The disappearance of Malaysian Airlines Flight 370 (MH 370) carries many interesting naval and defense lessons. The first and most important is that the widely held belief that nearly anything on the surface of the Earth is more or less visible at all times, thanks to satellites and other sensors, is an illusion. Most satellites provide fleeting glimpses of the Earth. The only ones that stare continuously at the surface, from geostationary orbits, are so far away that their sensors are incapable of tracking most relatively small moving objects.
Many years ago the U.S. Navy considered building a constellation of radar satellites specifically to detect and track Soviet bombers that might attack its carriers. Because satellites low enough to use radars to detect and track aircraft do not hang over particular spots on the planet, the project required enough satellites so that at least one was always in place over the area of interest (mainly the North Atlantic and Norwegian Sea). It was soon obvious that the constellation was unaffordable. A lot has happened in technology since the 1980s, but the physics of satellite motion has not changed. Nor has radar physics. It would still take an unaffordable satellite constellation to maintain surveillance of much of the world’s airspace.
Tracking Can Be Tricky
Airliners like MH 370 are self-tracked. Thanks to the modern technology of GPS satellites, they can always tell where they are. Tracking means reporting those positions. If for some reason someone on the airliner manages to turn off the reporting devices, the plane disappears from the tracking system. In the case of MH 370, the only remaining tracking came from an aircraft reporting system contacting its distant base via a system of INMARSAT satellites in high (geostationary) orbits. The current search is based largely on INMARSAT data. The airplane received a timed INMARSAT signal and sent back a reply. Precise timing of the reply indicated the distance to the airplane when the signal was sent, and that in turn indicated a circle on the Earth’s surface from which the signal came.
In fact, geostationary satellites aren’t stationary; they move in figure-eight shaped paths because the Earth is not a perfect sphere. That movement imposes Doppler on signals. The Doppler provides a crude estimate of the position of the airplane sending its signal up to the satellite. In 1982 the Argentines used this to find the British task force heading for the Falklands. Some readers may remember that the British fleet was found in the open ocean by an Argentine Boeing 707 airliner requisitioned for the purpose, and that it soon suffered an attack by long-range Argentine aircraft (which proved ineffective).
At the time, it was widely imagined that naval forces had gained invisibility by trading in their high-frequency (HF) radios for satellite systems. The satellite uplink, the only part emitted by a ship, was a narrow beam difficult for any enemy to intercept. It turned out that the downlink carried the Doppler imposed on the uplink by satellite motion, and the downlink was anything but narrow. When the coalition force built up in Saudi Arabia for the attack on Kuwait in 1990-91, the Russians managed to track the buildup by, among other things, the satellite technique. The United States later spent heavily to overcome this problem by having its satellites create entirely new downlink signals which would not include whatever was mixed into the uplink. It is not, incidentally, clear to what extent the Doppler processing technique is unambiguous. That circle around the Indian Ocean is quite large.
The story of the pinger assumed to be on the aircraft’s black box is relevant to antisubmarine warfare. Probably the single most difficult aspect of ASW is signal processing, picking a very weak signal out of a great deal of ocean noise. Signal processing is based on the idea that the noise is random, so that if the signal persists continued listening will pop it up. Many other technical solutions to signal processing work the same way. One way to help the signal processor is to compare what is received to a template based on the expected signal. The more complicated the signal, the better the chance that it will pop up above the noise. The template approach may create false alarms that explain multiple claims that the black box had been found several miles (even tens of miles) apart, even though it is a high-frequency signal supposedly undetectable beyond about 1.6 km (about 0.87 nautical miles).
It seems odd that anyone had much hope of detecting the signal if the airplane were indeed on the bottom so far below the rated maximum depth of the Bluefin-21 unmanned underwater vehicle (UUV) searching for it. Stated depths suggest that the searchers were hoping to pick up extremely faint signals by powerful signal processing and that they were willing, in effect, to accept a high rate of false alarms. Another way to say this is that the fainter the signal, the higher the false-alarm rate that must be accepted in order to detect it at all. It now appears that all of the detections were just that. That must be a sobering commentary on future ASW against very quiet objects such as small, armed UUVs.
