Curtin researchers in search for acoustic evidence of MH370

Media release

Curtin University researchers have been examining a low-frequency underwater sound signal that could have resulted from Malaysian Airlines Flight MH370.

The signal, which was picked up by underwater sound recorders off Rottnest Island just after 1:30 am UTC on the 8th March, could have resulted from Flight MH370 crashing into the Indian Ocean but could also have originated from a natural event, such as a small earth tremor.

However, there are large uncertainties in the estimate and it appears it is not compatible with the satellite ‘handshake’ data transmitted from the aircraft, which is currently considered the most reliable source of information.

Scientists from Curtin’s Centre for Marine Science and Technology along with colleagues from the United Nations’ Comprehensive Nuclear-Test-Ban Treaty Organisation (CTBTO) and Geoscience Australia have been involved in the search for sounds that might help with search efforts.

Dr Alec Duncan, Senior Research Fellow and part of Curtin’s Centre for Marine Science and Technology team, explained that a passive acoustic observatory 40 kilometres west of Rottnest Island that forms part of the Commonwealth-funded Integrated Marine Observing System (IMOS) had provided the potential lead.

“Soon after the aircraft disappeared, scientists at CTBTO analysed data from their underwater listening stations south-west of Cape Leeuwin and in the northern Indian Ocean. They did not turn up anything of interest,” Dr Duncan said.

“But when the MH370 search area was moved to the southern Indian Ocean, scientists from Curtin’s Centre for Marine Science and Technology decided to recover the IMOS acoustic recorders located west of Rottnest Island.

“Data from one of the IMOS recorders showed a clear acoustic signal at a time that was reasonably consistent with other information relating to the disappearance of MH370.

“The crash of a large aircraft in the ocean would be a high energy event and expected to generate intense underwater sounds.”

Dr Duncan said the signal could also have been due to natural causes – such as a small earth tremor – but the timing made it of interest in the search for MH370.

“It has since been matched with a signal picked up by CTBTO’s station south-west of CapeLeeuwin.

“A very careful re-check of data from that station showed a signal, almost buried in the background noise but consistent with what was recorded on the IMOS recorder off Rottnest,” Dr Duncan said.

“The CTBTO station receives a lot of sound from the Southern Ocean and Antarctic coastline, which is why the signal showed up more noticeably on the Rottnest recorder.

“Using the three hydrophones from the Cape Leeuwin station, it was possible to get a precise bearing that showed the signal came from the north-west.

“Comparing the arrival time of the signal at the IMOS recorder with the time of its arrival at the Cape Leeuwin station, it was possible for Curtin’s Centre for Marine Science and Technology team to come up with an approximate distance to the source of the sound along this north-west bearing.

Dr Duncan said Curtin’s Centre for Marine Science and Technology team would continue to work with search authorities.

“Although we have now completed our analysis of these signals, Curtin’s Centre for Marine Science and Technology still has several recorders deployed that could conceivably have picked up signals relating to MH370.

“Due to various factors, we consider it very unlikely that they would have done so and have therefore not recovered them to date.  We will, however, be carefully analysing their recordings when they are recovered in due course,” Dr Duncan said.

High-resolution diagrams and a map detailing the estimated uncertainty region are
available to download through Curtin Media Images on Flickr.


Share your thoughts on this story (comments are moderated in advance).

This story has 19 comments

  1. Melissa Everton says:

    Where are “the details” such as maps??

    • Susanna Wolz says:

      Hi Melissa, you can find the maps by heading to our Media Centre via and scrolling to the bottom of the page for the photos.

      Thanks for your interest!

  2. Chris says:

    Where can a copy of the audio be obtained? We found the pictures thank you!,

  3. S.Sadasivan says:

    Intriguing ! Excellent caution on findings !
    Is it possible that the sound of aircraft still airborne with engine on (perhaps losing altitude as fuel is running out in the final moments of flight ) is also buried in the underwater recordings ?

    Can raw digital time signals be made available ?

    Wonderful work !

    Best wishes

    • Alec Duncan says:

      That is an excellent question. The huge differences between the properties of air and water mean that very little airborne sound makes it into the water and vice-versa. However, if a sound is loud enough in the air it can be detected underwater. For example submarine sonars can pick up the sound of the loud turboprop engines used on many maritime patrol aircraft if the aircraft is close enough to the submarine and flying low enough. If MH370 had flown within a few kilometres of an underwater receiving system at low altitude it would probably have been heard, however this is very unlikely to have happened.


      • S.Sadasivan says:

        Thanks. As one who has done some work on exploiting acoustic modality towards aircraft detection, I have great admiration for pioneering work by Australian researchers in this field. The paper “Flight Parameter Estimation using Instantaneous Frequency Measurements from Spatially distributed underwater Acoustic Sensors” by Lo and Ferguson of DSTO is highly relevant here. The authors use acoustic Doppler effect to passively estimate parameters of the overflying turbo-prop, as you have correctly mentioned. May be it is a daunting effort, but pouring over underwater acoustic recordings of the time may unravel – admittedly slim chance – data of relevance to the MH370 mystery.

        Sorry for the bother and best wishes.

  4. Seb says:

    Wonder how the timing of this coincides with the apparent earth tremors that were detected at the time.

    I would also be interested in the raw digital information captured – you apparently have more than one? Would be nice to hear them in stereo (without any lossy sound reprocessing).


