The discovery last week of what appeared to be a piece of Malaysia Airlines Flight 370 on the shores of Réunion Island seemed at first blush a giant leap toward solving the famously perplexing mystery. Officials declared that, based on photos, the part could only have come from a Boeing 777. And since only one 777 has ever been lost at sea, physical evidence of the vanished plane seemed at last to be irrefutably in hand.
This marked a huge break in the case, since before now not a single piece of wreckage had ever been spotted. The only evidence that the plane had gone into the ocean was a series of difficult-to-decipher signals received by the satellite company Inmarsat. The incongruity led some, including me, to question whether the plane had really wound up in the Indian Ocean at all. Back in February, I explained in New York how sophisticated hijackers might have infiltrated the plane’s electronic bay in order to spoof the satellite signals and take the plane north to Kazakhstan. MH370 wreckage on the shores of Réunion makes such explanations unnecessary.
Investigators hope to glean from the six-foot-long chunk important clues about where and how the plane went down. The piece, called a flaperon, forms part of the trailing edge of the wing, and was located just behind the right engine. The front part of it looks dinged up but more or less intact, but pieces on the side and much of the rear part have been ripped away. That damage might have taken place in the ocean, but if on inspection it appears to have been caused by high-speed airflow (as a plane might experience in a steep dive) or impact with the water, it could shed light on the flight’s final moments.
The fact that the debris was found on Réunion itself provides a hint as to where the plane went down. The island lies on the far side of the Indian Ocean from the suspected crash area, a distance of some 2,500 miles. The ocean’s strongest east-to-west current, the South Equatorial Current, runs about a thousand miles north of where searchers are currently looking. Should the search area be moved up? In the coming weeks oceanographers will be refining their models in order to figure that out. To lend a hand, biologists will examine the barnacles and other sea life found living on the debris in order to determine how long it was in the water and what part of the ocean it passed through.
But, as if steeped in the weirdness of all things MH370, the Réunion flaperon came wrapped in an unexpected layer of ambiguity.
All airline parts carry identifying labels, much as cars carry Vehicle Identification Numbers etched on the engine block. In the normal course of things, this plate should have been attached to the rib end of the flaperon and allowed investigators to make an instantaneous identification. As fate would have it, the plate is missing.
It’s going to be a tricky job, and the stakes are high: MH370 has unnerved the aviation community like no crash before. Until we can figure out what took it down, the danger is ever-present that it could happen again.
While the world’s attention is on the flaperon, however, the sonar-scanning of the seabed on the other side of the Indian Ocean promises to tell us even more about MH370’s fate. If the small flotilla of search ships can locate the plane’s primary debris field on the ocean floor, they’ll likely find the black boxes that can tell us exactly what happened to the flight. But even if they don’t, they’ll reveal something important about what happened.
The area they’re scouring was defined through analysis of the Inmarsat satellite data. Part of the data tells investigators that the plane must have wound up somewhere along a broad arc 3,000 miles in radius. Another part, subjected to a new and complex form of analysis, showed that the plane headed in a generally southern direction. Where, exactly, depends on how it flew. If the plane flew slowly it would have taken a curving path and wound up north of a subsea feature called Broken Ridge. If it flew fast, its path would have been straighter and taken it south of Broken Ridge.
Among the attractions of the latter option was that it fit with an easy-to-imagine scenario: that, after flying up the Malacca Strait, whoever had been in control became incapacitated and the plane flew straight south on autopilot as a “ghost ship” until it ran out of fuel. Once that happened, the plane would have quickly spiraled into the ocean within a few miles of the final arc, meaning that the debris would have to be located within a fairly small area of seabed.
Last October, after months of internal debate, Australian officials decided that the straight-and-fast scenario was more likely. They laid out a 60,000-square-kilometer search grid and hired contractors to begin scanning. Their confidence in their analysis was so great that they reportedly kept a bottle of Champagne in the fridge, ready to be popped at any time. The longer they searched without finding the plane, officials said, the more their confidence grew, because they knew the plane had to be inside that box.
