The infrasonic array I53US, located in the forest north of the University of Alaska in Fairbanks, is surrounded by many sources of man-made infrasound. There is a bombing and gunnery range to the south of Fairbanks that is associated with nearby Eielson Air Force Base and the US Army base at Fort Wainwright. Explosions from bombs of up to 2000 pounds in size and the detonation of other ordnance south of Fairbanks are frequently detected by I53US. Mining excavation blasts at two large open-pit mines north of Fairbanks produce almost daily infrasonic signals at I53US. The research Poker Flat rocket range 34 km to the northeast produces detectable infrasound at I53US from large sounding rockets that have up to four successive stages of rocket motor ignition. Aircraft operations at both the military and civilian airfields around Fairbanks produce a great deal of local infrasound.

 The heating and ventilation fans in many UAF buildings produce infrasound tones that are within the CTBT/IMS infrasonic passband. In particular we have identified the IARC building next to the Geophysical Institute as the source of two constant tones at5 6.18 Hz and 6.37 Hz that appear in the I53US data in the 1 to 10 Hz passband.  These are very close to the Nyquist frequency of 10 Hz for the sampling rate of 20 Hz for I53US data. These two tones are virtually monochromatic.

Mine blast signals at I53US

Over its years of operation, the I53US array located in Fairbanks, AK has detected many signals from local mining activities.  The two local mines that have provided us with the most signals are Fort Knox and True North (Figure 1).  Both mines use ripple-fire surface blasting techniques to aid in extracting ore from the ground.  These surface explosions create low-frequency pressure waves in the air which can be detected at Fairbanks with the I53US array. There have been more than one thousand mine blast signals detected at I53US since 2002. The signals from mining operations are a very important resource for the Fairbanks Infrasound Group.  This is because these are signals which occur often through out all seasons and for which we are provided with background information.  Thanks to continuing cooperation with the engineers at Fort Knox and True North, the location, time, and magnitude of the detonations are known.  This allows us the unique opportunity of conducting a long term study to estimate the accuracy and uncertainties associated with our analysis tools.  It also allows us a chance to test our atmospheric modeling algorithm and determine if it can accurately predict which atmospheric profiles cause the signals to be refracted down onto the I53US array and which do not.

The mine signals are some of the clearest signals detected by the I53US array.  As seen in Figure 2, the amplitude of the average mine signal is much greater than that of the background noise.  The shape of the signal is referred as an N-wave due to fact that it resembles an upper case “N.”.  Explosions often result in N-waves because they create a region of high air pressure (compression) immediately followed by a region of low air pressure (rarefaction).  All eight sensors record nearly the same waveform differing only in arrival time.  Using these time delays, the trace-velocity and direction to the source can be estimated.

Plotted in the upper eight panels of Figure 2 is a mine signal from Fort Knox as detected by the eight sensors of I53US and filtered from 0.1 to 10.0 Hz. This signal is approximately 2 Pascals peak-to-peak and has the shape of an N-wave. Plotted in the bottom panel is a phase aligned overlay of all eight data series. By using the differences in arrival times, we estimate that this signal propagated through the array of sensors with a speed of 327.0 m/s (approximately the speed of sound) and originated 58.7 degrees east of north.

Rocket launches, Bombs, Ordnance Detonation, Howitzer shots, and an Aircraft Crash

When large multiple stage rockets are launched at the UAF Poker Flat research range, that is 34 kilometers to the north of I53US, infrasound from the ignition of each rocket stage can be observed at Fairbanks. An example is shown in Figure 3 of the infrasound wave train and the spectrogram observed on February 6, 2002 at I53US from the launch of a Black Brandt four stage rocket. The zero time in the wave train plot is at 09:40:00 UT. The first arrival from the ignition of the first stage can be seen at 09:40:32. The second arrival from the second stage ignition can be seen at 09:40:48. The wave train from the rocket continues for about 3 minutes. This rocket signal was recorded at sampling rate of 100 Hz at the DOE prototype array at UAF.

Infrasound from a helicopter passing over the DOE array was recorded as shown in Figure 4. The wave train and its spectrum showing the Doppler shift in the observed frequency of the signal as the helicopter approaches and then recedes from the array can be seen in Figure 4. A picture in Figure 4, of the Howitzer shot announcing the beginning of the Midnight Sun foot race on UAF campus, can be seen together with the observed infrasound signal observed by the DOE array.

In Figure 5 other infrasound signals observed with the 4 microphone DOE array can be seen from a large explosion near Delta Junction and the crash of a British jet aircraft near Eagle Alaska 100 miles east of Fairbanks. The azimuth to the crash location, as determined by the infrasonic signal by the UAF array, was helpful in identifying the crash site.

A very large signal with peak-to-peak amplitude of about 4 Pa was observed at I53US as is shown in Figure 6. This explosion signal resulted from the detonation of ordnance at the Blaire Lakes site south of Fairbanks.

A 2000 pound bomb dropped in the bombing and gunnery range in the Tanana valley south and east of Fairbanks produced the 7 Pa peak-to-peak amplitude signal at I53US in 2003 that is shown in Figure 7.

Howitzer signals were observed at I53US in the 2004 and 2005 from the start of the annual Midnight Sun run. The howitzer signal on June 20, 2004 at I53US is shown in Figure 8. These howitzer shots are from a location that is very close to the I53US array. They are used as calibration signals for the Locator software that was developed at UAF to give both the azimuth and range to a source that is within a distance of ten array apertures (17 Km) of the array center. The locator uses the curvature of the signal wave front to determine the range.

In Figure 9 a map is shown of the I53US array microphones (in blue) and the location of the howitzer (black circle marked 2004MS) as determined by the Locator software for the 2004 signal of Figure 8. The red X gives the actual ground truth position of the howitzer. One can see from Figure 9 that the Locator gives excellent results for sources that are just a few kilometers from the array.