  5. Keith says:

    If the most likely location of the sound came from the area around the Maldives, why does the listening station at Chagos (about 300 miles south of Maldives) show no information? Is it not more likely the sound came from an area closer to Australia nearer the two listening stations that actually detected the sound, possibly in the area where the underwater pings were located?

    • michal says:

      Who is in charge of the listening station at Chagos? Does it belong to Deigo Garcia?

      Could you please explain to me the reason you wrote that the most likely location of the sound was from the area around the Maldives.

  6. Pat Janssen says:

    I was very interested in your map showing the probable area of the source.

    Seeing you know the propagation speed and the time of arrival, would it not be relatively easy to add some time arcs to the map to show where the wave would have been at different times?

    These times could then be compared with likely crash times to greatly reduce the area.

    Nice work, thanks,

  7. Pat Janssen says:

    Hi again,
    No need to answer my question, I’ve just done a quick back of the envelope calculation based on 1.5km/s and realised that you would need to know the crash time reasonably accurately to refine the probable area.

    You mention the southern array on Chagos having no signal. Is there not also a northern array there? Did that not have a signal as well?


  8. michal says:

    This is a question about flight 370. Underwater sound recorders off Rottnest Island detected what could be a plane crash, just after 1:30 am UTC on the 8th March. This matched other sound recordings from Cape Leeuwin.

    Could you tell me the location of where the sound came from. It was reported that the location was north-west of Cape Leeuwin. but where exactly?

    • David B. says:

      They didn’t exactly narrow down the location. There was some leeway in the measurements, according to the ATSB report’s section on these recordings, such as a +/- 4s uncertainty in the difference of arrival times of the sounds at the stations. That resulted in an area of uncertainty (yellow narrow polygon) for the signals. In other words, the source of the signals could have been anywhere within that polygon, possibly as far north as what appears to be the Yemen/Oman coast, but most probably near the Maldives.

  9. Michel E. says:

    To investigate a potential acoustic signal in relation with the impact on the sea surface has been a judicious idea .
    As far as I correctly understood how the location area was elaborated from the data of the discovered signals, assuming they come from the same event, it seems that the horizontal variability of the sound speed has not been considered. The consecutive effect increases a lot with the range. Therefore, the polygone of uncertainty is wider in the SW-NE direction (originated from bearing), and in the SE-NW direction (originated from the time diiference) as well. Therefore, the south-east corner of this exented polygon is much closer to Perth; that makes this new area compatible with the ones determinated through the handshakes.
    Besides, the signal, originally very short if generated by an impact, becomes at the reception time spread by the multipathes. Did one try to perform an autocorrelation processing in order to get the range from the knowledge of the environment ?

  10. Alec Duncan says:

    You are correct that horizontal variations in sound speed can cause the acoustic path to deviate from the expected path. We looked into the magnitude of this effect, which was investigated by a PhD student of ours some years ago in another context, but as it amounted to a small fraction of a degree for the direction of interest, did not include it in our calculations. Unfortunately the difference isn’t enough to explain the discrepancy with the satellite handshake data.

    With regard to the suggestion of using the arrival time and (reasonably well) known speed of sound to estimate the range, we have done this based on the earliest and latest times the plane is likely to have hit the water, and the results are consistent with the uncertainty box derived from the arrival time differences at Rottnest and Cape Leeuwin. Unfortunately the speed of sound is large (about 1480 m/s) and the uncertainty in the impact time is also large (more than half an hour), so this doesn’t reduce the size of the uncertainty box by much. It also assumes that the sound we heard was the actual impact. It could also have been the implosion of sinking wreckage, but that would have been expected to occur some time later, which would confound this calculation.


  11. Stefan Simion says:

    Interesting that more data potentially exists which has not been exploited yet (other recorders not yet read out). But this analysis should ideally not be biased by the SATCOM analysis, for instance. Would it be possible to elaborate what the “various factors” are, which make it “unlikely” thet the remaining recorders have interesting data? Is it actually very expensive to get these data? Thanks

  12. Alec Duncan says:

    The main factors are nearby offshore seismic surveys that were going on at the same time – these produce a large amount of low frequency noise in the same frequency band – and poor propagation paths involving considerable distances over relatively shallow water (less than a few hundred metres depth). Also, each recorder is only recording for 1/3 of the time. Schedules vary but they are usually something like 10 minutes recording out of every 30. Costs involved in recovery are of order $40,000 for the more remote loggers, so you have to weigh that up against the likelihood of them having something useful.


  13. Don says:

    I think satellite ‘handshake’ data transmitted could be and is wrong take into account when the first said about the satellite ‘handshake it was said it could have gone north or south wow then it went for only south and now they have searched the spots the satellite ‘handshake’ said it most likely is wow don’t you think oh god they just might not be right I say the best lead might just be the sounds recorded by Curtin University and I wish people would wake up to it here you have a spot where something happen that really fits the puzzle yet your not looking

  14. Don says:

    I think from everything I read you will find that the sounds record do fit in with the satellite ‘handshake’ data transmitted from mh370 go back look at the beginning and look at every thing that’s has been said since I think there is a very strong link with the sound and the satellite ‘handshake’ from mh370 remember they gave a big circle at first with the satellite ‘handshake’ from mh370 if you look at the big circle they first gave you will find they both meet and include they said the plane flow longer than they first thought I strongly believe the sound you have is mh370

Your comment

Your email address will not be published. Required fields are marked *