Quoting WingedMigrator (Reply 83):
Quoting WarrenPlatts (Reply 80):Significant IMHO is the fact that there are not very many such objects--that in itself is could be an important clue.
The sample of debris suffers from extreme selection bias. Among the debris washed up on distant shores, all of them floated. Among the few that were identified as coming from the aircraft, all of them bore distinctive identifying markings. Everything else is either on the ocean floor or was overlooked as the usual flotsam. This isn't some clue; it's common sense. |
There are several sources of sample bias:
1. selection bias
2. density bias
3. crash mode bias
Of the three, what we are most interested in #3: in particular, the energy of the crash will affect the character of the debris.
A high energy crash should result in a crash site relatively close to the 7th arc: we should expect the initial pool of floating objects to be measured in the thousands; the objects in general will tend to be small sized (cf. Silk Air 185); the fuselage would break apart, releasing interior objects that can float. Moreover, the range of debris objects should be bilaterally symmetrical in their distribution.
A low(er) energy crash resulting from controlled flight inputs would tend lose parts from the trailing edges of control surfaces (a la US Airways 1549); the overall number of pieces should be much less, perhaps on the order of 100 or so; some larger pieces would not be unexpected. Depending on which side impacted the water first, some bilateral asymmetry in the distribution of found parts could be expected. Also, the crash site could be up to 130 nm below the 7th arc.
As for density bias, as the sieving in this case is due to the density, and it's going to be present in any overwater crash. For exterior objects, these will mostly be composite materials, and as such tend to be concentrated on control surfaces. So at least for debris objects deriving from the exterior, the mere fact that the found objects float can't really distinguish between high and low energy impacts.
The big question in my mind is why more objects have not been found. Drift studies show that roughly 20% of all floating debris should have washed up on a beach somewhere already. If there were "thousands" of objects as some contend, then hundreds of objects should have washed up by now. If there were only a hundred objects, then only a few tens of objects should have washed up.
The question is whether selection bias could account for the dearth of found objects so far even if a surfeit of objects has washed up?
Sources of selection bias: (1) income/education of finder; (2) population density; (3) presence or absence of trash collection; (4) formal search parties; (5) media intensity; (6) presence or absence of distinctive markings.
Treating these in no particular order, it is striking that of the African finds, they were all found by relatively well-off, white people: two were foreign tourists, one was a professional archeologist. Drift studies show that the debris should be concentrated in Madagascar, southern Somalia, Kenya, Mozambique, and to a lesser extent South Africa. Of course Somalia is pretty much a no-go zone these days. In Madagascar, the average person makes do with $2/day. In Kenya, the main attractions are the interior national parks, and recent terrorist attacks have drastically slashed the number of tourists visiting the last couple of years. Mozambique has fantastic beaches, but the country suffers from a lack of basic infrastructure.
On the other hand, South Africa and Australia are modern countries. Australia is famous for its beach-combing, outdoors-loving culture. Drift studies suggest that perhaps 4-5% of objects should have washed up in Australia. If there were "thousands" of objects, then at least 100 objects should have washed ashore in Oz. Yet nothing has been found, despite extensive coverage of MH370 in the media there.
As for South Africa, certainly the public there has been sensitized by the recent finds, as evidenced here by the fellow who found the porta-potty toilet. Yet nothing else besides the Rolls-Royce object has been found.
Formal searches: there have been few. I think Malaysia dispatched a team to Mozambique. Nothing has been found. There were extensive formal searches in Reunion, Mauritius, and possibly Rodriquez. Nothing more has been found there, the "cheat line" object found by Mssr. Begue notwithstanding.
Certainly more formal searches could be undertaken for a tiny fraction of the cost of the $100 million underwater search. I would suggest a formal search of the NE coast of Madagascar could potentially be productive.
As for lettering, my experience looking at the Aircraft Maintenance Manual is that virtually every part on the exterior has a 5 letter code stenciled on it somewhere. To be sure, all three finders of the African objects report that it was the markings that really piqued there interest. On the other hand, at this time, people know what to look for: honeycombed aluminum sandwiched between composite skins.
The lack of reports from places like Madagascar and Somalia could be explained by selection bias. However, the lack of more reports from South Africa and Australia are really hard to explain on the basis of selection bias. If there were a lot more objects to be found there, they would be found. Therefore, the lack of such objects represents a real pattern IMHO.
[Edited 2016-03-29 06:11:05]
