Malaysian Jet

Here's a question:

One thing I've learned from watching Air Crash Investigations, as well as similar shows and reading about air crashes in general, is how often a minor error can turn into a disaster.

Of course, this is rather understandable in the context of flying an airplane. After all, the pilots react in the way they think will save the aircraft, very often by resorting to a set procedure. The error often lies in misidentifying the problem.

The question is this: can a small error in the investigation of a crash lead to completely misdiagnosing the crash? In the context of MH 370, is there some subtle error, overlooked by a huge number of people, keeping the search in the wrong area?

Quote: Nareed

The question is this: can a small error in the investigation of a crash lead to completely misdiagnosing the crash? In the context of MH 370, is there some subtle error, overlooked by a huge number of people, keeping the search in the wrong area?

As I understand each ping was analyzed for doppler effect to determine direction of travel, and the analysis underwent a peer review process with space agency experts and contributions by Boeing.

Any analysis like that is subject to accumulation of errors. If you are continually off by 1/10 of a degree, but the errors are always in the same direction, by the time you get to the end of your projected path, you could be very wrong.

You are much better if you have random distribution of errors. Some clockwise and some counterclockwise. You still have an "area of uncertainty", but you can add a probability distribution that makes sense.

Yes, often a small error... but often induced by design principles or a small error that is "obviously waiting to happen".

an intersection "abels"... obviously dangerous named with well known and often used phrase of "when able" . so "descend when abels" becomes to the pilot "descend when able" and so he descends into terrain rather than when he reached abels.

an overloaded airplane just out of el cheapo maintenance that skipped three steps to save money and a routine 'gear up' combine to make the airplane uncontrollable and pilots never think in the next fifty seven seconds that things were fine while the gear was down.

a descend at "3.3 degrees" that gets entered in the button pushing mode as "descend at 3,300 feet per minute".

a kneecap that nudges the yoke just enough to make the computer think the pilot is flying the aircraft and so altitude hold is turned off while everyone concentrates on a blown out seventeen cent light bulb that indicates nose gear is or is not down and locked.

a pilot who signs paperwork just before take off that reads "empty oxygen generators" when its "expired (but full) oxygen generators" and some spanish language clerk interprets 'expire' as if it were spanish for empty, rather than english for out of date with the result the plane takes aboard a hundred pounds of ultra sensitive explosives that burn at five thousand degrees.

that is why there are INTERIM reports and public invitations to comment.

that is why investigators do a "four corners" preliminary determination. If you've got nose, one wing tip, tail and the other wing tip all at the crash scene, you look at terrain, fuel, etc, but if find the tail section a mile away from the crash scene you look at bomb or overstress, corrosion, maintenance, etc.



Pilots can make minor mistakes but investigators have the advantage of devoting time and effort to testing things.

It's interesting, (or maybe not) that we don't have at least one random shot of the plane by satellite. I know we don't have too much reason to monitor the open ocean, but you figure all the spy satellites and everything else do pass over areas anyway. Or maybe it's classified.

If you look at googles unclassified satellite photos, pretty sure you could spot a plane with the resolution, and maybe some clever algorithm. There's always open non-cloudy places.


.

australia is still working on its 'non existent' open sea surveillance which airplane surveillance, satellite surveillance, cell phone data and radar data to spot immigrant boats and illegal fishing boats.

france is interested in tasman sea and timor sea activity (noumea and New guinea)

ain't much interest in staring at miles and miles of open water 1,7oo miles west of Perth. No real threats from there.

think also of politics, if you release your military data... your enemies know your capability.
that is why af 447 from rio to paris disappeared and only minute segments of non aviation radar tapes are made available.

comarco swift departed the search area a few days ago.

coyo septimo, spanish tuna fishing long-liner following ziz zaggin course south of the search area reported nothing.

Quote: Pacomartin

Any analysis like that is subject to accumulation of errors. If you are continually off by 1/10 of a degree, but the errors are always in the same direction, by the time you get to the end of your projected path, you could be very wrong.

You are much better if you have random distribution of errors. Some clockwise and some counterclockwise. You still have an "area of uncertainty", but you can add a probability distribution that makes sense.

The thing about the unexpected is that no one expects it. Therefore it tends to come as a big, often nasty, surprise. No one expected metal fatigue in pressurized jet aircraft. I can't think offhand of other examples concerning air travel, but I'm sure there are a few.

There have been no more news regarding additional bits of wreckage, nor of any organized search in the area of Reunion island. The piece of flapperon found hasn't been analyzed yet. So there's really nothing more at the moment.

What I'm saying is, if MH370 winds up near Reunion rather than where everyone thought it was, there will be some explaining to do.

Despite French officials hedging, the piece of wreckage has been identified as being a part of MH370.

The thinking is that it was carried there from the crash area by ocean currents, therefore there's no plan to change the search area. If further pieces are found elsewhere, their locations could be used to help triangulate the likely crash location more precisely.

On a side note, the most idiotic headline of the week goes to a local free newspaper (I forget its name), which declared yesterday "Mystery solved." Meaning we now know the plane crashed. Well, yes, but that was the likeliest assumption. What we don't know yet is why it veered off course and why it crashed.

Quote: Fleastiff

australia is still working on its 'non existent' open sea surveillance which airplane surveillance, satellite surveillance, cell phone data and radar data to spot immigrant boats and illegal fishing boats.

When we do find the crash site, someone should go back and look at any appropriately targeted satellite data from that date. Assuming it is kept.

It would not necessarily surprise me if something can be found in the data once they know exactly when and where to look.

Unless the dang things are turned black over open ocean they are still seeing a track somewhere in their orbits/

the trouble is that satellite imagery can only help pinpoint the LKP ... Last Known Position. winds and currents still exert their effects and by the time images are analyzed its all historical data...better than none but debris shown on an image will not be there by the time you send a long range Orion out to look more closely.

Quote: Fleastiff

the trouble is that satellite imagery can only help pinpoint the LKP ... Last Known Position.

Well, the LKP probably has an area of uncertainty (AOU) to begin with. That AOU simply grows with time. In the Navy you have a fleet of slow moving P-3's which are lept aloft. The P-3 was based on the a commercial airline introduced in 1957 that was capable of only 600 km/h but could stay aloft for long periods of time.



The P-3 tries to localize the contact using sonobuoys before the AOU gets too big.