It has been three years and seven months since flight MH370 vanished in the heart of a quiet night above the South China Sea. The Boeing 777 had been travelling northeast from Kuala Lumpur to Beijing, where it was scheduled to land at 6.30am on March 8, when the co-pilot signed off from Malaysian airspace with the now infamous words “Good night, Malaysia 370”.

That was the last ever heard from the 239 people on board, 153 of whom were Chinese, and the last time the whereabouts of the doomed flight can be calculated with any real certainty. After 10 days of frenzied media speculation, in which the Malaysian authorities’ complete (and embarrassing) lack of knowledge of the flight’s location was broadcast across the world, Australia took the lead role in the search. That move was heralded at the time as “the pros” taking over – a confidence that in hindsight can be seen as completely misplaced given Australia’s aviation watchdog closed its investigation last week, not an inch closer to the truth.

But in the swirling confusion that followed the disappearance, hopes were high that those “pros” could make a breakthrough that would put an end to the (mostly flawed) theories that had begun to circulate in the international media – was it a hijack (where could it land without detection?), was it terrorism (why did no group claim responsibility?), was it pilot suicide (why no note and why such a complicated route?).

Indeed, what the “pros” did next was remarkably successful in helping the authorities regain control of the media narrative, in helping to reassure a worried public that, even if the plane’s exact location was not known, everything was nevertheless under control.

Seemingly against all odds and logic, those in charge declared the plane’s final resting place to be somewhere in the Southern Indian Ocean – thousands of kilometres in the opposite direction from where it was heading.

WATCH: MH370 ‘descended rapidly’

So certain were they in this counter intuitive conclusion that one of the largest marine searches ever undertaken took place. That search continued for 1,046 days, cost A$198 million (HK$1.2 billion) and mapped 710,000 sq km of sea floor (the largest ever single hydrographical survey) – 120,000 sq km of it in high resolution. Yet by the time the search was suspended in January 2017, those – much heralded – efforts had come to little. Four shipwrecks had been discovered, but no trace of MH370 was ever found.

“It is almost inconceivable and certainly societally unacceptable in the modern aviation era … for a large commercial aircraft to be missing and for the world not to know with certainty what became of the aircraft and those on board,” the Australian Transport Safety Bureau (ATSB) said in its final report, published on October 3.

Remembering the boldly confident statements that punctuated the search process, this admission of blatant defeat may come as a surprise to the public.

But a closer look at the operation, its initial legitimacy, and the vague premises on which the search area was based, shows this mission had every reason to fail, from day one.

That raises uncomfortable questions: to what extent did the authorities ever believe the search would succeed? Did they believe their own bluster and, if not, why not? Was the entire process an elaborate charade, aimed more at silencing a braying media than it was at finding the truth? An attempt to delay that admission of defeat until a time when pictures of the weeping relatives of 239 lost souls were not leading daily news bulletins across the world?

KNOWN UNKNOWNS

From the beginning, various authorities involved in the search did their best to give the impression that, amid the countless unknowns surrounding the flight’s fate, there were nevertheless at least some things they did know.

A little more than two weeks after the plane disappeared, Malaysian Prime Minister Najib Razak, in a news flash on March 24, announced with seeming certainty something the Australians were unable to prove in more than three years of searching for the plane: “[MH370’s] last position was in the middle of the Indian Ocean, west of Perth”.

Next of kin were informed by text message: “Malaysia Airlines deeply regrets that we have to assume beyond any reasonable doubt that MH370 has been lost and that none of those on board survived.”

This was a new hammer blow for the relatives, who were being asked to accept their loved ones should be declared lost forever – without a single piece of debris having been found.

WIZARDS OF OZ

The truth is, in late March 2014, there was enormous confusion over the case. Malaysia had lost control of events to such an extent that the international media gathered in Kuala Lumpur were having a field day. It looked like anyone could do a better job than the Malaysians. Since Australia had already taken on the mantle of “head of search and rescue operations” down under, it somehow seemed natural and reassuring that the Australians took the lead of the overall search operation now that it was moving south.

