The First 30 Years of Infrasound Research at UAF

In 1965 the United States National Bureau of Standards requested that the University of Alaska install an array of infrasonic microphones in Alaska to study a particular type of infrasound that seemed to be associated with geomagnetic activity. These signals were observed at low latitude stations. They appeared to be coming from a source at high latitudes that moved from NE to NW as the night progressed. The Geophysical Institute at the University of Alaska at Fairbanks UAF, as a center for the study of auroral and geomagnetic storm phenomena, was happy to accept the challenge. The NBS provided the equipment for this first Alaskan infrasonic array. Four capacitor-type NBS microphones (N-4) were installed in a spaced-array with an aperture of about 5 kilometers around the Fairbanks area. Signals from the array were brought back to the GI building with rented phone lines. A 1000 foot long noise-reducing Daniels type pipe was used at each microphone as a spatial filter to reduce boundary-layer wind noise. The analogue signal from each microphone was recorded on individual paper charts. Coherent signals were identified on the paper chart records by time-shifting the four individual superimposed waveform traces by hand on a light table to obtain waveform matching. From the sequence of delay times measured from the phase-aligned records the horizontal trace-velocities and the back azimuths of the signal were determined. The system was operated in two different pass bands: (1) from 0.015 to 0.10 Hertz for long period waves, and from 0.10 to 1.0 Hz for short periods.

During the first years of operation of the Fairbanks array it was discovered that the supersonic motion of auroral-electrojet arcs would produce infrasonic bow waves that propagate to the surface, with high trace-velocities, from azimuths that are parallel to the direction of motion of the source-arcs. Over the next decade two more infrasonic arrays were installed in Alaska to study auroral Infrasound. The first was at Palmer, Alaska southwest of Fairbanks, from 1967 to 1968. The second was north of Fairbanks on the Yukon River at Stevens Village from 1971 to 1973. A forth array was installed in Inuvik in the Northwest Territories of Canada to look for auroral infrasound north of the auroral oval from 1969 to 1971.  The auroral infrasound studies were extended to northern Europe in Sweden at the Kiruna Geophysical observatory from 1973 to 1976. Kiruna and Fairbanks have the same geomagnetic high latitude so that auroral infrasound observations could be compared at stations along the same parallel of geomagnetic latitude at stations separated by 180 degrees of longitude.

In Antarctica at 77.7 degrees south latitude in Windless Bight near the US station at McMurdo Sound an infrasonic array was operated from 1976 to 1985 in order to observe natural infrasound in the South Polar Region. Auroral related infrasound was found to be present at all six of the above mentioned high latitude infrasonic arrays. Natural infrasound from volcanic eruptions, earthquakes, marine storms, bolides, avalanches and mountain associated lee-wave turbulence have all been observed at each of the  six  infrasonic arrays operated by the Geophysical Institute of the University of Alaska. Infrasound produced by the enormous eruption in 1980 of Mt. St. Helens in Washington state (Figure 1) was observed at Fairbanks and in Windless Bight, Antarctica.

Recent History of Infrasound Research at UAF

In 1996 the G.I. at UAF converted an historic set of magnetic digital infrasonic data tapes from the Fairbanks and Antarctica infrasonic arrays into a new set of CD-ROMs using the new CSS 3.0 format for the infrasonic data. This historical infrasonic data from 1980 to 1983 was used in 1999 to produce a Library of Typical Infrasonic Signals in a four volume set as well as a 9 volume set titled Identification of Coherent Infrasonic Signals in an Historical Data Set as observed at the Windless Bight and Fairbanks infrasonic arrays.

In the summer of 1999 the Geophysical Institute was funded by the US Department of Energy to install a four-sensor infrasonic array at Fairbanks to test its operation under Arctic conditions. Part of the task was to develop a field-type calibrator for Chaparral Model 4 microphones (a prototype of the later Model 5) that could be used in the Arctic and the Antarctic. A wind-noise reduction system using short lengths of soaker hoses was tested during winter conditions and compared to the noise reduction of (1) a 305 meter Daniels pipe; (2) a 70 meter hexagonal ring of rigid pipe vented with hypodermic needles every five feet, and (3) a 30 meter aperture radial array of 12 soaker hoses. The 1999 Fairbanks array geometry is shown in Figure 2 below. A microphone enclosure used in the 1999 array is shown in Figure 3 below together with the calibrator unit used for field calibrations of the sensors.

