The mystery of flight MH370


Continuing airsearch (Image source: Shen Ling REX)

“Data! Data! Data!” he cried impatiently. “I can’t make bricks without clay.”

If anything teaches us that the modern media is for the most part bat-shit crazy the continuing whirlwind of speculation does so. Even the usually staid Wall Street Journal has got into the act with speculative reports that MH370 may have flown on for hours under the control of persons or persons unknown… sigh.

This is as Sherlock Holmes might have remarked more than a three pipe problem, and this mystery will take more than fifty minutes to resolve. So like Holmes I’m not going to speculate without the facts, as such is a fertile breeding ground for narrative fallacies of the worst sort. However, there are at least some conclusions we can safely draw from the current mess.

First, that search and rescue relies on knowing roughly where to search. When you have no real idea and can only draw a circle on the map then the difficulty of your job goes up as the square of the radius. So at some point no amount of assets will be able to cover the ground, and thus you’re back to relying on expert judgement under uncertainty, e.g. guessing, as to where to look. Knowing that the aircraft is within a reasonable distance of it’s intended flightpath allows one to collapse the search field and makes the job technically feasible. Very Bayesian.

Secondly you cannot infer from a transmission ceasing that an aircraft has crashed, there are simply to many scenarios involving human agency, equipment failure or both that could cause the loss of a system, or systems, but not the aircraft. Put simply lack of evidence (a signal) is not evidence of lack (a crash site).  In practice the only way to reliably determine a crash site is to have a crash survivable system, independent of all other onboard aircraft systems, and unlikely to be affected by known catastrophic or common cause events (rotor bursts, fires, human action) that can reliably squawk or mark the crash site in the ocean for a period of time. In other words something that can reliably signal ‘here I am, come and get me’. Unfortunately the current crew deployed or fuselage hard mounted systems and data recorder pingers just don’t pass muster.

Finally, the ocean is still a vast and empty place, and when our technology falls away it can still swallow up a plane or an aircraft as it did in the case of AF447 before, and will probably do so again in the future.