As time went by, however, a problem emerged: The plane wasn’t there. After six months, there was a 99 percent probability that the search had covered the calculated end point, and that number only kept climbing toward 100. Authorities stopped talking about how sure they were that it was in the 60,000-square-kilometer area, and announced that they would expand the search zone to twice that size.
What went unremarked upon in the general press was that there was no theoretical justification for the authorities to continue the search in this way. To get so far from the final arc, the plane would have to have been actively piloted, because only a conscious pilot could have kept the plane out of a death spiral. So the ghost-ship scenario was out the window. A plane held in a glide by a conscious pilot could travel for a hundred miles or more, far too huge an area of ocean to scan. The only reason to search the extra 60,000 square miles was that, for the authorities, it was better than admitting they had no idea what they were doing.
It also kept them from having to contemplate other unattractive alternative scenarios. Perhaps the plane didn’t fly straight and fast, but slow and curvy, and wound up north of Broken Ridge. It’s hard to imagine why someone would fly like this, but then again it’s hard to imagine why someone would sit patiently on a six-hour death flight to nowhere. If a slow, curvy flight was what happened, then again a terminal death spiral could by no means be assumed, and the required search area would be impossibly large.
To be sure, none of these scenarios make a lot of sense. But then, so much of what we know about MH370 is baffling. If the perps flew into the southern Indian Ocean because they wanted to disappear, why didn’t they just fly to the east instead of turning back over the Malay peninsula? If the aim was suicide, why not just put the nose down and crash right away, like every other suicide pilot we know of? And why did the perps turn off the satellite communication, and then turn it back on again, a procedure that — by the way — few airline pilots know how to do?
Though it has earned much less attention from the world press, the failure of the seabed search actually tells us a lot about what did or did not happen to MH370. And what it tells us is that this case is as weird as ever.
This piece originally ran on the New York magazine website on August 4, 2015.
For Dennis – http://www.watoday.com.au/business/aviation/mh370-intact-part-lifts-odds-plane-glided-not-crashed-into-sea-20150806-git4mq.html
@Matty
Thx for the link. I was leaning in that direction, but those kinds of conclusions are far outside my area of competence. We shall see what the analytics tell us.
Within several months do not be surprised to see a LOT of MH370 debris washing up on the east coast of Africa. After all, seems like pieces of the plane are floating westward, and there is just no way they can miss the African continent. Who knows what will end up there.
I’ve just come across all this and have a simple theory that the Russians shot down MH17, an identical plane to 370 and didn’t let anyone near it for days so they could extract bits of plane and dump them in the Indian Ocean to eventually wash up somewhere….
@Matty
I see the new “zombie awakening” hypothesis of the IG being the new rallying point. Sad really.
Richard,
“Are these the best estimate from the models, or the best estimates with known errors allowed for?”.
I am afraid none of the above. Those are just more or less standard engineering approaches. Best estimate from the models requires quite more work.
Yes, wind has notable influence. When I was involved in the drogue studies, our estimates were up to 5 cm/s “on top” of currents for winds of approximately 8 m/s. But this indeed depends on the shape of a buoy above/below water. Such a wind could displace such a buoy by 2,000 km over the period of 500 days.
>I am afraid none of the above. Those are just more or less standard engineering approaches.
Perhaps you could explain that answer rather more. My original question was as follows. The drift studies refer to circulation models that have been derived from buoy data (and other information). Those models will have an error (i.e. in current velocity and direction at any point). When the models are applied to an actual situation to predict the track of test particles, are the errors propagated into the prediction? For example standard weather forecast synoptic charts do not do that, they show a best result of the prediction model.
What you say about the speed of buoys with respect to the underlying current is very interesting, I guess the 5cm/s (~0.6% of the wind speed) corresponds to the 1.2, 1.5 and 1.8% used in the CSIRO work.
Hi Jeff,
Interesting as always,
Would you say at this point, that your previous theory about Kazakhstan is for sure out-of date?
