If you were leading a high-profile international aircraft investigation, in command of the world’s most qualified technical experts and in possession of all the relevant data, would you bother listening to a rag-tag band of internet commenters, few of whom actually work in the space or aviation industry, and none of whom have access to all the data?
Most likely, you’d say: certainly not! But as time goes by, and the puzzle remains curiously impenetrable, you might find it worthwhile to pay a listen to what the amateurs were saying. You might even abandon some of your own conclusions and adopt theirs instead.
This appears to be the case in the search for Malaysia Airlines Flight 370, which disappeared en route from Kuala Lumpur to Beijing back in March. From the beginning, the authorities running the investigation — first, Malaysia’s Ministry of Transport, and later the Australian Transportation Safety Board (ATSB) — held their cards close to the chest, releasing very little information about the missing plane and maintaining a posture of absolute conviction. The investigators’ self-confidence reached its apex in April, when their methodology led them to an area of ocean where underwater accoustic signals seemed to be coming from pingers attached to the plane’s black boxes. Officials assured the press that the plane would be found in “days, if not hours.” But then it wasn’t. A scan of the seabed found nothing; the pingers were a red herring (perhaps literally!). Back to square one.
Meanwhile, on the internet, a group of amateur enthusiasts had come together from all around the world to trade ideas and information about the missing flight. The group, which came to call itself the Independent Group (IG), emerged from various online comment threads and eventually grew to about a dozen individuals. This was a truly spontaneous, self-assembling crowd: there was no vetting of credentials, no heirarchy of any kind. (Full disclosure: I count myself among this group.) Basically, if you seemed to know what you were talking about and could comport yourself in a collegial fashion, you were accepted into the crowd.
While the mainstream press was reporting the ATSB’s pronouncements as received wisdom, the IG was raising red flags. IG members were among the most vocal critics of the ATSB’s contention that the accoustic pings probably came from black-box pingers. And later, after a public outcry led Inmarsat to release a trove of data received from the aircraft, and the ATSB issued a report explaining how it had come to identify its current search ear, the IG dove into the new information with abandon, quickly identifying holes in the data and weaknesses in the official approach. In a pair of papers, the group recommended its own search area, hundreds of miles to the southwest of the ATSB’s officially designated zone.
Today, the ATSB has released an update to its earlier report, explaining why it has decided to reassess its conclusions and move its search zone to a new area — one that overlaps, as it turns out, with the IG’s recommended area. (In the graphic above, the white bracket shows the ATSB area; I’ve added a yellow dot to show the IG area.) Needless to say, this has caused elation within the ranks of the IG, who see the move as vindication of their methods, and indeed validation of their combined efforts over the last few months.
A few observations on the new report:
— One of the reasons the ATSB gives for the shifting of the search area is the recognition that Inmarsat data related to an unsuccessful ground-to-air telephone call attempted at 18:40 indicated that the plane had already turned south at that time. The IG had been basing its analyses on this data point for months.
— Since the June report, the ATSB has improved its BFO model by taking into account various factors — such as temperature shifts caused by the Inmarsat satellite passing through the Earth’s shadow and the mis-location of the Perth ground station in an important Inmarsat algorithm — that IG member Mike Exner has been working through in detail for months.
— The ATSB has fundamentally changed its approach in how it is assessing the plane’s likely path. In its June report, the focus was on what I call the “agnostic” approach: it generated a large number of flight paths based on as few initial assumptions as possible, then graded them based on how well they fit the timing and frequency data received by Inmarsat. This resulted in a population of potential flight paths that fit the data well, but did not make any sense in terms of how a plane might be flown. Some of the routes, for instance, involved multiple changes in heading and airspeed. Today’s report explicitly excludes such flight paths. The ATSB and the IG alike now assume that the last several hours of the flight were conducted without any human input — the crew were presumed to be incapacitated by hypoxia or other causes, so the plane flew on autopilot until it ran out of fuel and crashed. This has been the IG’s starting point for ages, and the fact that the ATSB has now adopted it is a major reason for why the two group’s search areas have now converged.
— You can see in the graphic above how an emphasis on matching the Inmarsat data will tend to lead you in one area (“Data error optimisation”) while an emphasis on routes that comport with real-world autopilot functioning will lead you to another (“Constrained autopilot dynamics”). To be sure, they overlap, but the peak area of one is far from the peak area of another. I think it’s important to realize this, because it helps us to understand why it has been so hard to get a handle on where MH370 went, why the official search area keeps moving, and why knowledgeable people have been furiously debating possible flight paths for months: the BFO and BTO data just do not match up that well. In order to arrive at its recommended area, the IG has been willing to accept much wider deviations from Inmarsat data points than the ATSB has been comfortable with.
— Finally, it’s worth nothing that the ATSB approach is superior to the IG’s in one important regard: it is at heart statistical, looking at families of potential routes rather than proposing and assessing one at a time. There is a tendency, as an individual–and I have fallen into this myself–to cook up a solution, run an analysis, and to be so impressed with the result that one wants to shout about it from the rooftops. (Ask me about RUNUT some time.) The IG has come up with a search area essentially by pooling together a bunch of individual solutions, each of which is generated by a different set of procedures and different set of assumptions. It’s a herd of cats. To really move the ball forward a more rigorous approach is needed, one that takes each procedure and sees how it would play out if the assumptions are methodically modified.
The upshot is that, since the early days of the investigation, the attitude of search officials has changed radically. Once dismissive of amateurs’ efforts to understand the incident, they have clearly begun to listen to the IG and to turn to it for insight and ideas. Indeed, you could say that since the release of Inmarsat data and the issuance of the ATSB report in June, the search for MH370 has become effectively crowdsourced: a de facto collaboration between the professionals and a spontaneous assemblage of knowledgeable experts.
UPDATE:
The overlap between the ATSB’s analysis and the IG’s is more evident in the image below, courtesy of Don Thompson. It shows the fan of values calculated by ATSB to match likely autopilot settings.
@Brock
Actually, seven heart beats which happen to all be very close together, making them potentially quite useful indeed. How many hourly pings were there again?