Certainly, this was the view of former US naval officer Stephen Ganyard, who told Good Morning America, “The good news, here, is that we have the Australians now in charge of this investigation. We’ve seen a lot of inconsistency out of the Malaysian authorities all week. But now we have the real pros on the scene.”

Not everyone shared his optimism. That Australia helped coordinate the surface search was one thing. But that it should be in charge of the underwater phase puzzled many, including American aviation journalist and author of The Crash Detectives, Christine Negroni.

The Australian Transport Safety Bureau (ATSB) was still tainted by its shoddy investigation into another aviation disaster in which a Pel-Air WestWind air ambulance was forced to ditch near Norfolk Island in 2009, having found itself unable to land in bad weather and with insufficient fuel to divert. The two and a half year investigation by the ATSB prompted a national scandal when a television documentary aired allegations of misconduct by the ATSB. A subsequent Senate inquiry found the ATSB’s accident report was deeply flawed and had unfairly blamed the pilot.

Claim of new evidence in search for MH370

“It seemed to me that the Australians including the [head of the ATSB] Martin Dolan were eager to become the heroes in solving the world’s most riveting air mystery. In an interview in June 2014, Dolan told me enthusiastically that coordinating the search was ‘the challenge of a career’,” recalled Negroni.

Beset by public humiliation, did the ATSB see the MH370 mission as an opportunity for redemption? To cleanse its name and bury past shame? Possibly.

Whether it was up to the task is another matter. The man appointed to head the initial search effort was the much decorated Angus Houston, former Chief of the Australian Defence Force, who had a great reputation in the military but no previous experience of civil aviation accidents nor of underwater searches.

Indeed, while Australia was willing, it was not Malaysia’s first choice. That distinction went to the French air accident investigation bureau for its role in investigating the fate of AF447 – a case with striking similarities to that of MH370. The Air France A330 jet had crashed during the night in the middle of the Atlantic Ocean on its way from Rio de Janeiro to Paris on June 1, 2009, in 3,980 metres of water. Its hull and black boxes were located and recovered in less than two years after a search that cost just US$40 million (including the recovery) – a quarter of that spent on looking for MH370.

The Malaysian government approached the French team but Jean-Paul Troadec, the bureau’s former director who lead the search for AF447, declined.

WATCH: Australia says MH370 search could resume in future

Legally, Australia was under no obligation to take charge of the operation, let alone at its own expense (initially it had shared the cost with Malaysia). Almost 3,000km from the Australian coast, the search area is in international waters. Therefore, any search and rescue responsibilities Australia may have had would have been limited to saving lives – not locating wrecks. Yet, with unclear motives and poor credentials, Australia volunteered. Why?

SEARCH FOR A SEARCH SITE

Of course, the Australians can’t be blamed for not finding the plane if the plane was simply not there in the first place. But neither can Malaysia be blamed for identifying the wrong search site. During his March 24 television address, Najib clearly said he had been “briefed by representatives from the UK Air Accidents Investigation Branch. They informed me that Inmarsat [a British satellite communications company] has been performing further calculations on the data.”

In other words, the Malaysian prime minister was just repeating what the British had told him.

And what the Brits had told him was based on some rather flimsy evidence. They had established the crash location solely on the basis of “handshake pings” – the tiny, invisible signals that bounced between MH370, a ground station in Perth, Australia, and an Inmarsat satellite 36,000km above earth. There were just seven such “pings” on the night of MH370’s disappearance yet this was apparently enough for mathematical extrapolations to establish “beyond reasonable doubt” the fate of the plane and its 239 passengers.

Even if we put full faith in those extrapolations, in trying to deduce the flight route, the scientists had to make substantial assumptions about the speed and altitude of the plane, assumptions for which there is no proof. The slightest inaccuracy in any of these assumptions would change the final crash site drastically. With so many vagaries, how can anyone be sure of anything?

To make matters worse, Inmarsat withheld its full set of raw data for years from independent experts, inexplicably taking an attitude that was both secretive and hostile to the media rather than proudly explaining how it had helped solve the mystery.

