waves in the presence of spatially uncorrelated noise fields. Einaudi et al.⁵ (1989) describe the processing approach used to detect signals. An example of the power spectral output as a function of time is shown in Figure 5. This is for the cross correlated, summed signals of four individual channels. Interestingly, the algorithms are largely based upon an analog correlator (Brown⁶). These analog computing systems successfully and efficiently analyzed infrasonic signals well into the late 1970s. We envision that the parallel application of a number of processing techniques will lower detection thresholds, increase confidence in signal parameters retrieved, and help discriminate between signal types. Even though the seismic, hydroacoustic, and electromagnetic communities have greatly differing signal and noise environments, there should be significant payoffs in adapting cross-disciplinary processing approaches.

Figure 5. Spectral intent of an infrasonic signal associated with an avalanche or a function of time.

4. EXAMPLES OF INFRASOUND DETECTION
AND IMAGING OF NATURAL AND CIVILIZATION PROCESSES
There is substantial evidence that a variety of phenomena radiate infrasound. However, our understanding of sound generation mechanisms is limited. In addition to developing theoretical models, we need to measure infrasound under conditions when the source region is defined by other remote sensing systems. This area-of-need is highlighted in section 6, where opportunities for synergistic studies are outlined.

Table 1 summarizes a range of phenomena, providing references, and indicating imaging potentials. Some items are identified with a question mark in parenthesis, signifying that at this point the capability may be possible, but has not been demonstrated.

Table 1. Infrasonic Observatories and Potential Areas for Data Interpretation and Imaging

More detail on meteor detection from Bedard and Bloemker¹⁴ emphasizes the potential of infrasonic observing systems for providing important information concerning geophysical phenomena. One important capability is distinguishing meteors interacting explosively with the lower atmosphere from those producing bow shock waves at high altitudes. Even at long ranges (> 1000 kilometers), it may be possible to triangulate on lower - atmospheric impact points using multiple observations and guide the recovery of meteorite fragments. Figure 6 shows an example of long range detection using a single observatory. The explosive interactions act as a point sources, while the higher altitude shocks often show changes in azimuth, elevation angle, and frequency as a function of time. For example, we have recorded frequency shifts from 1 to 5 Hz over intervals of several minutes. On the other hand, many meteor events show no frequency shifts.

The progressive frequency changes often observed along a meteor trail may indicate changes in meteor mass with time. Theory predicts larger meteors to have lower dominant acoustic frequencies. Thus, as ablation reduces meteor mass a shift toward higher frequencies is expected. The great differences in the magnitudes of frequency shifts observed suggest it may be possible to deduce information about composition from the acoustic signatures. Inferring ablation rates acoustically seems a fruitful area for future research.

Figure 6. An example of long range meteor detection from a single infrasonic observatory. This is related to a meteor near Bakersfield, California on October 4, 1996.

Other infrasonic measurements of signals with the characteristics of meteors (shifting azimuths, high acoustic phase speeds, and short durations of a few minutes or less) indicate that the numbers of meteoritic objects may be larger than observed by conventional means. For example, on October 3-4, 1996 during about a 21 hour interval we detected 53 signals, we infer to have originated over the continental United States. Other observing periods detected shock like signals moving over head (e.g. 12 in less than an hour in one case) during times of no known meteor swarm. When coordinated with other observing techniques (e.g. radar, satellite, surface optical), infrasonic systems could help expand our knowledge of the statistics of meteors. The planned International Monitoring System for nuclear test ban verification will be ideal for performing such studies.

5. OPPORTUNITIES AND USES FOR ACTIVE SOUNDING AT
LOW AUDIO AND INFRASONIC FREQUENCIES
Sound detection and ranging (SODAR)²¹ is an active remote sensing technique primarily for studying the lower boundary layer of the atmosphere. Combined acoustic/electromagnetic systems called radio acoustic sounding systems (RASS)²² provide atmospheric temperature profiles. Because of reduced atmospheric attenuation, the use of lower acoustic frequencies can extend the range capabilities of these systems. Moreover, other opportunities for sounding exist and Table 2 reviews some of these.