There is no track record for the IG. It’s just a bunch of folks, some of who individually touted all sorts of nonsense that lead us up various garden paths for weeks (in some cases creating false hope for the families) who have clubbed together in the hope that the errors in their diverse theories and assumptions will cancel out and end up with a slightly better projection for the SIO terminus than the ATSB. And, if it doesn’t, no harm, no foul, it’s just an informal club anyway, nobody’s really responsible for any of their ever-changing “conclusions,” they’ll all just slink away at the end of this thing with no “institutional” responsibility entailed. Color me unimpressed with the “gravitas” of the IG.
I think the re-boot of the SATCOM and any odd flight maneuvers over the Malacca Strait are not a mystery at all. Flight tracking sites show a busy airway, many planes are coming in and leaving, the planes are descending and climbing. The pilot of MH370 had to avoid a collision. The operation might indicate, however, that this part of the flight was not a part of the initial plot.
As I stated in posts months ago I don’t have a full grasp of the science or the math involved in the frequency analysis, but I did notice Jeff’s reference to a “herd of cats”. So the question that popped into my head was,” If one were to herd cats, how would one go about doing that?”
Also, for some reason the bouncing signals, pings, had me thinking about a ping pong ball going back and forth. Paddle on one end the sending station, the table is the satellite, and the plane is the paddle on the other end. What if the ball is out of round? What if the ball is out of balance? What does that do? I guess doppelganger shift still trips me up?
Anyway, keep up the good work and pursuit of a solution for the families Jeff.
@Luigi: While ATSB chased rainbows, IG has said s38 since June; had Inmarsat data been disclosed in March, IG would’ve said s38 since April. ATSB is now back at s38. Didn’t think any of that needed explaining.
Best of luck in your search for analysts with the “gravitas” to draw the “conclusions” your theory requires.
(I’m done. Next volunteer?)
Back to the graph on the title page of the ATSB MH370 Update of 8 October 2014 …
Thinking about it a bit longer, I’ve come to the conclusion that the green probability distribution for the optimised paths is the result of the BFO error model defined on page 26 of the ATSM MH370 Definition report of 26 June 2014:
– a random BFO error defined by a gaussian distribution with a standard deviation of 5 Hz
– a uniform distribution for the AES fixed frequency bias (FFB) between 147 – 152 Hz
The ‘flat’ portion of the green distribution between longitudes of approximately 92E and 94E is apparently the result of the uniform distribution of FFB. Think of it as a succession of gaussian distributions, each for a different value of FFB. The northern and southern ‘tails’ of the distribution result from the gaussian distribution of random BFO errors. I initially rejected this interpretation because FFB’s ranging from 147 to 152 Hz result in a much larger spread on the 6th arc, while different locations of the turn south between 1827 and 1840 UTC produce a smaller spread . But then I reasoned that selecting the ‘best 100’ paths truncates the distribution and that truncation may well explain the smaller spread seen in the distribution.
Looking at this process with some detachment, I find it definitely odd:
– Randomly create 6000 sets of errors
– Calculate 6000 error-contaminated paths
– Select the ‘best 100’ paths with the smallest errors
– Plot a probability distribution for the selected paths
Extrapolating this process to its logical end, the ultimate ‘best’ path is obviously that with the smallest RMS error, or with no errors at all. The minimum-error path can be found directly, as demonstrated in Inmarsat’s paper in the Journal of Navigation. Inmarsat’s ‘Example Reconstructed Flight Path’ is also fully compatible with Autopilot constraints and is one of ATSB’s ‘best 100’ constrained paths. It is also almost identical with the ‘zero error’ path for FFB=149 Hz that I presented earlier.
Is the emperor naked?
@Brock
Immarsat has some gravitas here: they figured out the plane was flying for hours and ended up in the SIO, and at least they have published their analysis in some sort of scientific journal. The Malaysian military have some gravitas, because they did actually track the plane in real-time, and their leaks offered by far the most important insight into what happened to the airliner. The Daily Mail has some gravitas because they did a bit of real leg work and found out useful stuff about the guy flying the plane. The authors of “Goodnight, Malaysian 370” have some gravitas because they did a ton of research and pulled it all together into a rational, comprehensible book on the incident. The ATSB and the IG, not so much. Reminder: they haven’t found the plane. So far, these folks have been equally unsuccessful at that task.
http://www.duncansteel.com/archives/1104
The Independent Group
The following individuals have agreed to be publicly identified with this statement, to represent the larger collective that has contributed to this work, and to make themselves available to assist with the investigation in any constructive way. Others prefer to remain anonymous, but their contributions are gratefully acknowledged.
Brian Anderson, BE: Havelock North, New Zealand
Sid Bennett, MEE: Chicago, Illinois, USA
Curon Davies, MA: Swansea, UK
Michael Exner, MEE: Colorado, USA
Tim Farrar, PhD: Menlo Park, California, USA
Richard Godfrey, BSc: Frankfurt, Germany
Bill Holland, BSEE: Cary, North Carolina, USA
Geoff Hyman, MSc: London, UK
Victor Iannello, ScD: Roanoke, Virginia, USA
Barry Martin, CPL: London, UK
Duncan Steel, PhD: Wellington, New Zealand
Don Thompson: Belfast, Northern Ireland
Jeffrey James Wise, BS: New York, NY, USA
I’d like to personally thank the Independent Group for their amazing contribution to the search for MH370. I have tremendous respect for everyone on this list (as well as those not named) who donate their time and talents to help locate the 239 human lives lost in the most perplexing aviation mystery we’ve ever known. I pray we’ll be able to resolve this crisis and reveal the truth of how and why it happened so a similar tragedy can be averted in the future.
~LG~
@Gysbreght
Here is another idea for explaining the green curve.
I trialled fitting with heading and speed changes allowed at each arc (and minimising, to zero if possible, both BTO and BFO errors) with no useful results; the required speeds vary wildly. I have also tried adding noise as a sort of Monte Carlo but again not achieved anything useful.
