Hydroacoustic research for monitoring marine animal behaviour has recently been appropriated for an entirely different purpose – the search for Malaysian Airlines Flight MH370.
When the aircraft first disappeared on 8 March and the search centred on the Gulf of Thailand, scientists from the United Nations’ Comprehensive Nuclear-Test-Ban Treaty Organisation (CTBTO) immediately analysed data from their underwater listening stations (intended to detect clandestine nuclear tests) south-west of Cape Leeuwin and in the Northern Indian Ocean, but did not ‘hear’ anything of interest.
But when the search abruptly shifted to the Indian Ocean, Dr Alec Duncan, Associate Professors Robert McCauley and Alexander Gavrilov of Curtin’s Centre for Marine Science and Technology (CMST) joined the search. They operate several hydroacoustic sensors as part of the Commonwealth Government-funded Integrated Marine Observing System (IMOS), and immediately recovered a passive acoustic recorder off Rottnest Island.
“The crash of a large aircraft would be a high-energy event, creating significant underwater noise,” explains Duncan. “We looked for a low-frequency signal, something energetic enough to travel a large distance underwater, and found a signal that was consistent with the timing of the disappearance of MH370.”
They immediately re-analysed the data from the CTBTO sensor off Cape Leeuwin, and found a signal, almost buried in the background noise, consistent with the Rottnest signal.
“The signal was extremely weak, but strongly correlated across all three independent hydrophones at Cape Leeuwin,” says Duncan.
“More importantly, those hydrophones are two kilometres apart – you can use the very slight difference in arrival time of the signal at each sensor to calculate the direction it came from.”
Comparing the difference in arrival time of the signal at Rottnest and at Cape Leeuwin then enabled calculation of an approximate distance to the source along the north-west bearing. Unfortunately, this placed the source further out in the Indian Ocean than the search area defined by the satellite ‘handshake’ data transmitted from the aircraft, considered the most reliable source of information about the crash.
Duncan elaborates: “Rottnest and Cape Leeuwin are relatively close together on the oceanic scale, so the accuracy of the calculation was limited. We’re confident that several sensors have detected the same event, but that may not be an aircraft impact. It is a low-frequency, high-energy signal with a fairly long ‘tail’ – it could be geological in origin. No-one knows what an aircraft crash sounds like underwater after being distorted by several thousand kilometres of ocean, so we can’t identify the spectral signature unambiguously.”
Limited data, collected opportunistically, and overlaid with the myriad other sounds of the ocean. But Duncan, Gavrilov and McCauley had calculated a potential crash site for the missing aircraft within a week of its disappearance, and passed the information onto the search authorities before MH370’s black box stopped transponding.
At the beginning of September CMST recovered another passive IMOS acoustic sensor, from Scott Reef off northern WA. McCauley and Duncan immediately identified a strong, complex signal that correlated with the Rottnest and Cape Leeuwin signals. After filtering out overlying whale calls, clicks and higher-frequency noise, they confirmed its arrival time. Because Scott Reef is much further away from Cape Leeuwin, it allowed a more accurate determination of the source position: close to the Carlsberg Ridge, a geologically active mid-ocean spreading ridge south-west of India.
“The satellite data and the acoustic data can’t both be correct, so our acoustic data is likely to be a signal of geological origin,” says Duncan. “But there are other pieces of evidence that make us reluctant to completely rule out the possibility of it being the aircraft: the timing is consistent with being shortly after the last aircraft transmission, and the location is where the MH370 would have run out of fuel if it had stayed on course from its last known position. You have to accept the satellite data, but if it eventually transpires that there is a problem with that data, our location is the first place you would search.”