As a result of the Fairbanks 1999 prototype array study we proposed that an eight sensor array of microphones be used in a pentagonal pattern of 5 sensors with a 1.7 kilometer aperture surrounding an equilateral triangle of three sensors at its center for both the infrasonic arrays in the Arctic at Fairbanks and in the Antarctic at Windless Bight. The impulse response diagram for the newly proposed pentagon-triangle array has almost no spatial aliasing over the passband from 0.02 to 10 Hertz. Later in 2000, our new array design using a pentagon-triangle pattern of sensors was accepted as an optimum design for Fairbanks and the Windless Bight sites. The pentagon-triangle 8 microphone array can be seen in Figure 4 on an aerial view of Windless Bight with Erebus Volcano in the background.

The University of Alaska was requested to participate in the installation and operation of infrasonic stations in the Arctic and Antarctic. The Geophysical Institute was tasked with the installation and operation of infrasonic arrays both in Fairbanks, Alaska and in Windless Bight, Antarctica. Our first task was to go the US Antarctic Palmer Station on Anvers Island off the Antarctic Peninsula in January of 2000 to set up an infrasonic test array to make measurements of the background wind noise level over a one month period. Model # 2 Chaparral microphones and R. M. Young model 05103V wind instruments were deployed at Palmer Station, at Torgersen Island, at Janaus Island and at the location of the old British Site. Infrasound from the calving of icebergs in Author Harbor near Palmer station as well as microbaroms signals from marine storms during quiet periods at the microphone sites provided the necessary assessment data. We concluded from our one month period of operation at the test array at Palmer Station that this would be another suitable Antarctic site for an infrasonic array.

We returned to Antarctica in January of 2001 to install an infrasonic array in Windless Bight at the same location where UAF had operated the first Antarctic infrasonic array from 1976 to 1985. Within 47 days an 8 microphone array was installed using surface-laid wires from a centrally located hybrid diesel/photovoltaic generator to supply 65 volts dc to each individual microphone site. The digital output of each Chaparral Model 5 microphone as well as the meteorological data were telemetered to McMurdo Station from each microphone site. The quality of the infrasonic data as well as that from the hybrid power supply in Windless Bight was monitored at the McMurdo infrasonic hub. Data is sent by satellite in real-time to the Geophysical Institute for analysis and operational quality-control monitoring.

On the University campus in May of 2002 work was begun on the installation of the Fairbanks infrasonic array in the forest north of the University campus where 20 years of   previous infrasonic array operations had proven it to be a very quiet wind noise environment. A pentagon-triangle array, with an aperture of 1.7 km, of eight Model 5 Chaparral microphones was built using commercial power at each microphone site. Wind noise reduction at each site was accomplished using a set of four rosettes of radial pipes that are vented at the end of each pipe. The resulting wind reducer at each site is a total of 18 meters in diameter with a total of 96 low impedance inlets, with 24 inlets installed on each rosette. The microphone, the digitizer, an ac to dc power panel and a Free-Wave radio at each site are housed in a 40 inch square box made of ¼ inch thick steel to protect the equipment. Infrasonic pressure time-series data are telemetered from each site to the Geophysical Institute building. The monitoring of the data from both the Fairbanks and Windless Bight arrays is done at the infrasonic hub in an air-conditioned room using both manual observation of the real-time data and an automatic computer system. Over the several years of operation of these two infrasonic arrays the data reliability has been near the 99% level. The infrasonic microphone output data from both arrays are stored   locally on hard drives. The infrasonic data is eventually all transferred to CD-ROMs for archiving at the GI of UAF.

Many man-made and naturally occurring infrasonic signals have been analyzed from these two infrasonic array sites from the following disparate sources. Natural Sources:  volcanic eruptions, avalanches, auroral and geomagnetic activity, large Alaskan earthquakes,  microbaroms from marine storms in the Gulf of Alaska and the Ross sea in Antarctica,  the floating  Ross Ice-Shelf interactions with the  Ross Island land mass, Mount Erebus Strombolian eruptions and volcanic-tremor, mountain associated infrasound, large bolides. Man-Made Sources: Aircraft generated sonic-booms, cannon fire and bombs at the USAF Tanana Valley test ranges, excavation explosions at Alaskan gold mines, aircraft crashes, and helicopter and fixed-wing aircraft, rocket launches from Poker Flat Range north of Fairbanks. The very large infrasonic signal from the crash of an A-Ten aircraft on February 26, about 20 km southeast of the  Fairbanks array is shown in Figure 5.