What we have so far is – as I understand – unmarked wing. There is still nothing more but assumptions regarding Indian Ocean version.
@Woo, Once the piece is confirmed as having come from MH370, we can take the Kazakhstan theory off the table. But you’re right, we’re not quite there yet.
Even if/when the French investigators and Boeing are convinced that this one item (the flaperon) came from 9M-MRO, the question as to how it got to La Reunion can’t be answered with 100% certainty IMO. By all means call me cynical, but the best news we have had lately is that the French may be able to get more information out of the Malaysians.
I just skimmed the new comments. And I am delighted to see that some critical thinking has set in again.
I’m not advocating paranoia. But let’s keep our eyes open.
And please, don’t put those who ask the very legitimate question how this part wound up on the beach of this tiny island, automatically in the fringe corner. I’m sure the investigators look into this as well.
@Jay, yes this arson business with the avionics shop is one of the weirder aspects of this strange case. And good that you mentioned the series of minor incidents that befell MAS after mh370. It was officially suspected as being sabotage and was investigated under that hypothesis. Of course we never heard more about it.
@Arthur Sorkin, I like your remarks about scientific thinking.
I’m open to all out-of-the-box ideas as long as they are plausible. But I think the idea that parts of mh17 have been pilfered in order to disguise them as parts of mh370 is highly unrealisti IMO. I confess that it has crossed my mind, too, and just to be sure U would feel better if the Dutch could confirm that they have salvaged the flap of mh17, but how could that work? Mh17 was shot down by a BUK- missile over land and that has impacted all or at least many parts of that plane like an ID code. Whatever happened to mh370 was probably a very different scenario and the engineers should be able to tell the difference.
CNN correspondent Saima Mohsin (@SaimaMohsin) sent out this tweet this morning:
“#Malaysia tells #CNN
1. A serial number matched #MH370
2. The paintwork matched
3. There was a MAS maintenance seal on the flaperon”
Thoughts? To me, #1 seems impossible, that would have been the kind of definitive proof that would have satisfied the French, and I think we would have heard about it. Most likely they were referring to the part numbers painted on the skin, which are unique to 777s but not MH370. #2 is very plausible–it would add weight to the theory that the part is from MH370 without fully proving it. As for for #3, I’m intrigued as to what this means, I’ve sent a DM to Mohsin on Twitter to see if I can find out more.
@jeffwise: To add more to the confusion, Malaysia claims more aircraft debris was found on Reunion, which was quickly denied by France. It appears Malaysia was referring to the window part, now believed to be part of a sewing machine. Why would Malaysia believe that France would go along with this lie?
This is playing out for the world to watch in a bad way for Malaysia.
http://bigstory.ap.org/urn:publicid:ap.org:2438ecd4e24749d6b1ddfd617ce733a8
Jeff, I know what this is going to sound like but what are the odds someone put this in the water. I realize that the odds of it being found are huge but without the I D numbers on the plate which is really the only real damage on this flap I am suspicious this is a plant. Sorry this just won’t go away.
@ Oleksandr & Richard – FWIW, a test performed during the AF-447 investigation showed a floating human body travels at about 2% of the wind speed.
@VictorI, thanks for sharing the article.
U.S. Secretary of State John Kerry… “if it is found to be conclusively from the aircraft — will help to bring some sense of closure.”
TOTALLY WRONG Mr. Kerry.
Knowing the flaperon is MH370’s brings NO CLOSURE for family members, only anger at how it’s been handled since day 1.
Determining how, when, and where the flaperon disengaged from MH370 that leads to the resting place of MH370 along with the understanding of how and why this tragic incident occurred will help bring some sense of closure.
Officially it has not been confirmed the flaperon was MH370’s. We still wait to here from the professionals, not grandstanding political ‘leaders’.
Next we await to learn how did the flaperon disengage from wing? (low speed water, high speed water, angle of attack, flaperon position, in air flutter, physical strike to object, etc…) We are a long way from uncovering all of Malaysia’s hidden truth’s. The Malaysian’s could short cut this drama with full disclosure, but they chose to shamefully hide and torture NOK. But I’m hopeful the world continues to press until the truth is uncovered and the entire is incident laid out with reasonable clarity.