About three years ago, I contacted British astrophysicist Duncan Steel to share my concerns about Inmarsat’s mind-boggling calculations. Steel is one of the leading scientists of the self-named Independent Group that worked separately but in cooperation with the Australian team. Instead of reassuring me, Steel highlighted further complications, including the inclination of the Inmarsat satellite, which had been greater than normal at the time. Another complication was that Inmarsat’s software had been designed when its ground bases were only in the northern hemisphere, and had not been fully updated to account for the ground base in Perth. In three years of investigating, I have not come across a single person – even in the satellite industry – who felt confident in the extrapolations based on the satellite’s “handshake pings”.

WATCH: Don’t abandon search, MH370 kin plead

The final report into the search succinctly sums it up: “The data available was very limited. The type of data and the scientific methods used for its analysis were never intended to be used to track an aircraft or pinpoint its final location.”

So, to sum up, the search was based on a small sample and an untested method – something that would alarm any scientist. The biggest challenge of this search was that it was impossibly vague.

FOLLOW THE DEBRIS

The only tangible thing that could have lent credence to the projections would have been to find debris. Yet at the time the search was launched no debris whatsoever had been found. Indeed, to this day, only a few pieces of debris that can categorically be linked to MH370 have been found – despite numerous claims that other debris “almost certainly” from MH370 has appeared on the southeastern coast of Africa, a region where several plane crashes have occurred in the last decades, involving Boeing planes on several instances.

Launching the search then contradicted a maxim known to everyone involved in investigating plane crashes over water: “Find debris first, work out the crash zone second, look underwater third.”

“First and foremost, you must find pieces of debris and formally identify them as coming from the plane. The ocean is a dustbin – and from a long way out you arrive in the whirling currents of the Indian Ocean Gyre,” said Troadec, who led the search for AF447.

Troadec said it would be useless to begin underwater searches with such uncertainty about the crash position. “Once you have debris, experts can work out the point of crash by drift calculations based on winds and currents. But the margin of uncertainty rises quickly as the days go by,” explained Troadec. Time is of the essence.

WATCH: Wreckage found on Reunion Island may be from MH370

With every passing minute, any debris is breaking apart, becoming water logged, sinking and dispersing, making the task of identifying the crash zone ever more difficult.

Planes are made of millions of components, all built as light as possible, meaning most parts are buoyant. Even when a plane stays mostly in one piece after impact – as AF447 did – thousands of pieces break away and float for a while. About 90 per cent of AF447 sank in one piece, with most passengers still belted to their seats, even so, several thousand pieces of debris and several bodies were collected from the surface.

“You find everything in the sea. The sea doesn’t hold on to things. And if debris is found, the aircraft will be found,” a British naval officer told me shortly after MH370 disappeared. He mentioned French solo sailor Eric Tabarly, whose body was found in 1998 in the Irish Sea by fishermen 40 days after he had fallen overboard.

The absence of debris was not due to a lack of investigation. In May 2014, a meeting between Australia, Malaysia and China in Canberra heard that Australia had carried out 334 air patrols, involving 3,137 hours of air reconnaissance. Its search had involved 10 civilian aircraft, 19 military aircraft and 14 ships. Chinese efforts had involved 21 satellites, 18 ships (including eight equipped with helicopters) and five aircraft, covering an area of 1.5 million sq km. China had also asked 88 Chinese-registered vessels inside the zone (68 merchant ships and 20 fishing vessels) to help. Despite all this, as well as dozens of beach cleanings along the western and southern coasts of Australia, no debris that could be conclusively linked to MH370, its fuselage, its cargo or its passengers was found.

WATCH: Chinese MH370 relatives still suffering

GOING UNDER

This complete absence of debris did not dishearten the Australians, nor did it lead anyone to double check the assumptions on which the search area was based. Everyone was happy to pretend this mission impossible was feasible and under control. And so the underwater search launched with great fanfare and hope.

The imposing Australian Border Force Ocean Shield cast off from Perth, with its gleaming red hull 110 metres long, its crane, its helicopter landing pad and all the acoustic equipment needed to listen for the “pings” emitted from MH370’s black box beacons. It had just days before the beacons would stop transmitting, leaving little time to make good use of the cutting edge equipment on loan from the US navy.

