Classification: Documented

Flight MH370: 239 people, seven hours of signals and an ocean without an answer

The Boeing 777 vanished on 8 March 2014 with 239 people aboard. Radar, satellite records and identified debris steered the investigation toward the southern Indian Ocean, but have not revealed an impact site or cause. This dossier follows the evidence, the searches and the operation concluded in 2026.

Sepang, Selangor, Malásia · See stories from this place and nearby

Before the map, 239 people

In the early hours of 8 March 2014, Malaysia Airlines flight MH370 took off from Kuala Lumpur International Airport in Sepang for Beijing. The Boeing 777-200ER, registration 9M-MRO, carried 227 passengers and 12 crew. The manifest stood for families, colleagues and lives across many countries; reducing the case to a contest of theories erases the people on board. Malaysia is eight hours ahead of UTC: some reports date the occurrence 7 March UTC, without referring to a different flight.

Nocturna’s marker identifies the confirmed departure airport, not an impact site, which remains unknown. The planned route crossed the South China Sea toward China. The operation initially appeared normal. The loss was understood in layers: first the expected identification vanished from civilian control; only later did combined radar and automated messages indicate that the aircraft had likely kept flying.

The handoff between two controllers

At 01:19:30 Malaysian time, the last recorded voice transmission acknowledged an instruction to contact Ho Chi Minh control. Malaysia’s report identifies the voice as the captain’s and detected no anxiety in the exchange. At 01:20:31, the aircraft passed waypoint IGARI. Soon afterward, its transponder ceased to provide the identification and position expected by secondary radar; MH370 did not establish normal contact with Vietnamese control. That is a loss of contact to civilian aviation, not evidence that the aircraft crashed at that minute.

The 2018 report describes delays and coordination failures between control centres as the aircraft failed to respond. That interval affected the initial response, but does not itself explain why the plane changed direction. The crucial distinction is between what controllers could see that night and what investigators reconstructed months later from several kinds of record.

The turn captured by radar

Radar records, including military surveillance, showed a track that diverged from the Beijing route, crossed the Malay Peninsula again and headed toward the Strait of Malacca and northern Sumatra. The ATSB locates the last positively fixed surveillance position near Sumatra’s northern tip. Radar did not then provide a continuous track to the aircraft’s final destination. A recorded turn is data; who commanded it and why remain questions.

Malaysia’s report found it hard to attribute the sequence of manoeuvres to a particular system failure and considered manipulation of controls more likely. This identifies neither a person, an intention nor a single mechanism. Statements that “radar proved the final destination” are wrong. Radar constrains part of the path; the much longer final stretch was inferred from another type of trace.

Signals that were not GPS coordinates

After normal communications ceased, the aircraft’s satellite terminal still exchanged automatic signals with a ground station. These handshakes were neither passenger calls nor messages stating a position. Analysts used mainly burst timing offset (BTO), restricting distance from the satellite, and burst frequency offset (BFO), helping assess motion and possible tracks. Each timing measurement gives an arc of possible positions, not a point. Aircraft performance, fuel consumption and satellite movement also enter the calculations.

The last logged exchange was at 00:19 UTC on 8 March, or 08:19 in Malaysia, roughly seven and a half hours after takeoff. The analysis favoured a course toward the southern Indian Ocean, and the seventh arc became the search reference. These signals conceal no GPS map. Location along the arc, altitude, changes of heading and the impact time depend on modelling, so search areas can change as evidence is reassessed.

Fragments that reached distant shores

On 29 July 2015, a flaperon was found on the French island of Réunion. French investigators tied it to aircraft 9M-MRO using internal component numbers and manufacturing records; confirmation was announced in September. Other components examined in Mauritius, Tanzania, Mozambique and elsewhere received different identification grades: confirmed, highly likely, likely or unidentifiable. It is inaccurate to call every collected object a proven MH370 part.

Indian Ocean drift brought material to coasts far from its origin. CSIRO researchers simulated currents and winds against dates and places where debris was, or was not, found. Working backward narrows possible regions but does not yield one coordinate: fragments float differently and can go unseen for months. Confirmed finds establish that the aircraft ended in the ocean; they do not replace the main debris field or flight recorders.

