In 1960 sound from explosive charges detonated off Perth, Australia was detected at Bermuda in the North Atlantic. Click on the figure to see the paths the sound may have taken to get to Bermuda.

In 1960, in a culmination of decades-long research on long-range acoustics by the Lamont Geological Observatory of Columbia University, Palisades, New York, a sequence of nine 200- and 300-lb explosive charges were detonated just west of Perth, Australia. Lamont's R/V Vema was used to deploy six 200-lb charges at 5-minute intervals on 1 March 1960, while the Australian Navy's HMAS Diamantina was used to deploy three 300-lb charges (for a higher signal level) three weeks later on 21 March 1960. The sounds from these explosions were detected by a hydrophone at the Observatory's Bermuda SOFAR Station in the North Atlantic, about 19,820 km away. The signals from the Vema were rather weak, so that the three additional shots with higher yield were requested; the Vema and Diamantina had co-operated during the Vema's test shots. Time was accurately determined at the Bermuda SOFAR station using the U.S. Department of Commerce's WWV radio timekeeping. Since the Bermuda SOFAR station was part of the U.S. Navy's developing acoustical locating system, a precursor to SOSUS, one can assume that the receiver location and time keeping were particularly accurate.

The R/V Vema explosions off Perth were located at 33.600°S, 113.483°E. The HMAS Diamantina explosions were located at 33.217°S, 113.717°E, about the same location as the Vema. The hydrophone of the Bermuda SOFAR station was located at 32.167°N, 295.417°E (64.583°W). The explosions occurred at different depths between 732 and 1800 m (the pressure detonators were not exact). The hydrophone receiver at Bermuda was located on the sound channel axis at 1323 m (and another nearby on the bottom - 3000 m?)

In the late 1980's, a question arose concerning the travel times that were recorded and the acoustic path that was required to give that travel time. While there is an acoustic path free of bathymetric obstructions along a geodesic on the earth's elipsoid (WGS84) between Perth and Bermuda, that turns out to be not quite the correct path. The ocean's thermal structure bends, or refracts, the path horizontally (Oceanography 101: It is colder at the poles than at the equator), so that the actual path is more complicated.

Munk, O'Reilly and Reid (1988) calculated an acoustic path, allowing for horizontal refraction, using a ray along the sound channel axis. This calculation suggested that the actual direct path would intersect Southern Africa (see figure below). It seemed there was no direct acoustic path between Perth and Bermuda, so that Bermuda was in the acoustic shadow of the African continent for sound from Perth. The calculation was not quite accurate, however, because the 2-D refraction calculation overestimates the refractive tendencies. The acoustic pulses cycle up and down in the water column, and spend most of their time in deeper water where horizontal sound speed gradients are less. Heaney, Kuperman and McDonald (1991) suggested that acoustic modes 1 and 2 have unobstructed, but bottom interacting, paths from Perth to Bermuda, with one path close to the unrefracted geodesic, but also another path passing just south of Africa (see also Munk, Worcester and Wunsch in their 1995 book "Ocean Acoustic Tomography"). Using the Del Grosso equation for the speed of sound in seawater (now known to be the most accurate of available equations), the measured travel times very closely match the calculated travel times, 13382 seconds, or 3 hrs, 43 min. and agree with the double pulse arrivals that were observed.

In any case, Munk et al concluded that global acoustic transmissions are very sensitive to oceanographic conditions. These considerations led to the notion that long-range acoustic transmissions might be a sensitive measure of the ocean's climate.

Assuming 2-D horizontal refraction at the sound channel axis, Munk, O'Reilly and Reid calculated the refracted path between Perth and Bermuda and found that the direct path would intersect Southern Africa. The two dimensional assumption is not quite right, however, and more accurate calculations based on mode phase speeds gives unobstructed acoustic paths between Perth and Bermuda (See figure).

Smoothed bathymetry between Perth and Bermuda along the geodesic path is shown in black. The colored background is sound speed (red is fast, blue is slow; the sound channel axis is blue). The magenta region shows an acoustic ray path with near-zero launch angle. Only acoustic energy very near the sound channel axis (low modes) can travel between Perth and Bermuda. In the Southern Ocean, the acoustic energy is trapped at the ocean surface. The figure is reversed from that above; Perth is at left, Bermuda at right. The large peak in bathymetry at about 4000 km range is the Kerguelen Plateau; Heard Island sits upon the Kerguelen Plateau.