The ATSB report update states the following for the data optimisation model:
“At each step, speed and/or heading values were varied to minimise the error between the calculated BFO of the path and the recorded value”.
This covers a range of possible models including constant speed courses with heading (course) changes that fit the BTO, as well as more unconstrained models. I note that the only detailed course we have is from the Ashton et al paper table 9 and that has only two speeds, plus heading changes at each arc.
I have looked at all the single speed courses that are consistent with a turn between 18.25 and 18.40UT and scored them against a BFO model that gives similar results to the Ashton et al table 9 (I did not use two-speed models, for my taste there is not enough data to justify adding another parameter to the fit). From the track and BFO models I can read off the lowest BFO r.m.s. error for tracks terminating in latitude bins on the final arc (I used bins of 0.5deg). Because tracks that are not consistent with the turn are excluded, this gives a slew in the error/latitude plot, rather a smooth trend with latitude. I converted those r.m.s. errors into probabilities using the Gaussian (normal) distribution and an BFO error of 2.4Hz (1 sigma, as I have used before), plotted those against latitude on the final arc and blended them with a Gaussian as described previously. As I stated, this is not a rigorous process.
The result (with the original ATSB plot) is shown below.
https://www.dropbox.com/s/qmx1yy2mrpn6hgm/green-curve.jpg?dl=0
The graph reproduces most of the features of the ATSB graph, though the peak at low latitude is not as prominent. The details are sensitive to the BFO offset and the maximum latitude of the final turn.
So in this interpretation the green curve presents the statistical balance of final locations of MH370 using single speed models, and folding in the implications of a turn between 18:25 and 18.40UT.
Thanks Brock, for noting the consistent IG record, pointing the way to the mostly likely area. My personal confidence in the IGs work has never been higher.
We continue to work intensely, 24X7, on the further refinement of the estimate because even the best estimate still has large margins compared to the sonar coverage speed. Thus anything we can do to give Discovery a tighter estimate will reduce the time to finding MH370.
The IG is presently hard at work to refine the satellite ephemeris, BTO Bias and BFO Bias terms, all of which have the potential to shrink the uncertainty and improve the best estimate *at the margins*. An update is in the works. None of this work is expected to move the point out of the current area Equator has recently surveyed, or Discovery has started to search. But it could shave a few weeks, or even a few months off the time required to find MH370 within that 320 X 35 NM section of the 7th arc.
I want to acknowledge that working with the IG has been a reel privilege. The IG is a dedicated group of brilliant people, all of whom are experts from diverse fields. I also want to acknowledge the important work of Inmarsat, ATSB and other independent researchers that have conducted thoughtful research and contributed to the IG’s understanding. It is no accident or coincidence that ATSB and IG now agree on the best site to search fist, AND the general area within which to look around that site. MH370 families and friends should be patient while the search continues at the necessarily slow pace, limited by the speed and coverage of the technology, but confident it will payoff in the end. We are on the right track now.
Please have a read.
Submitted this yesterday, but maybe the spambots ate it:
http://www.twitlonger.com/show/n_1sdejua
And Rand – Thought you’s like 😉 And while SOS instrumentals are sick, most of their songs are with vocals – by the killer Corinne Drewery. It’s all heaven, I’m telling you. Please tell Tokyo I’m coming soon!
You’re welcome, Mike.
I meant it when I said it pained me to lock horns with you and your group. It was in fact the huge analytical gap between the IG (outstanding, transparent) and the JIT (flaky, shadowy) which compelled me to demand accountability of the latter with my “Day 200” open letter (whose data requests I can report were, in the spirit of transparency, summarily ignored). How can these people continue for months to hold unchecked power, when outside groups are thinking rings around them? And with so many lives hanging in the balance, we have no room at all to ASSUME those errors were all honest.
The IG’s September decision to deem s21 plausible was, to me, far beneath its own standards, and, perversely, served as a lead blocker AGAINST my quest for JIT accountability. With the signal analysis seemingly wrapped up, I thought fighting through this lead block (which I still assume is there by accident, despite the coincidental speeds and coordinates) would “distract” only from our slow, patient vigil over the towfish.
But I now understand you remain hard at work on the signal data. Thanks so much for continuing to refine it, and apologies for ever having distracted you from it.
Some within the IG had at one point hinted, I thought, about developing independent performance limit analysis. I’d think such work might prove invaluable in corroberating your Inmarsat data interpretations, and further focusing the search. Was this dropped?
Thanks,
Brock
Seriously: can anyone explain why the ATSB would ask Indonesia to watch its shores?
8 months of plastic drift, starting at 7th arc & s38 (thanks to adrift.org.au):
https://twitter.com/Brock_McEwen/status/525797062705946624
Back to the graph on the title page of the ATSB MH370 Update of 8 October 2014 …
Thinking about it a bit longer, I’ve come to the conclusion that the green probability distribution for the optimised paths is the result of the BFO error model defined on page 26 of the ATSM MH370 Definition report of 26 June 2014:
– a random BFO error defined by a gaussian distribution with a standard deviation of 5 Hz
– a uniform distribution for the AES fixed frequency bias (FFB) between 147 – 152 Hz
The ‘flat’ portion of the green distribution between longitudes of approximately 92E and 94E is apparently the result of the uniform distribution of FFB. Think of it as a succession of gaussian distributions, each for a different value of FFB. The northern and southern ‘tails’ of the distribution result from the gaussian distribution of random BFO errors. I initially rejected this interpretation because FFB’s ranging from 147 to 152 Hz result in a much larger spread on the 6th arc, while different locations of the turn south between 1827 and 1840 UTC produce a smaller spread . But then I reasoned that selecting the ‘best 100’ paths truncates the distribution and that truncation may well explain the smaller spread seen in the distribution.