Searching an ocean floor

The initial search focused on the South China Sea and Strait of Malacca before combined radar and satellite interpretation shifted it to the southern Indian Ocean. Under Australian coordination, the seabed effort reached remote water with irregular topography and depths of up to about 6,000 metres. Teams first had to map the bottom, then scan sonar strips and inspect promising targets. Mapping bathymetry and searching for debris with high-resolution sonar are different tasks.

By the January 2017 suspension, the ATSB had covered more than 120,000 km² with sonar; 82 especially promising targets were investigated and ruled out as MH370. A further Ocean Infinity effort in 2018 brought the total coverage to close to 200,000 km², according to the ATSB, without finding the main wreckage. A “searched” area describes the coverage and quality achieved by a campaign. It does not mean that every point in the ocean was seen or exclude impact hypotheses outside it.

The final stretch: simulation versus observation

Aircraft performance and fuel models suggest the prolonged flight ended near fuel exhaustion. The final satellite exchanges, especially a log-on attempt, helped shape descent scenarios and guide the seventh-arc search strip. Experts also examined wing-part damage. Malaysia’s report judged an outboard flap probably retracted when it separated; that weighs against some controlled-ditching stories, but does not describe every final second.

No investigator witnessed the impact. Without flight-data and cockpit-voice recorders, the main wreckage and a physical sequence of breakup, one cannot confidently select among all simulated last movements. The absence of a detected distress signal or beacon also does not prove everyone aboard was incapacitated: the report notes that the flight-data-recorder locator beacon battery had expired. That maintenance issue matters to a search; it is not a demonstrated cause of disappearance.

What the 2018 report did and did not find

With the evidence available, Malaysia’s team judged the communications loss before the diversion more consistent with manual deactivation, interrupted electrical power or radios not being used than with a specific mechanical failure mode. It likewise considered control inputs the more likely source of the changes in track. At the same time, it said it could not definitively exclude a system failure or third-party intervention. Its formal conclusion was that the actual cause could not be determined.

This wording blocks an unwarranted leap from probability to accusation. A track compatible with human input does not establish who acted, whether coercion occurred, the cabin’s condition or a motive. Hijacking, deliberate action, depressurisation and technical emergencies carry different strengths and difficulties; none was settled by the missing recorders. This dossier does not turn identifiable people into culprits without evidence.

The 2025-2026 return

On 25 March 2025, Malaysia contracted Ocean Infinity to search a new estimated 15,000 km² area in the southern Indian Ocean under a “no find, no fee” arrangement. According to the official 8 March 2026 update to families, operations after signing came in two phases: 25-28 March 2025 and 31 December 2025 to 23 January 2026. There were 28 operational days, surveying about 7,571 km² inside the defined area. Weather and sea conditions interrupted work; by that update, nothing had confirmed the aircraft’s location.

On the same date Ocean Infinity reported 151 days at sea and more than 140,000 km² mapped since 2018. Those are company-wide cumulative MH370-search figures, not the area of the latest campaign and not figures to add directly to the Malaysian 7,571 km². The company left the area on 23 January 2026. On 29 June, Malaysia’s transport minister announced an extension of the agreement from 1 July 2026 to 30 June 2027, covering about 7,428.54 km² still to be surveyed. He indicated that search assets might return between November 2026 and April 2027, subject to operational conditions. Extending the contract does not mean a new phase has already begun or produced results.

What another discovery could change

Established: 239 people left Sepang; the aircraft diverged from its route, exchanged automatic satellite signals for hours and left confirmed parts in the Indian Ocean. Unknown: the impact point, the sequence in the cabin, the cause and the location of the main wreckage. As of September 2026, the latest completed-search result cited here is from 8 March; the 29 June contract extension allows for a future phase. The consulted sources report no confirmed discovery of the main wreckage.

Start with Malaysia’s report: its timeline and its conclusion differ from accusatory headlines. Then compare ATSB search maps, CSIRO drift work and the 2026 update. Ask whether each item is an observed position, inferred distance, identified component, calculated scenario or sonar result. Finding the fuselage and recorders could turn this mosaic into a causal investigation. For the families, the missing answer is not entertainment: it is a prolonged loss that maps have not yet closed.

Open in Google Maps