In the night of 31 May to 1st June 2009, flight AF 447 from Rio to Paris disappeared off the Brazilian coast, without any emergency message, without witnesses and outside radar coverage. As soon as the alert was raised on the morning of 1st June, significant air and naval resources were mobilised to try to find some trace of the airplane and any possible survivors. It was only 5 days later, and during the following days, that bodies and floating debris were found drifting on the surface of the sea, north of the last known position automatically transmitted by the airplane, a little less than 5 minutes before the impact.
On 10 June 2009, a first undersea search operation began. Its aim was to detect the acoustic signals that should be transmitted, for a certified period of 30 days, by beacons attached to the airplane's flight recorders. These searches were undertaken using receivers lent by the US Navy, towed by surface vessels, and a receiver on board a French Navy submarine. The acoustic searches were brought to an end on 10 July 2009.
This operation was followed from 27 July to 17 August 2009 by a new attempt to localise the wreckage with the aid of sonar on board an IFREMER vessel. This operation was also unsuccessful in its attempt to find the wreckage.
Thus, in August 2009, the BEA defined a new strategy to search for the wreckage in a particularly difficult environment:
A vast search zone (17,000 km2) corresponding to the area of the circle into which the airplane crashed into the sea;
Sea depths that go down as far as 4,300 metres;
A very rough sea bed;
A region in which the currents are not well known.
In order to be able to define the impact site, the BEA then launched a study to model the sea currents so as to estimate the drift of the recovered debris. This work was entrusted to a group of specialists from eleven oceanographic institutes and organisations and resulted in the definition of a reduced area of about 2,000 km2 in which the wreckage should, with a high degree of probability, be located. The third search campaign was thus launched on 29 March 2010 on the basis of this hypothesis.
The hope of quickly localising the wreckage led the BEA to mobilise versatile and complementary naval resources that made possible both localisation of the wreckage and the immediate recovery of the recorders and, if appropriate, of elements of the wreckage useful to the investigation.
This campaign took place in two stages, in April and then May 2010. It was based on two vessels, one equipped with Remus 6000 autonomous underwater vehicles (AUV) as well as means for observation and lifting, the other with a towed sonar (the latter did not participate in the second stage).
During these two stages an area of over 6,000 km2 was covered: the initial zone of 2,000 km2 was extended to the adjacent zones. The operation came to an end at the end of May 2010 without having detected the wreckage of the airplane.
The lack of success of the sea searches undertaken between June 2009 and May 2010 meant that a review was required in order to answer the following questions:
Did the resources used make it possible to localise the wreckage?
Were the zones explored the ones that might contain the wreckage?
Work undertaken by the BEA with the support of Metron Inc made it possible to exploit the results of the searches undertaken up to now. It was possible to deduce that the zones that had been searched previously using sonar underwater imagery did not need to be covered again.
However, the wreckage may be located in the other zones in the 40 nautical mile (74 km) circle centred on the last known position that have not yet been explored by this type of equipment. This is why the new campaign is based on the strategy of systematically searching all of the zones that have not yet been explored by means of sonar underwater imagery. It will begin in the 20 nautical mile (37 km) circle centred on the last known position.