Looking at this process with some detachment, I find it definitely odd:
– Randomly create 6000 sets of errors
– Calculate 6000 error-contaminated paths
– Select the ‘best 100’ paths with the smallest errors
– Plot a probability distribution for the selected paths
Extrapolating this process to its logical end, the ultimate ‘best’ path is obviously that with the smallest RMS error, or with no errors at all. The minimum-error path can be found directly, as demonstrated in Inmarsat’s paper in the Journal of Navigation. Inmarsat’s ‘Example Reconstructed Flight Path’ is also fully compatible with Autopilot constraints and is one of ATSB’s ‘best 100’ constrained paths. It is also almost identical with the ‘zero error’ path for FFB=149 Hz that I presented earlier.
Is the emperor naked?
Richard Cole (richardc10)
Posted October 25, 2014 at 1:45 PM
@Gysbreght
Here is another idea for explaining the green curve.
I trialled fitting with heading and speed changes allowed at each arc (and minimising, to zero if possible, both BTO and BFO errors) with no useful results; the required speeds vary wildly. I have also tried adding noise as a sort of Monte Carlo but again not achieved anything useful.
The ATSB report update states the following for the data optimisation model:
“At each step, speed and/or heading values were varied to minimise the error between the calculated BFO of the path and the recorded value”.
This covers a range of possible models including constant speed courses with heading (course) changes that fit the BTO, as well as more unconstrained models. I note that the only detailed course we have is from the Ashton et al paper table 9 and that has only two speeds, plus heading changes at each arc.
I have looked at all the single speed courses that are consistent with a turn between 18.25 and 18.40UT and scored them against a BFO model that gives similar results to the Ashton et al table 9 (I did not use two-speed models, for my taste there is not enough data to justify adding another parameter to the fit). From the track and BFO models I can read off the lowest BFO r.m.s. error for tracks terminating in latitude bins on the final arc (I used bins of 0.5deg). Because tracks that are not consistent with the turn are excluded, this gives a slew in the error/latitude plot, rather a smooth trend with latitude. I converted those r.m.s. errors into probabilities using the Gaussian (normal) distribution and an BFO error of 2.4Hz (1 sigma, as I have used before), plotted those against latitude on the final arc and blended them with a Gaussian as described previously. As I stated, this is not a rigorous process.
The result (with the original ATSB plot) is shown below.
https://www.dropbox.com/s/qmx1yy2mrpn6hgm/green-curve.jpg?dl=0
The graph reproduces most of the features of the ATSB graph, though the peak at low latitude is not as prominent. The details are sensitive to the BFO offset and the maximum latitude of the final turn.
So in this interpretation the green curve presents the statistical balance of final locations of MH370 using single speed models, and folding in the implications of a turn between 18:25 and 18.40UT.
Hi Richard,
Thanks again for your elaborate and well explained reply. I think you basically agree with me, although you express it differently.
In your first paragraph you write that with your model you found “no useful results”. I find that difficult to understand, because it is actually quite easy to find paths that fit the BFO’s at each arc at reasonable speeds, for example 448 kt on average for FFB=149 Hz.
Do you agree that: “At each step, speed and/or heading values were varied to minimise the error between the calculated BFO of the path and the recorded value” produces a single path, and therefore the ‘best 100’ optimised paths are a selection of the paths produced by the error model of the report of 26 june?
Apologies for a copy/paste error.
Brock:
We have spent considerable time looking at the performance limits, focused primarily on the south western end of the arc. At some point, I’m sure we will include some more specific results in a report.
The problem is that the fuel limit or performance limit is a function of so many variables, like atmospheric temperature, altitude, TAS etc. that it is not that easy to analyze all the cases necessary to understand all the path model limits. As most understand by now, the limit does not vary simply as a function of end point latitude. Due to the complexity, it is a time resource management problem more than a physics problem. Plus, we have plenty of Boeing and engine manuals, charts, etc., but we lack the measured PDA numbers for the specific engines on 9M-MRO, many times requested, never released. So we can only get close with the data available. Still hoping ATSB will convince Malaysia to release the information so we can fine tune these estimates. All that said, we have looked closely at the southwest end, and we are in close agreement with ATSB on that end of the arc. Because the fuel limit crosses the 7th arc close to the point we derive independently from Inmarsat data, it provides additional confidence in the ~37-38 degree area, and less in areas southwest (too little fuel) or northeast of that point (excess fuel). It is not exactly orthogonal data, but it is a good proxy for one more dimension.
@airlandseaman: thanks, Mike – great information. Happy to hear the SW intersection is lining up, and eagerly (yet patiently) await publication of your detailed work.
I offer my services as the IG’s resident pit bull. Would Boeing have 9M-MRO’s measured PDA numbers? And/or Rolls Royce? If so, I’m happy to do the legwork required to get one of them to help you out, in the absence of the investigation’s (promised) transparency.
@JS
“@Will – I did the drift site. I could not find a point that drifted to Indonesia. It was always South Africa or the southern coast of Australia.”
Using my fat finger on the iPad, I did get a couple of sims with Indonesia, India and all around the NIO shores. But mostly E African and W & S Australian shores as per your results. Conclusion: Drift results very sensitive to initial conditions (or the size of my finger) as expected.
“But even if I missed place the duckie, I’m still not buying this. You are suggesting, I think, that someone waited to remind us where to look. The modeling isn’t that precise, though. If they were doing it as a reminder, I’d expect one every few months. It’s easy enough to say, “Hey, we’re waiting for X until debris drifts up.”
There is no(t much) point in reminding after 2, 3, etc months, since debris would still be in the middle of ocean. No one there to look.
“It’s a lot like missing persons cases in snow covered, frozen over areas. They don’t say, “Ok, start looking… NOW.” They say, “We’re looking, but we don’t expect to find much until the thaw.”
Apples and Oranges. Missing (deceased) persons in snow covered frozen areas don’t move (drift) about. Initial searches would happen, as did for MH370 via planes for debris, boats for accoustic pings, etc. until those means’ usefulness were exhausted.
When the thaw starts and the tourist are coming back, they would remind them, keep a look out now because the snow has gone. Or, in the case of MH370, look now, because some sims suggest that debris may wash up now in Indonesia.
Anyhow, I would have thought, they’d included Australia and Africa in that recommendation, given the drift simulations.
