As usual early in air accident inquiries, there is growing confusion over what happened in what sequence to Air France flight AF447 early on 1 June over the mid Atlantic.

It all seemed a lot simpler 24 hours ago. The A330-200 had broken apart in mid air after flying through wild weather. There is widespread wreckage on the ocean. A series of automated status messages sent to the operations centre in Paris revealed an unprecedented series of electrical failures as well as a pressurisation failure.

But it is less clear by the hour as to what happened in what order, and if the reports in the French media are a guide, different people in the airline are saying slightly different things.

Even the official statements lack total consistency.

Initially Air France confirmed that the early signal was one of a short circuit followed by a depressurisation followed by four minutes of a flurry of messages detailing an unprecedented series of electrical faults.

However in the last day not even Paul Louis Arslanian, the head of France’s air accident investigation agency the BEA is sure in public interviews as to whether the jet broke up in flight or on impact with the sea.

He now says there is not enough information to establish which came first, or even the exact time of the accident.

About the only shred of information that seems fixed in sequence is that the last automated report from the Airbus was a vertical cabin speed alert.

Some have interpreted this as meaning the cabin speed was positive or climbing not falling, and others that it reflects the falling air pressure, which is most likely wrong, but all bets ought to be off until Air France reveals exactly what it knows, publicly.

The airline did reveal that the electrical faults messages included malfunctions in the ADIRU or air data and inertial reference units which provide speed, attitude and other information to the primary flight control computer.

This raised similarities with the ADIRU problems encountered by Qantas A330-300s, even though the units in the Qantas jet that was most affected, which was flying QF72 between Singapore and Perth on 7 October when it suffered a serious control upset and made an emergency landing at Learmonth, were not of the same design or manufacture as those in the Air France jet.

Other elements in the cloud of confused speculation include a ground accident in which the tail area of the crashed jet tangled with that of another Air France Airbus at Paris Charles de Gaulle airport in 2006.

It could be an important clue, as in ‘could be’ not ‘must be’.

The Paris public prosecutor’s office has announced that it will move quickly to open ‘une information judiciaire‘ or preliminary investigation.

The only possible value in this would be if Air France, for the sake of clarity, was to table the full details of the last voice communication from the cockpit, about 30 minutes before the crash, and the four minutes or so of automated status messages.

Otherwise this investigation will most likely be rapidly shunted to one side pending the official accident inquiry, which is required by law and international protocol to make a preliminary report within 30 days of an accident, meaning the end of this month, as is the ATSB in Australia in its investigations.

Among the reports that lack credibility are those quoting the pilots expressing concern that they were flying directly into a black storm cloud.

It was night. Everything was black.