A Towed Ping Locator (TPL) resembles a small, bright, yellow stingray. Once in the water, it can detect black box signals from up to 2km away. But the black box beacons are powered by a battery that lasts about a month, after being triggered by contact with water. The beacon’s job is to transmit an inaudible ultrasonic signal, saying “Find me. I am here”. The ping locator’s job is to pick this up and transmit it to the surface crew.

In all likelihood, only one of MH370’s two black boxes was actually emitting (the one attached to the cockpit voice recorder). As revealed in the interim report Factual Information, published in March 2015, the flight data recorder’s beacon had expired in December 2012. This fact was not admitted at the time of the crash and is not even mentioned in the latest report. Such an admission would have embarrassed Malaysia Airlines for its appalling maintenance standards and cut by half the success chances of the Australian search. Why share such bad news when the motto was “hope”?

So it was with near zero chance of success that the Ocean Shield lowered the ping locator into the water.

Yet once again, this did not prevent authorities from providing an illusion of success, helping to satiate a public hungry for progress. Within minutes of entering the water, the ping locator appeared to strike gold, detecting a ping. Almost simultaneously, a Chinese ship hundreds of kilometres away also detected pings and soon after that another TPL picked up yet more pings. A press release was issued on April 5 saying “signals consistent with those emitted by aircraft black boxes” had been found.

This was cause for false optimism. Black box pings have an unmistakable identity that derives from a combination of two features: their frequency of 37.5 kHz and their intermittence, a one second interval between each signal. None of the pings matched these criteria. But the show was too good to be spoiled. Television stations were breaking audience records.

“Call it a triumph of science, or incredible luck, but on the very first path, the Ocean Shield [towing the TPL …] detected a steady series of pings,” CNN commented on April 8 – one month after the plane was lost, when the beacons’ battery lives would already have been in overtime.

CNN, intoxicated by raptures of the deep, kept serving up its Australian pings 24/7, but scientists familiar with underwater searches were growing increasingly alarmed. “I went on CNN about 15 times to try and make people understand there was absolutely no chance the pings detected were from MH370,” said Paul-Henri Nargeolet, a globally respected wreck hunter. A former naval officer and clearance diver, he has led countless underwater assignments including six expeditions to the Titanic since 1986. He was also part of the search for AF447.

“I know the US Navy people who were on board [Ocean Shield], they would never make such elementary mistakes. A 33 kHz ping, or a 35 kHz ping, cannot become or be treated as ‘similar’ to a 37.5 kHz ping, either because of the pressure or because of failing batteries as I heard it explained. That’s sheer nonsense,” said Nargeolet.

William Meacham, a former archaeology professor at the University of Hong Kong, tried to warn CNN it was making a big mistake in suggesting the pings could come from the MH370 black boxes. Pointing out that pingers were also used by scientists to tag marine creatures, he drew up a list of those that may well have been inside the search zone: 86 loggerhead sea turtles, 30 flat-backed turtles, 30 hawksbill turtles, 14 green turtles, seven humpbacked whales and five dugongs – not to mention a great white shark carrying a 36 kHz pinger, which had crossed the Indian Ocean from South Africa to Australia’s west coast. Was the Australians’ greatest hope nothing more than a great white?

As irrelevant as these detections were, they created so much excitement in the news that the Australians decided to launch the autonomous submarine Bluefin-21 to feed the media appetite for anything ping-related.

WATCH: Australian deputy PM on debris found in Indian Ocean

“There was one signal which has been analysed very closely, which was a very strong signal and it had all the characteristics of being from a man-made device and the characteristics of the transmission were very, very similar to those … from an emergency locator beacon; our experts have established a datum on the ocean floor – probably the most likely place where you might find wreckage of the aircraft or a black box,” Angus Houston told a Chinese journalist from Xinhua, adding he was “very hopeful that we will find something”.

When, at the end of May 2014, Michael Dean, the US Navy’s deputy director of Ocean Engineering, said on CNN that everyone now agreed the pings had not come from the black boxes, a US Navy spokesman appeared on the same channel a few hours later to dismiss the comments as “premature and speculative”.