Table 2. Further Possibilities for Low-Frequency Acoustic Sounding

6. OPPORTUNITIES FOR SYNERGISTIC STUDIES COMBINING INFRASONIC OBSERVATORIES WITH OTHER REMOTE SENSING SYSTEMS
Table 3 outlines remote sensors valuable in combination with infrasonic systems to provide more complete interpretations of both sound generation mechanisms as well as the dynamical processes causing the sound. Once acoustic source mechanisms are understood techniques can be applied with more confidence for detection, monitoring, and warning applications. For example, a theory suggested that avalanche signal generation involves roll instabilities modulating changes of state, hence producing cyclic temperature changes. Experiments measuring avalanche dynamics with a Doppler radar, Doppler lidar, and radiometers measuring temperature changes will give critical information to test this infrasonic generation mechanism. A range of important collaborative studies are possible.

Table 3. Opportunities for Synergistic Studies Combining Infrasonic Observatories with Other Remote Sensing Systems

7. Resource Materials on Infrasound
This section provides a combination of review materials and bibliographies as resources for additional background on infrasonics. Thomas et al (et al.²⁷,1971) provide a bibliography on infrasonic waves covering instrumentation and data analysis, source mechanics, insopheric effects, ground level observations, and theory. Georges and Young²⁸ (1972) published an extensive review focusing on geophysical sources at lower frequencies. More recently, a National Academy of Sciences report reviews the application and research needs for nuclear test ban international monitoring systems (Lay et al.²⁹, 1997). A report by McKisic³⁰, focused on the monitoring of atmospheric nuclear explosions, also provided an annotated review of papers on natural and civilization sources of infrasound. Finally, an excellent general reference is the text of Gossard and Hook³¹ who summarize the generation, propagation and detection of infrasound and acoustic-gravity waves.

8. REFERENCES

1. Cook, R.K. and A.J. Bedard Jr "On the measurement of Infrasound" Q.J. Roy. Astro. Soc. 67,pp 5-11, 1972.

2. Cordero,F., H. Matheson, D.P. Johnson "A nonlinear instrument diaphragm" J. Res. Natl. Bur. Standards, 58, pp333-337, 1957.

3. Daniels, F. B. "Noise reducing line microphone for frequencies below 1 Hz" J. Acoust. Soc. Amer. 31, pp529-531, 1959.

4. Bedard, A. J. Jr "The DC summator: theoretical operation, experimental tests, and possible practical uses" Fluidics Quart., 9, pp26-51, 1977.

5. Einaudi, F., A. J. Bedard Jr., and J. J. Finnigan "A climatology of gravity waves and other coherent disturbances at the Boulder Atmospheric Observatory during March-April 1984" J. Atmos. Sci., 46, pp303-329, 1989.

6. Brown, R. S. "An automatic multi-channel Correlator" J. Res. Natl. Bur. Standards, 67C(1), pp33-38, 1962.

7. Bedard, A. J. Jr., G. E. Greene, J. Intrieri, and R. Rodriguez "On the feasibility of detecting and characterizing avalanches by monitoring radiated subaudible atmospheric sound at long distances" Proc. a multidisciplinary approach to snow engineering, Santa Barbarbra, Ca, pp 267-275, 11-14 July 1988.

8. Bedard, A.J.Jr. "Evaluation of atmospheric infrasound for monitoring avalanches" Proc. 7th Int'l Symp. on Acoustic Remote Sensing and Associated Techniques of the Atrmosphere and Oceans, Boulder, Co, 3-5 Oct. 1994.

9. Bedard, A. J. Jr. "Seismic response of infrasonic microphones" J. Res. Natl. Bur. Standards, 75C(1), pp41-45, 1971.

10. Donn, W. L. and E. S. Posmentier "Ground-coupled air waves from the great Alaskan earthquake" J. Geophys. Res., 69, pp5357-5361, 1965.