Cheers,
Will
MuOne – It might even be that they have some tags in the water slowly making their way towards Indonesia that they have been monitoring? Even if they did I have always been under the impression that the sea level on the north side of the Indon archipelago is about a metre higher than the south side because of the current that moves through it from the top, so 777 debris landing there would be a complicated matter? Those beaches are also pretty well littered to begin with.
But why not leave open the idea that it may show up in Australia/Africa as you say? And drift rates don’t just rely on current, there is the composition/buoyancy of the material and size of the object as well so I reckon I would have put some time frames up day one to have people ready. To me it’s a belated and odd sort of directive. If there is a fair degree of guesstimation involved then 7 months is a fair while to sit on it. Either they did indeed sit on it tactically or it’s suddenly come into the frame. I’m leaning towards the latter as I suspect they are becoming uncomfortable.
And it looks as if Mike and I are on opposite ends of the confidence-pessimism spectrum, but I won’t mind being wrong.
@Matty – your second question nails it. Why omit all of the other coasts suggested by the drift model?
I tried to reverse engineer a drift that ends at Indonesia. It appears to be possible with crash sites further north including the Bay of Bengal and the rough area of DG, though I’ve never subscribed to the latter.
Something interesting along those lines – if a piece had washed up somewhere, other than Indonesia, would they attempt to locate the crash site in reverse? One would think, sooner or later if there’s debris floating it will land somewhere. If the first landing occurred in Chile, for example, would that be used to try to identify the location of the breakup?
@Matty, JS,
The confidence in the current most probable SIO crash site is only now rising after many refinements of the ISAT data based path modelling.
It would be wrong to suggest that they could have created any useful forecast of flotsam arriving on shores anywhere back in March, April, May,…
The drift models are very sensitive to initial coordinates, so the earlier high level of uncertainty of where the plane ended up would have made any forecast early on a futile exercise. They needed to be careful, not to cry wolf.
Cheers
Will
@Matty,
“Either they did indeed sit on it tactically or it’s suddenly come into the frame. I’m leaning towards the latter as I suspect they are becoming uncomfortable”
I guess we agree somewhat, I am leaning to the latter too. However my interpretation of why is different ;o).
Cheers
Will
@Gysbreght
I don’t think we are doing the same thing to get the green curve but your idea is as good as mine. I used a single value of FFB in my simulation.
Single path only from the error minimisation model – that makes sense, though personally I couldn’t get a fit that was physical. Since Analysis C gave 5000 candidate paths something must be added, presumably a simulation of noise. With that idea I still couldn’t get candidate paths that stretched across the width of the green arc; Analysis C (June) and D1/D2 (October) show a narrow spread. The point-by-point minimisation of error concept is unappealing as it implies fitting the noise (we have discussed this before :)). The wording of the October report allows the possibility that the D analysis did not use point-by-point error minimisation – the Ashton et al track certainly does not.
Best 100: I agree that analysis A and B can be used to generate a paths and scored against the BFO model. Picking the ‘best 100’ does not have any useful meaning and I don’t believe that idea was used in generating the red and green curves. Those curves used a scoring of statistical significance of the candidate tracks.
Inmarsat’s ‘Example Reconstructed Flight Path’ compatible with Autopilot constraints? -That path has a variable speed and track at the start, but from 20:41 onwards it appears straighter with constant TAS. I don’t think there is a statistical basis for using two speeds and it doesn’t have much impact on the final range of candidate destinations. However, there must be pressure to include an autopilot compatible mode for at least part of any proposed track since curved tracks are criticised for being inconsistent with the way large aircraft fly.
Hi Richard,
It is certainly interesting that you get a probability distribution curve similar to that of the ATSB by using entirely different assumptions.
It seems to me that you are making several statements that are in conflict with the explanations given in the two ATSB reports. But I don’t want to appear to be nit-picking nor do I want to engage in a ping-pong game of yes,it did say that, no it didn’t.
Therefore I will limit myself to your last paragraph. A large aircraft will fly in any way that a pilot wants it to fly, within its performance and handling capabilities and structural limits. A large part of a NORMAL flight will be on autopilot. In NORMAL flights the autopilot is controlled by the FMS most of the time but can be controlled by the pilot if there is any change of the flight plan, f.i. in this flight the ATC instruction Direct-To-IGARI. The airplane diverted from the FMS flight plan at IGARI, so if it was flown on the autopilot after the diversion it is likely that the autopilot was controlled by the pilot rather than the FMS. Between IGARI and 18:40 UTC the autopilot selections were obviously changed several times. At 18:22:47 someone changed the electrical configuration of the airplane’s systems. So what is the latest time that someone could have changed the autopilot selections, and what evidence or logical argument supports that?
@JS: yes, they’d try to reverse engineer (though they might not find a unique solution: distinct “multiple streams ending in one river” possibility).
Agree: why ONLY Indonesian shores? Very curious.
On a hunch: I plugged in the “most likely” coordinate for the Curtin U. event (I know it was subsequently downplayed, but a hunch is a hunch), and got this:
https://twitter.com/Brock_McEwen/status/526417488603795456
I threw out an accidental shoot-down scenario here a month ago. I’ve also wondered about the possibility of a DEFENSIVE shoot-down, where the plane was (thought to be) targeting either DG, or Will & Kate (vacationing in Maldives March 8). This ATSB directive is consistent with that scenario.
Gysbrecht,
I believe it’s more likely that the flight continued with the AFDS in LNAV mode, flying a route defined in the FMS. The route is simply a series of waypoints, these may be nav db waypoints or even lat-longs, very simple to enter & build the alternate route. I recently reviewed the ‘Beijing Lido’ radar track again & that evidences a turn at a position on the edge of the WMD-413A military manoeuvres area, called out in the related AIP, towards another waypoint on the edge of the KL FIR.
@GuardedDon, do you think that all of the published post-17:21 primary radar tracking of MH370 was Thai, and if so, what is your opinion as to why the Malaysians have not published any of their own primary radar tracking of MH370?
Thanks!
@GuardedDon,
sure, very simple, if he had plenty of time and nothing else to do, just type the identifiers in the keyboard and press a few buttons for each waypoint. It’s even simpler to turn a knob and press it, and that can still be done at any time.