Accident . . . or Something More?
All of this omits the most remarkable aspect of the story. MH 370 is being treated as a tragic accident, but there has been no explanation relating an accident to the supposed course of the airplane. Early in the search, when it was reported only that the airplane had veered off course back toward the Malaysian coast, several aviation-safety experts offered a scenario: a cockpit fire causing the pilots to turn toward an emergency airfield, only to be incapacitated by fumes. The airplane would keep flying until it ran out of fuel and then dive steeply into the deep part of the Indian Ocean.
Unfortunately the fragmentary radar data indicate that the airplane made several maneuvers after breaking contact with air traffic control. At the least, it crossed the Malay Peninsula and it seems to have flown up the Malacca Strait before turning south (this last turn was not confirmed, it seems, by any radar). It was also reported that the airplane descended to 5,000 feet, which would place it under most shore radar coverage. As of the beginning of May, the course estimated on the basis of satellite pings took the airliner south and east toward the western coast of Australia. The last satellite contact was partial, and the search as of early May was centered on a spot it indicated.
It may be that the airplane was under the control of its inertial navigation system, which had somehow reset itself to take the aircraft far from its intended path. Normally such a system is set by the pilot before takeoff to guide him to the intended destination. In most if not all cases, moreover, the inertial system tells the pilot that he is at a waypoint (requiring a turn), but it does not turn the airplane. That is evident from the few cases in which U.S. pilots have missed their destinations. If the autopilot was not turning the airplane onto a succession of new courses, then someone else likely was. It is then difficult to understand why the end of MH 370 is still being handled as another episode in aviation safety. Surely it took someone at the controls to make all of those turns.
A really suspicious person would point out that the radar contacts were only presumably with MH 370. The surveillance radars involved do not paint a picture of what they see. Unless they are coupled with some form of interrogator, they do not indicate identity. Anyone determined to confuse investigators could fly an airplane with the wrong transponder code, in which case we have absolutely no idea where MH 370 went after it broke radio contact with Malaysian air-traffic control. The last more-or-less certain contact was a good-night message to Malaysian air-traffic control less than an hour after takeoff. The only other specific contact is a report of a visual sighting of an airliner in Malaysian Airlines’ livery over a small island some hours later, but such sightings are not always reliable.
No group has publicly taken credit for hijacking MH 370, or for destroying it in the air. It seems clear that the aviation community would much rather be dealing with a horrific accident than with an apparent collapse of security. There is also an international issue here. Most of the passengers were Chinese, and China has made clear its displeasure with Malaysia’s treatment of the loss of the airplane. It did not help that Malaysia released an extremely short (five-page) preliminary report, which concentrated on lapses that delayed the report of the missing airplane for about four hours. It was interesting that the Malaysian prime minister said shortly before the report was released that he was not convinced that the airplane had been lost.
Whoever (if anyone) was at the controls might be aware of the ways in which airliners remain in contact with the Earth, including the satellite pingers. Anyone really sophisticated would be aware of the way Doppler makes a satellite contact meaningful for location. Signals can be faked. The electronic countermeasures community does that all the time, and anyone reading about signals intelligence will be profoundly impressed by the ingenuity involved. The last pathetic half-ping, received eight hours after the airliner took off, has been taken as evidence that it crashed into the water, but it might be evidence of a landing or of a fire ax finally used against the right wire on board the airplane. It is even possible that all the pings were deliberately faked.
For that matter, a four-hour gap in the story would leave time for an airplane to land and refuel, or to land and be destroyed (hence not to be visible on later satellite pictures of airfields around the Indian Ocean). Perhaps our picture of the world is not so complete as we normally imagine—with huge implications for military capabilities.