11. Donn, W.L. and M. Ewing "Atmospheric waves from Nuclear explosions" J. Geophys. Res., 67, pp1855-1866, 1962.

12. Greene, G.E. and A. J. Bedard Jr. "Infrasound from distant rocket launches" NOAA Tech Memo, U. S. Dept. Commerce, 1986.

13. Wilson, C. R. "Auroral infrasonic waves" J. Geophys. Res., 74, pp1812-1836, 1969.

14. Bedard, A. J. Jr. and R. Bloemker "Detection of space debris and meteor impacts using atmospheric infrasound" Proc. Annual Conf. SPIE, The International Soc. for Optical Engineering, 3116, pp177-191, July 1997.

15. Cotton, D. and W. L. Donn "Sound from Apollo rockets in space" Science, 171, 1971.

16. Daniels, F. B. "Generation of infrasound by ocean waves" J. Acoust. Soc. Amer., 34, pp 352, 1962.

17. Georges, T. M. "Infrasound from convective storms: examining the evidence", Rev. Geophys. Space Phys., 11, pp571-593, 1973.

18. Bedard, A. J. Jr. " Infrasonic detection of severe weather" Proc. 19th Conf. on Severe Local Storms Minneapolis, Minn., 14-18 Sept.1998.

19. Bedard, A. J. Jr, and R. Craig "Infrasonic detection of atmospheric turbulence in the vicinity of mountains" 8th Conference on Mountain Meteorology, Flagstaff, Az, 3-7 August 1998.

20. Delclos, C., E. Blanc, P. Broche, F. Glangeaud, and J.L. Lacoume "Processing and interpretation of microbarograph signals generated by the explosion of Mount St. Helens" J. Geophys. Res., 95, 5,484, 1990.

21. Bedard, A.J. Jr and M. J. Sanders "Thunderstorm-related wind shear detected at Dulles International Airport using a Doppler acoustic/microwave radar, a monostatic acoustic sounder and arrays of surface sensors" Proc. Conf. on Weather Forcasting and Analysis and Aviation Meteorology, pp 347-352, Oct. 16-19, 1978.

22. May, P. T., K. P. Moran, and R. G. Strauch "The accuracy of RASS temperature measurements" J. Appl. Meteorology, pp1329-1335, 1989

23. Donn, W.L. and D. Rind "Natural infrasound as an atmospheric probe" Geophys J. Roy. Astro. Soc., 26, pp111-133, 1971.

24. Bush, G. A., Ye. A. Ivanov, S.N. Kulichkov, and M. V. Pedanov "Estimation of the characteristics of a pulsed ground source by remote acoustic techniques" Izvestiya, Atmos. and Oceanic Physics, 25, pp861-866, 1989.

25. Tartarskii, V. I. "personal communication concerning the suggestion of A. M. Obukhov for the measurement of tornado wind speeds" 1998

26. Bedard, A.J. Jr. "Highway and Concert Noise Reduction" The Military Engineer, 88, no. 576, pp. 43-45, February - March 1996.

27. Thomas, J.E., A.D. Pierce, E.A. Flinn, and L.B. Craine. "Bibliography on Infrasonic Waves". Geophys. J. Roy. Astro. Soc. 67, pp399-425, (1971)

28. Georges, T.M. and J.M. Young. "Passive Sensing of Natural Acoustic Gravity Waves at the Earth's Surface". Chapter 12 in Remote Sensing of the Troposphere, V.E. Denn, Editor. U.S. Gov. Printing office. 1972

29. Lay, T. and the panel on basic research requirements in support of comprehensive test ban monitoring. "Research Required to Support Comprehensive Test ban Treaty Monitoring". National Academy Press, P. 138. 1997.

30. McKisic, J.M. "Infrasound and the Infrasonic Monitoring of Atmospheric Nuclear Explosions: A literature review". Final report PL-TR-97-2123. Department of Energy and Phillips Laboratory, National Tech. Information Service, p. 310. 1997.

31. Gossard, E.E. and W.H. Hooke "Waves in the Atmosphere". Elsevier, N.Y, N.Y., 1975.