@Luigi:
Re your question – sorry for the delay. My cousin and I have been playing phone tag. He’s en route to Europe now, so we’re planning to talk mid-week.
From his VM today:
“If the Malaysian plane is in the ocean, where is the debris?”
All they had to say regarding debris:
We may be in need of some vigilance from the public in the coming months as eventually debris from this plane will show up somewhere. We don’t know exactly where at this point or in what timeframe either.
@Matty, re: “all they had to say regarding debris” – who is “they”? Do they trump the current ATSB MH370 site’s “debris” paragraph? From today:
“The ATSB continues to receive messages from members of the public who have found material washed up on the Australian coastline…but drift modelling undertaken by the Australian Maritime Safety Authority has suggested that if there were any floating debris, it is far more likely to have travelled west, away from the coastline of Australia. It is possible that some materials may have drifted to the coastline of Indonesia, and an alert has been issued in that country, requesting that the authorities be alerted to any possible debris from the aircraft.”
A keyword (drift/ed/ing) search of amsa.gov.au turns up several articles – the most recent from Dec. 2013. I’ve asked AMSA for details – the exact coordinates of their starting (Mar.8) debris field in particular.
The ATSB is tagging Indonesia as a likely “first shore” for debris originating at “s38 & 7th” (IG=ATSB current best estimate crash site).
On the weekend, I tweeted a drift.org.au screen cap suggesting the opposite was true: W.Oz was likely; Indonesia was NOT. But perhaps I chose a poor site, or used a good one badly. I eagerly invite experts in this field to weigh in.
The only other sources I checked (also posted to this blog) were SIO pilot charts which suggest that, if AMSA’s drift model sent debris west from s38 in March/April/May, it did so against prevailing currents and winds.
@Brock. I think the ATSB is just trying to say they don’t have the resources or inclination to respond to everyone in Australia who calls to say that they found some rubbish on the beach.
FWIW, I agree that the debris would have drifted east, but unfortunately they gave up too soon trying to retrieve any of what the satellites spotted.
@Bruce: well-argued, and quite possibly the true underlying motive. I, too, read “stop calling us, Mrs. McGillicuddy!!” between the lines.
But the statement: “far more likely to have drifted west” (from s38) – according to sources and opinions (including yours) thusfar presented – is not true; I can’t in good conscience ASSUME it is a “WHITE” untruth. The event being investigated is far too grave for such an assumption.
Just like the proven untruth underpinning the March search move closer to shore (“flew faster, therefore move from S2->S3”), the Australian government needs to come clean on exactly WHY it is untruthing.
“We’re cutting corners on the search to save a few bucks” is by far the LEAST concerning on a long and chilling list of possible reasons for deliberately deceiving the public. Why wouldn’t they just come out and admit this, then, and eliminate suspicions of complicity in a cover-up?
[Also published on Twitter — with ALL URL’s:
http://bit.ly/1zdV5ys%5D
On March 10, 2014, Stéphane Berthomet, a former counter-terrorism officer based in Quebec turned author and blogger, was told by a confidential source that despite MH370’s disappearance from radar, the aircraft would have continued to give its position for several hours.
As Berthomet relates, armed with that tip and Inmarsat’s statement that it gave its MH370 data to SITA (which in turn, gave it to Malaysia Airlines), he approached SITA directly. SITA did not confirm Berthomet’s tip — but it didn’t deny or contradict his information either.
Per the background piece above, on March 13, 2014, three days after Berthomet’s tip (and two days after he first wrote about it on his blog), WSJ published an article by Andy Pasztor and Jon Ostrower claiming the same: MH370 “Flew On for Hours”.
Note the interesting difference between the two: while Pasztor and Ostrower subsequently retracted their story (because Boeing and Rolls Royce strangely, denied it), Berthomet is standing by his — and confirmed the same via tweet:
So again, this all begs the question: why did the WSJ reporters change THEIR story?
Is it possible (I think so) that Pasztor and Ostrower’s unnamed source is the SAME person who gave the tip to Berthomet? Did Pasztor and Ostrower later retract because they decided that their (unnamed) source was not credible, or because using an unnamed source made their piece weaker (let’s note that US media outlets serially publish huge stories relying on ‘unnamed sources’)? Or was their source someone inside SITA? Is it possible that Pasztor and Ostrower were planning to use the ‘flying for hours’ story from the unnamed source to corner Boeing and Rolls Royce into confirming that their data indicated the same? If so, that clearly backfired, because Boeing and Rolls Royce denied that they received any data after 1:07 am (17:07 UTC). And as Berthomet points out, Hishammuddin Hussein cited their denial in a tweet to do the same.
It’s all very confusing, isn’t it? But I hope that people get this: Boeing and Rolls denying that THEY have ACARS data from MH370 after 1:07am does not mean that the plane didn’t fly for hours after that time. Nor does it mean that Boeing and Rolls Royce are unaware of OTHER (likely, proprietary) information indicating that the plane flew on for hours. What it does mean though, is that to the extent MH370 flew on after 1:07am, Boeing and Rolls Royce are clearly saying, “you didn’t get that from us.”
And that’s exactly why I believe (and tweeted) that Boeing and Rolls Royce’s denial of the WSJ story — and WSJ’s subsequent retraction — should be a red flag.
Nihonmama:
I don’t see the red flag here. People confused “flew on for hours = GPS position data in ACARS Messages, or the like” with simple handshakes – “is was somewhere in the air”. Once it was clear what everybody meant, it was easy to see what happened with all the press errors. People were saying all kinds of non-sense, like the engines had their own transmitters, which was not true. All engine data went through the ACARS App and the Inmarsat hardware. It is no surprise that people unfamiliar with how all this works got very confused before the facts got straightened out. You will have to ask Andy why he did not simply clarify, rather than retract, but Boeing has no data, and that is 100% consistent with “the plane flew on for hours” (based on handshakes BTO/BFO data, but not containing any position data from any nav source).