It was as if confusion about the true nature of all these erroneous pings was supposed to last a little bit longer.

One question lingers: was this incredible fiasco due to incompetence on the Australians’ behalf or was it an orchestrated show primarily aimed at satiating a media hungry for MH370-related news?

MISLEADING LEADS

The absence of any real leads did not stop authorities from showering the media with hopeful announcements. And when it came to such announcements, the Australian prime minister at the time, Tony Abbott, was in a class of his own.

Abbott announced “new and credible information”, the “best leads so far”, felt “very confident” … “The best brains in the world are applying themselves to this task. All of the technological mastery that we have is being applied and brought to bear here, so if this mystery is solvable, we will solve it,” he boasted on March 31, while visiting the Pearce Air Base north of Perth, the air search headquarters. On April 11, during a trip to China, he went one step further. He was “very confident the signals we’re detecting are from the black box from MH370. We know the position of the black box flight recorder to within some kilometres”.

Naturally, this unequivocal claim came as a shock to the families. Unfortunately, it was just a false statement that he was later challenged to explain in parliament.

Another questionable announcement came via Angus Houston on April 14, when “Ocean Shield detected an oil slick … approximately 5,500 metres down wind and down water from the vicinity of the detections picked up by the TPL on Ocean Shield”.

Was this supposed to be an indication of the MH370 crash site? The official narrative was that MH370 crashed having run out of fuel. And even with some remaining fuel in its tanks, did this experienced air force officer really think for a second a slick could still be found 38 days later? The plain answer is no. So, why pass on such information, as useless as it was misleading, while adding, to faint caution, that the liquid “still has to be analysed”?

Meanwhile, the Malaysian transport minister, Hishammuddin Hussein, was struggling to keep up with the Australians’ pace of mock good news. Seemingly out of the blue, he said on April 19 the next 48 hours were going to be “crucial”. Then, nothing more: no pings, no black boxes and no MH370 wreck. Nothing crucial, then.

On April 28, 2014, three weeks after the last pings were picked up, the Australian prime minister threw in the towel and brought the ping hunt to an end, calling the sea search “the most difficult in human history”.

“We focused on the best leads we had,” he assured everyone, without detailing exactly what “leads” he had in mind. Abbot said it was now “highly improbable that we will find the slightest debris on the surface … 52 days after the crash, most of the debris would have become waterlogged and sunk.”

It was yet another statement that would be thrown overboard when, 500 days later rather than 52, the first significant piece of potential debris washed ashore on Reunion Island, a French territory. Abbott then declared, without waiting for confirmation that it was a piece of MH370, that the find was “very consistent with the search pattern we’ve been using for the last few months”.

The moment the flaperon was found on Reunion Island, several oceanographic institutes produced studies pointing to very different potential crash zones. In November 2014, the ATSB crash investigator Peter Foley himself had predicted that “something is going to wash up somewhere on the beach, most probably in Sumatra”. That’s a long way from the African coast...

Australia, then, would seem to run Malaysia pretty close when it comes to the art of bungling a search operation, and in terms of providing deliberate or accidental misinformation.

The fiasco surrounding the pings, the sudden and unexplained changes in search zones, the opportunistic switches in the official line, along with a string of totally unfounded declarations, were surely not what was expected from the “real pros”.

Nevertheless, the ATSB stayed in charge until January 2017, continuing to provide the public and the families with fleeting snippets of good news – snippets that would slowly sink without trace as soon as they were subjected to the slightest scrutiny.

The natural conclusion of this sad story, if we are to follow the official line, is that a plane as big as a Boeing 777, loaded with electronics and equipped with several redundant communications systems – not to mention the hundreds of mobile phones of its passengers – can become perfectly stealthy in a few seconds, in one of the most closely monitored regions of the planet.

MH370 managed to do what decades and billions in research have not yet achieved for the most sophisticated military plane. Each and every one of the 10 million passengers who board a plane every day must hope their plane won’t be up to the same trick.

Florence de Changy has published her investigation into flight MH370 in French and Chinese. She is the Hong Kong correspondent for Le Monde and French National Radio