Shame on Mr. Kerry, trying to put a label on the discovery to convince the less informed that this incident can be closed and we all can move on. Mr. Kerry’s statement refreshes raises concerns over the US involvement, knowledge and silence.
@Benaiahu
Thinly veiled anti-US sentiment by an Israeli. Gets a bit tiring after awhile. US taxpayers are beginning to lean away from a sympathetic position. Don’t shoot yourself in the foot with nonsense. Save it for the real beefs.
@dj
The part actually is quite damaged. The trailing edge is missing 1-1.5 feet and is ripped in a jagged line. Only the leading edge appears too good to have survived a crash. In the clean configuration, the leading edge would have been tucked inside the wing.
@Jeff: 1) could be a case of “lost in translation”. I know this has caused confusion in the past (e.g. “downed”) – I’d want a few translators to get together and corroborate “serial number” before scratching my head too hard.
As all who’ve followed this saga will agree, reporting can be wildly inaccurate even in the ABSENCE of language barriers. Here’s what 5 seconds of internet keyword searching gave me (courtesy AirLive.net, July 30):
“UPDATE 12:38UTC Here is the serial number of plane debris: ‘657-BB'”
My views will probably not be shared by others.
While everyone wants to sling mud at Malaysia, France has done much to exacerbate the situation. Why is a French judge deciding how the the investigation will proceed? Why is a deputy prosecutor presenting initial findings? The job of a prosecutor is to assign blame, not solve the mystery of an airliner accident. By mixing up its (independent) criminal investigation with what is otherwise an international civil investigation, France is just making a mess of the whole process.
John Goglia has written an opinion piece for Forbes that explains it better than I can (use Google – he has written several articles worth reading.)
@DennisW,
1.0 Who’s an Israeli???
2.0 What are you tired of?
3.0 What does anything I posted have to do with US taxpayers???
4.0 What nonsense did I shoot my foot with?
5.0 Save what for real beefs?
I’m so lost with your post, it makes no sense to me regarding MH370. I don’t know all the personalities for bloggers here so forgive me for not knowing your positions. I’ll go back and read some of your posts to better understand. Look forward to you answers to above, maybe I can catch on to your point.
@dj, I think that’s an utterly valid reaction given the situation as it stands.
@VictorI, Agree 1000%
@SK999
Agree. It is the French being French more than anything else.
@Benaiahu
You know exactly what I am talking about.
@sk999: Independent of what the French prosecutor said regarding the criminal investigation, the bigger problem is that Malaysia came out with a statement that conclusively stated that the part is from MH370 when at the same point in time, the NTSB and Boeing have stated that the maintenance records for the flaperon of 9M-MRO don’t exactly match the maintenance that was performed on the recovered flaperon.
Why was there the rush to for Malaysia to make this proclamation before Boeing had definitively concluded the same? Boeing has stated that there is a discrepancy that needs to be resolved. Why should Malaysia override the concerns of the manufacturer of the airframe, whether or not those concerns ultimately are justified?
@DennisW: I don’t know what you mean. Please fill me in, either here or by email.
I apologize if this appears twice as my last post has not yet appeared.
@sk999: Independent of what the French prosecutor has said, the bigger issue is that Malaysia has pronounced that the part is from 9M-MRO at a time when Boeing and the NTSB have stated that the maintenance records for the flaperon of 9M-MRO and the recovered part don’t exactly match.
Why did Malaysia proceed with the announcement while the airframe manufacturer is trying to resolve this discrepancy? Are they worried that the testing will be inconclusive (or embarrassing) and so they wanted to control the message?
@VictorI
The ICAO rules are very clear that Malaysia, the country of registry in this case, is to have unrestricted authority to gather information and conduct the investigation. Under ICAO rules the French should have given the flaperon to Malaysia. That is what the Monday meeting in France was all about, and why I raised the issue earlier. France is keeping the flaperon on the pretext that it is investigating an act of terrorism involving French citizens (five were on board MH370).