Nihonmama
Posted October 24, 2014 at 8:32 PM
“How does Angus Houston, who’s sitting in Australia, *know* that “investigators will never be able to connect MH17 shootdown to “a particular group”?”
Via @PAIN_NET1 today:
“can’t have Houston he is not so obliging when it comes to the cover up of the duck up”
http://t.co/xppPlaOxrg
@airlandseaman:
You said:
“It is no surprise that people unfamiliar with how all this works got very confused before the facts got straightened out.”
Mike, please expand, if you will, on who you think got it confused:
1) Stéphane Berthomet
2) Andy Pasztor and Jon Ostrower
3) Boeing and Rolls Royce
4) SITA
5) All of the above
6) None of the above
7) OTHER
Bruce/Brock – without public vigilance they will never get debris and AF447 bits were showing up two years later. Taking the OZ coastline off the table reveals a certainty that they don’t really have. Being tied up with false leads would be a continuation of the original sea search….wouldn’t it?
@Nihonmama. I believe in the aftermath of the disappearance of MH370, there was a combination of confusion and disinformation regarding the satellite data. For instance, retired Gen Tom McInerney insisted that the plane had landed in Pakistan as relayed to him by his military sources. Supposedly, Boeing had given this information, gleaned from satellite data, to US intelligence.
McInerney in August was claiming that these same sources were saying that MH370 could reappear in an attack on 9/11 and he wanted the US to go to a DEFCON 1 alert level. He also said that “On the seventh of September, a major news network and publishing network are going to put out a book. It is going to be an earth shattering [account] of what’s happening and what happened. The fact is we may even see a 9/11/14 MH-370 surface again.”
Well, McInerney was dead-wrong about his predictions regarding MH370, including the release of the tell-all book. Assuming that McInerney genuinely received this information from his military/intelligence sources, this indicates that there is a faction that is deliberately leaking false information about the satellite data.
@LGHamilton and I have tweeted to Fox News host Uma Pemmaraju asking her to follow up with McInerney to better understand why his sources were leaking false information. We have not received a meaningful response. Only there are “issues with his intelligence sources” and he is “not ready to go public yet.”
https://twitter.com/LGHamiltonUSA/status/503270872212570112
Perhaps Monsieur Berthomet was a victim of this same disinformation campaign regarding the satellite data.
I have not heard one technical expert claim that GPS coordinates are transmitted as part of the satellite handshakes when ACARS is disabled. Many people confuse the ACARS data with the underlying transport layer for the satellite communications. It is hard for me to believe that Inmarsat personnel, including the engineers analyzing the satellite data, would participate in such a grand deception. I cannot say there is zero probability, but I would say it is extremely unlikely.
Victor
@VictorI
“Only there are “issues with his intelligence sources” and he is “not ready to go public yet.””
On the basis of the evidence above, in particular the latter, I suspect McInenery’s sources may be connected to J. F’n Tino.
Cheers
Will
@Victor
Taking this man the LEAST bit seriously is a grave flaw in judgement, with all due respect.
The only ‘disinformation’ being dished out by Mcinenery was furnished and dreamed up by him and him alone. He used the platform that was somehow (lord knows how) granted him to push a political agenda in the most fraudulent and disingenuous of manner. It was downright disgusting, and outright mendacity of the worst order.
Come on folks…nary a word further should be wasted on anything this individual has to say about MH370.
Cheers
Victor:
I’m at a loss to understand why, rather than respond to the essence of the post, you chose to conflate Berthomet with (Lt.) Gen McInerney — a FOXNews talking head, I might add (although, it’s quite interesting that he’s said repeatedly that MH370 is in Pakistan and no one’s checked him.) I’ll just say that if someone wants to argue that Monsieur Berthomet is dead wrong, then please make THAT case. Because comparing Berthomet to McInerney is not persuasive.
Moreover, if Monsieur Berthomet was a “victim of a disinformation campaign regarding the satellite data”, let the record reflect that his tip appears to be identical to that which Pasztor and Ostrower received (from somebody), reported and then retracted. But I don’t recall that you (or anyone) suggested that the two WSJ reporters were victims of a disinformation campaign. If they were, they sure haven’t said it (although they could). They simply retracted the ‘flying for hours’ piece without comment. And guess what? Something (still) doesn’t wash. But I don’t see anyone in the aviation press drilling into the reason for that retraction.
I did notice that you and others have already engaged Monsieur Berthomet on Twitter. It appears that he’s not only not giving up more info over the Internet (as a former counter-terrorism officer – that makes sense), but he also told @cryfortruth: “I can assure you that my source can not be the same as the @WSJ”. Even more interesting. How can Monsieur Berthomet say with certainty that his source is NOT the same as WSJ’s – unless, perhaps, he has additional — or better — information? So yes, I agree that there are a victim(s) all up in this mix, but I’ll also proffer that it may not be Monsieur Berthomet. Perhaps you’ll ask him.
As far as GPS coordinates being “transmitted as part of the satellite handshakes when ACARS is DISABLED”, I can’t answer that.
But I did receive information from a former colleague (who’s in a position to know) on this specific issue back in May:
https://twitter.com/nihonmama/status/464830613493526529
I went to every person I know in aerospace to try and confirm this, but hit a stone wall. That told me something too.
Rand,
I was just pondering the Des Ross article. It’s spot on. Most of what he is talking about I think Mary Schiavo nailed right from the beginning, that people may tend to be asleep at the switch at that hour, especially if the military and the air traffic control are stationed within the same edifice and not getting up off their derrieres and talking to one another. I would imagine us not hearing about if any jobs were lost or heads rolled due to all the missed red flags would be part of the saving face mentality and not divulging anything other than what is in the Preliminary Report. Although, didn’t someone say and I believe it could have been HH right at the beginning that they admitted they were sloppy in the beginning but in the end Malaysia would be viewed as doing the right thing, or in some such verbiage as that? I agree those tapes would indeed be something to add to the Wish List of desired information from the official investigative team, but it’s doubtful they will ever see the light of day at least not to the public.
