Investigation | Are we looking for missing flight MH370 in the right place?

Jonathan Durgadoo, Siren Ruehs, Arne Biastoch

The search for missing flight MH370 has so far been focused on a 120,000 square kilometre region south west of Perth, Australia. But now there are reasons to believe that the search area could be too far south.


On March 8th 2014, 239 lives ceased, and are still unaccounted for. Flight MH370, the now infamous Malaysian Airlines Boeing 777-200ER, departed from Kuala Lumpur for Beijing. Soon after take off, half way between Vietnam and Malaysia at the mouth of the Gulf of Thailand, the aircraft was recorded making a sharp turn west. MH370 was last spotted on radar about an hour and half into the flight over the Andaman Sea, south west of Phuket. Then it disappeared.

After disappearing from radar screens, the plane remained airborne for several hours. Evidence for that comes from the hourly automatic satellite log-on communications, so-called “handshakes”. The 7th and last of these hourly handshakes occurred approximately 7.5 hours after take off.

Australian-led search authorities initiated an extensive search campaign shortly after the disappearance. A priority search area was defined in the southeast Indian Ocean around “Arc 7”, the possible location of the aircraft during the 7th handshake. In defining the search area, the authorities also considered the maximum flight range given the amount of fuel carried by the aircraft.

Fast-forward 16 months to July 29th 2015. After months of fruitless search, a 2 meter long flaperon (part of an aircraft’s wing that helps stabilise the plane during take off and landing) beaches at Saint-André, on La Réunion Island, thousands of kilometres away from the search region. Within weeks the authorities confirmed that the flaperon indeed belonged to MH370.

That latest finding naturally begged several questions and fuelled several more conspiracy theories. From an oceanographic perspective, the question was simple, yet difficult to answer: could we track the flaperon back in time to establish where the plane had crashed, and if so, would that position coincide with the priority search area?

“Ocean models describe the motion of water using well-understood mathematical equations”

As an oceanographer this question appealed to my curiosity, and to that of my team. Since the flaperon beached, it must have floated and drifted, that is, moved with the currents. Drift of an object at the surface of the ocean can be affected by 3 factors: surface ocean currents, direct sail effect by the wind, and wave-induced drift. Under the assumption that the flaperon drifted horizontally and mostly submerged in the upper metre of the ocean, direct wind sail effect is negligible. This left us with the need to obtain accurate descriptions of surface ocean currents and that of wave-induced drift.

From the outset, it was clear that we would need the most up-to-date dataset for the period March 2014 – July 2015, and that such a dataset needed to be coherent, containing no gaps in space or time. Since no ocean observations satisfy all of these requirements, we used an ocean model to perform the task.

Ocean models describe the motion of water using well-understood mathematical equations, and are performed using state-of-the-art super-computers. The model dataset that we used also makes use of available ocean observations – from satellites and buoys – in order to provide the best possible coherent simulation of ocean currents for the desired time period. This model ran on a supercomputer at Mercator Ocean in Toulouse, France. The description of wave-induced drift was obtained from the European Centre for Medium-Range Weather Forecasts in Reading, UK.

So, the idea is that we could use an ocean model like this to track the flaperon back in time to establish the flight’s crash location. But the ocean is a chaotic place; it makes no sense to simulate the path of a single “virtual flaperon” backward in time. Therefore a “strength in numbers” strategy is what we used when we placed close to 5 million virtual model flaperons around La Réunion Island during the model month of July 2015. Using the modeled description of the upper ocean currents and of wave-induced drift, these virtual flaperons were backtracked until March 8th 2014 – the date the doomed flight crashed. Daily geographic positions for each virtual flaperon were stored, resulting in approximately five million independent possible trajectories.

“based on our analysis, the chance that the flaperon started its journey from the priority search area is less than 1.3 %”

Now, let us assume the satellite handshake analysis to be credible. This means that the flaperon would originate somewhere close to Arc 7 on March 8th 2014 (see article image). This means that we can refine our analysis by imposing a condition. Keep in mind that we have virtual flaperons starting at La Réunion in July 2015 and tracked backward in time until March 2014. If we only consider those virtual flaperons whose position on March 8th 2014 were, say, within 500 kilometres of Arc 7, then we reduce the number of possible trajectories to around 800 000. The positions of these virtual flaperons on March 8th 2014 are subsequently mapped giving an indication of the most probable locations of their origins (see article image).

While it is impossible to pin-point an exact location, we found that the origin of the flaperon is likely to be to the west rather than southwest of Australia. More importantly, based on our analysis, the chance that the flaperon started its journey from the priority search area is less than 1.3 %. This naturally raises questions. Should the search area be refocused elsewhere? Are the model assumptions that we made correct?

With just a single piece of debris, these questions are hard to answer. But since July 2015, there have been numerous other pieces of debris found along the shores of several east African nations, namely Mozambique, Madagascar, Mauritius, Tanzania, and South Africa. Several of these have been confirmed to be very likely components of the demised aircraft. Our next step is to refine our results by considering these other debris. Back tracking multiple pieces of debris using our ocean model method will further allow us to home in on the likely crash point of the plane. This is what we are now focused on.


Dr. Jonathan Durgadoo is an oceanographer interested in ocean circulation and the role of the ocean in climate. He is a post-doctoral fellow at the GEOMAR Helmholtz Centre for Ocean Research Kiel, Germany.