The ill will probably goes back further when France was not given full access to data it had requested from the Brits.
@Brock – I do not know how it’s done in the airplane industry, but I can tell you about Valves for Nuclear Power Plants. There is a Code that defines certain Valve Parts as “Safety Related.” The supplier and purchaser can agree to add other critical parts to that list. All of these parts will have permanent identification with the part number and serial number. The part number identifies the part (like 657BB). The S/N identifies on which Valve it was installed (like 9M-MRO). A separate QA documentation package gives a full pedigree for that part. Note, however, all of this information only applies to those Parts on the list. The balance of the parts are considered “Commercial Grade” and could come from your local hardware store.
I’ll bet a similar process applies to passenger planes. For example, I’ll guess a Tray Table will not have a Part Number or a S/N. Certain pieces or all of the pieces of the airframe will. I’ve no idea in which category the Flaperon falls.
@DennisW: Independent of ICAO protocol, which I agree gives control to Malaysia, why would Malaysia make an announcement before the test results were complete? I am not questioning Malaysia’s adherence to protocol. I am questioning the wisdom of Malaysia making a premature announcement before the conclusions were definitive. Surely they knew that Boeing was not fully persuaded, and surely they knew that Boeing would go public to refute Malaysia’s statement, which took only hours.
@VictoI
I can only assume that the Malays wanted to exercise their privilege to be “in charge”. They never delegated any authority to France as they did to Australia to conduct the search in the SIO.
It is all shadow boxing and muscle flexing as near as I can tell.
@VictorI
Have you actually seen a statement directly from Boeing or the NTSB? I have not. My assumption is that they are not authorized to make pubic announcements.
@DennisW: I have seen no official announcement from either Boeing or the NTSB.
VictorI,
One can only conjecture about why Malaysia made the announcement. It is possible that there are enough facts established (e.g., type of paint) that tie the part to MH370, even if there are other additional facts (e.g., serial numbers not yet accessible) that, once developed would, add further, perhaps even stronger, confirmation.
@Victor, DennisW, I think I see what’s going on now.
France is involved because they don’t like the answers they’ve been given. The arrival of debris in their territory gives them jurisdiction to do the investigation they wanted to do all along.
France will not confirm the identity of the part, because once they do, that ends their involvement under ICAO rules. Malaysia, on the other hand, is quick to identify the part, because they want to end France’s involvement.
As long as the part remains in France’s possession, and as long as they refuse to confirm its identity 100%, I believe the ICAO rule doesn’t apply, and they can continue to investigate.
I actually think it’s a good thing, though, even if it’s the French being French. We weren’t getting answers before. Now, at least, somebody has leverage over the Malaysians, and maybe we’ll get some answers. Maybe not all, but more than we’ve gotten before.
@sk999
The French have not specifically stated what the additional tests might be. Makes one wonder why not? My guess is that it is nothing more than a delaying tactic to embarrass the Malays and force them to provide a quid pro quo for being allowed on the investigation site.
What quid pro quo you might ask? Information which the Malays have so far withheld relative to details of the Malay criminal investigation.
All very annoying and murky. Nothing has been transparent in this entire debacle i.e. no radar data, no acoustic ping data, no unredacted ISAT data, no details relative to criminal investigation,… It goes on and on.
@JS
I think you are absolutely correct.
Richard,
Are you asking whether errors found by comparison of the known actual and predicted buoys positions are then propagated to improve accuracy of the model’s prediction? If you are talking about ATSB’s particle tracking (PT) model, from its description I lean to think they took previously calibrated currents, wind and wave data (atsb.com.au/publications/2015/mh370-drift-analysis.aspx). I.e. they took the “best available result” to feed into the PT model. I am not sure about calibration procedure for the PT model ATSB mentioned about.