Going back over the Preliminary Report what is baffling is MAS OPS. It does not say they (MAS) contacted Cambodia, it says they were looking at a projection of the flight not suitable for tracking. It was HCMATC who ultimately contacted Cambodia for verification if MH370 ever entered their airspace. It’s an overwhelming amount of incompetence, did they just flat out miss it, were they incredulous to the fact that something like this could ever happen and figured they have to be up there flying en route to Beijing, or was it something more severe than that, who knows. And there is still the matter of whether or not the military saw MH370 during their transponderless traversing in real time or on a loop delay and were the jets scrambled after the fact or not? Do we even know that yet?
It seems KLATC took the word of MAS verbatim and did not challenge them enough knowing full well that MH370 had never been heard from by HCMATC. And didn’t MAS tell them they established signals with the flight, KLATC bought that, I guess they had no recourse not to. I agree a lot could be learned from those tapes, probably a lesson in what NOT to do in an emergency, but whether or not any clues can be gleaned on what happened to the flight or where the flight is now I don’t know. Those tapes and the tapes of MH88 and the other flight asked to contact them in the Prelim. Report would be very interesting to say the least.
Cheryl – all very South East Asian to me. For the very same reasons I had no problem with the plane making it across India had the “data” allowed.
Debris – nearly 8 months later, is the absence of debris so far on Australian beaches the reason why they are not on the watch list? If it was going to land here we would have known by now?
@Nihonmama: Regarding the WSJ story, I think the reporters simply got the technical facts wrong, either through their own misunderstanding or through something misunderstood by their source. Once they learned of their mistake, they corrected themselves with a retraction. I do not see anything mysterious here. I agree with @airlandseaman regarding the confusion in the early days after the disappearance.
I chose to talk about McInerney because he is an example of somebody with a similar story about the satellite data that has been proven wrong. I don’t dismiss him outright because he is on FoxNews. I dismiss him because his predictions have not come to pass. (I would give a talking head on MSNBC the same due.) And I questioned whether the same erroneous source has fed information to others, like Mr. Berthomet, who is sure that his source is different.
I believe there is confusion between ACARS, engine monitoring,and the SATCOM link. It may be that even if the airline has not subscribed to the engine monitoring service, there is performance data that is still sent over ACARS to the engine manufacturer, which in this case is Rolls-Royce (RR), hence the belief that RR received data even though MAS was not a subscriber to the service. This might have caused some of the confusion about the “engines transmitting the data.” However, with ACARS disabled, I don’t see how RR could have received this data.
I will try to learn more from Mr. Berthomet in attempt to reconcile disparate facts in this case. I am not sure that he will be willing to share more than he has already written, but I am very willing to hear and digest as much as he is willing to share.
I will say that if Inmarsat really does have the GPS coordinates of each handshake, and this knowledge becomes public, it would tank the stock price of this publicly traded company. It is hard for me to imagine that the company would take this risk. Also, it would mean that its team of engineers and scientists are part of the cover up.
I think that marketers and corporate officials are capable of distorting technical facts for what they believe is for the good of the company. I have a hard time believing that a group of engineers and scientists would be asked to publicly do the same. At the most, they would be asked to remain silent. Instead we have technical personnel making public statements, appearing on camera,and submitting a technical article to a journal publication.
Victor – Engine data – truck engines have their own transmitters. Caterpillar, Mack, Deutz, Cummins, Perkins Diesel etc. All these trucks are hooked up and feeding. Why go though ACARS at all? I accept what you say but I guess it’s another one of those questions that MH370 has thrown up. The biggest turbo fan engines ever built don’t have their link?
@Matty: The cost, weight, drag, and complexity of each engine having its own data processor, RF electronics, and antenna system would be prohibitive with little benefit. The system would have worked fine as designed had an operator not disabled the ACARS.
Matty – Perth
I think you are confusing “transmitters’ with “data collection platforms” (DCPs). The RR engines, GE engines, and dozens of other devices on cars, trucks and aircraft have DCPs, but they do not generally have their own dedicated Radio Frequency Transmitters. This is definitely not possible with Inmarsat Transmitters (AKA “AES”). It is impossible to put an AES on an aircraft engine. Instead, the data collected from engines is routed via a copper or optical connection (ARINC 429 typically) to a computer in the equipment bay where the data is encapsulated in an ACARS message, which is then sent via another ARINC 429 connection to the same Inmarsat AES used to carry cockpit voice, cabin telephone, ATC and other traffic.
@VictorI,
I believe the facts are that MAS did not subscribe to a Boeing “airplane health monitoring” service, but did subscribe to a Rolls-Royce “engine performance monitoring” service. The latter is presumably part of a commercial agreement between the engine manufacturer and the operator, under which the former provides guarantees to the latter regarding engine maintenance costs and fuel consumption. At specific moments in a flight (i.e. not continuously) ACARS will aquire engine performance data, assemble them in a report, and transmit the report via SITA to Rolls-Royce, using VHF or Satcom communication systems, as appropriate. If both communication systems are unavailable, ACARS will hold the report in abeyance and send it when communication is restored. If at that time there are other messages to be sent, ACARS will transmit them in an order of priority, as described in detail in the AF447 accident investigation reports.
To my knowledge, ACARS can not be shut down from the cockpit, but can be ‘silenced’ by cutting the communication link. That has presumably been done, since no ACARS messages were sent when the Satcom system was re-powered, although at least one message was scheduled to be sent prior to that time.
Gysbreght Posted October 28, 2014 at 9:03 AM
This is consistent with my understanding as well. RR paid for the data, or had MAS pay for it pursuant to some contract between them. RR got 30 minute data, either via VHF, HF, or Inmarsat ACARS, or in some cases, downloaded at the end of the flight. When ACARS transmissions stopped (for unknown reasons), RR got no more data (after 1707).
@Gysbreght: Thank for making the distinction between airplane health monitoring and engine health monitoring. I agree that ACARS can be effectively disabled by disabling all three communications platforms (SATCOM, VHF, HF). In any event, our current understanding is that there is no GPS data without ACARS messages. It would be earth-shaking if we were to learn otherwise.