PT models often include stochastic forcing, for example, by the means of so-called random walking technique. Such models aim to simulate density distribution rather than position of an individual particle. It is practically unfeasible to account for all the forcing (currents, winds, waves) to make an accurate prediction of the location of an individual floating particle in 500 days. Thus calibrating such a model against actual position is not possible: stochastic component has to be “switched off”.
I’m not sure about “adrift” model as I didn’t find a proper description of this model except that it is based exclusively only on the buoys data. It does not appear it accounts for actual winds and waves at all (it does not ask for the date to release tracer). On top of it, it seems “adrift” uses eulerian approach.
Does this answer your question?
All,
With regard to ATSB’s particle tracking plot (atsb.com.au/publications/2015/mh370-drift-analysis.aspx) I noticed a very interesting thing. The presented plot includes positions of real buoys (magenta symbols), which are clearly split into two clusters: northern SIO (22-32S) and southern SIO (40-45S). Do these “magenta dots” depict the buoys dropped along the 7th arc? If yes, apparently the buoys that were dropped at the southern section of the 7th arc are trapped in the “roaring 40th”, while buoys that were dropped in the northern section of the 7th arc travelled towards Madagascar. Knowing buoys IDs, this would give possibility to define threshold latitude. In other words this might be an opportunity to refine the search area.
@Oleksandr
Yes. I noticed the same thing. However, the text below the graphic has this statement:
begin cut-paste//
Magenta symbols are positions of real drifting buoys (with sea-anchors at 12m) on the day. Their movement has been used to estimate the errors of the ocean current component of the total drift velocity.
end cut-paste//
What do you make of that? The only thing I know about the ocean is that it is cold when my kayak tips over.
@Lauren H,
Everything on an aircraft will have a part number but not necessarily a serial number. The flaperon has a serial number; that was on the missing data plate.
There have to date been no identified part numbers or serial numbers from the aircraft. 657BB is an access panel reference number; opening up the flaperon may reveal some part numbers that can be traced.
OZ
There is regular programme of buoy release to maintain the total World fleet of 1250. The buoys shown in the map must include numbers of that population plus any specially released for the MH370 work. I don’t think an inference can necessarily be drawn which are which.
BTW I don’t think I have seen a reference on this blog to the CSIRO page that refers to the analysis.
http://csironewsblog.com/2015/08/05/what-does-our-ocean-modelling-tell-us-about-the-fate-of-flight-mh370/
@Oleksandr – thanks for the answer. I was more asking about the estimated residual errors in the model after calibration on buoy data etc. and if those are then used as a scatter element in the predictions. A stochastic term may cover that.
I don’t feel there is any argy-bargy with the flaperon. Malaysia became unnerved by the possibility that the tests may be inconclusive and rushed out to sway perception. Call it media manipulation. They are dying for a ticket out of this place, since the beginning.
Dennis,
I was referring to the very text. If these “magenta symbols” represent buoys that were dropped along the 7th arc during air search phase (those could also be some generic-purpose buoys), most likely a group of southern buoys traveled to the south and became trapped between 40 and 45S. This means the flaperon would not travel to Reunion if the crash occurred to the south of certain latitude. And this latitude can now be established if IDs of buoys are known.
In addition, this would mean that ATSB’s PT model is not sufficiently accurate as it does not reproduce particles traveling to the south.
@Oleksandr
The animation at the CSIRO site shows the release of the MH370 hunt specific buoys. They appear to have a short life.
@Oleksandr
OK. I interpreted the fact that the bouys had long anchors to mean that they were largely unaffected by wind, and only (mostly) affected by current. I thought the bouys were largely intended to model the current, but that the PT model included the affects of wind and current both of which would affect the surface debris.
OZ,
This missing plate from the flaperon with the serial number – I am imagining this plate is tightly secured with rivets or bolts or screws or soddered, etc., in other words, “in like Flint” on the flaperon. How is this coming off on impact, forced off from air flow, what? It would seem it would take manual force with an electrical drill or screwdriver to get it off? I can get how the flaperon would be bent or the edge jagged from being torn off or dragged but the plate???