Guest Post: Why Did MH370 Log Back on with Inmarsat?

[Editor’s note: One of the most intriguing clues in the MH370 mystery is the fact that the airplane’s satcom system logged back on to the Inmarsat network at 18:25. By understanding how such an event could take place, we can significantly narrow the range of possible narratives. In the interest of getting everyone on the same page in understanding this event, I’ve asked Mike Exner for permission to post the content of a detailed comment he recently provided.  One piece of background: a lot of us have been referring to the satellite communications system aboard the aircraft as the “SDU,” but as Mike recently pointed out in another comment, it technically should be called the “AES.” — JW.]

Until we have more evidence to support the theory that the loss of AES communications was due to the loss of primary power to the AES, we must keep an open mind. Loss of power may be the most likely cause (simplest explanation), but the fact is we do not know why the sat link was down between 17:37 and 18:25. My reluctance to jump to the conclusion that it must have been due to the loss of primary AES power is based on decades of experience in the MSS (mobile satellite service) industry. It’s not just another opinion based on convenience to support a theory. Let me elaborate on a few possible alternative explanations.

The potential for loss of the pilot carrier, due to the orientation of the aircraft in relation to the satellite, was increased as soon as the airplane turned WNW. Between the time of this turn (circa 17:50) and the time of the FMT (final major turn circa 18:25-18:40), the aircraft was flying more or less toward the satellite where the antenna pattern was near a null. Don and I have both looked at the antenna pattern in some detail and concluded that the antenna pattern and coincidental direction of flight were unlikely to be so bad that the pilot carrier would be lost due to this geometry. Moreover, according to a MAS Press Conference on March 20, 2014, there should have been an ACARS message transmitted at 17:37, but none was received. ( bit.ly/QFbF6C ) At 17:37, the aircraft was still over Malaysia SW bound, so the HGA pattern would not have been an issue at that point. Taken together, loss of the pilot carrier due to antenna orientation appears to be a possible, but unlikely explanation for the outage.

Ionospheric scintillation has also been suggested as a possible explanation for the loss of service during this period, but there have been no reports of other aircraft in the vicinity suffering a loss of service, so this explanation is also unlikely. (Note: Ionospheric scintillation in the equatorial regions can be a big problem for VHF and UHF communications, but it does not affect communications in the L band as much.)

The MCS6000 AES, located in the back of the airplane, requires a continuous feed of INS data (position, speed, etc.) via an ARINC 429 link from the computers in the front of the plane. If the AES stopped receiving INS data for any reason, then it would not have been able to steer the HGA, or compute the required Doppler corrected transmitter frequency. Thus, it is very likely that the AES would be out of service if there was any loss of this 429 data link, or the information carried over the link. Given that there was no VHF voice communications after 17:19:24 and the Transponder Mode S data was lost after 17:21:13, it is certainly possible that the INS data flowing to the AES was disrupted due to a common failure in some piece of equipment in the E-Bay. This explanation for the loss of service cannot be dismissed as easily as the two previous theories.

However, there is one additional observation that tends to favor the loss of primary power theory over the loss of INS data theory (or the other two theories above). We note that when the AES logged on at 18:25:26, the BFO values for the first few minutes thereafter appear to have been drifting in a way that is more consistent with a restoration of primary power event than a restoration of INS data event. If the AES power had been on during the outage, the oven controlled reference oscillator would have maintained a stable frequency and there should not have been any significant BFO transients following the 18:25:26 logon.

In summary, there are multiple alternative explanations for the AES outage, but loss of primary power is the most likely explanation. Like so many other necessary assumptions, like the mode of navigation after the FMT, we have no choice. We must base the search on the most likely assumptions while maintaining an awareness that few of the assumptions have probabilities of .999.

637 thoughts on “Guest Post: Why Did MH370 Log Back on with Inmarsat?”

  1. @Nihonmama: Yes, both “technical folks” and “street detectives” will be required to solve this incident. One basic problem is that we have little or no detective work in the public arena, while there has been an abundance of technical work that has produced major areas of agreement (although not 100%). Sure, there are some non-technical folks that conjecture. However, much of the non-technical evidence remains anecdotal and incomplete, such as the Maldives sighting.

    I wish there was more energy directed at detective work. That would produce more evidence we all could work with and would validate or invalidate many of the assumptions about which we bicker, such as the timing of the turn south, or if it even happened.

    One of the few pieces of substantive investigative reporting I have seen is the Four Corners interview with Hishammuddin Hussein, but that was back in May, I believe. He was forced to answer (actually evade) questions about why the RMAF did not intercept MH370 as it crossed the Malay peninsula. Who else is in the media are asking questions like this?

    Absent a find in the SIO, or a whistleblower, witness, or leak, it will be difficult moving forward in a meaningful way, and we will be reduced to arguing amongst ourselves over questions that cannot be definitively answered.

    Many of us suspect that the official investigation, led by Malaysia, is tainted. So rather than bash the technical folks, ask yourself why there are no independent investigators trying to solve this mystery. We should strategize about ways we can get more data into the public domain. And produced more evidence that the technical folks can work with to validate or invalidate particular scenarios.

    Here is a short list of data that would be helpful:
    1. All raw radar data from all sources.
    2. All ATC communications, including between ATC centers and between KL and the RMAF.
    3. Surveillance video of the plane while on the ground at KLIA.
    4. The complete, unredacted satellite logs, including the payload data (not just the signaling data).

    All of this data is available today, but has not been released.

  2. I will not be surprised if nothing is found in the current search area, which is based on, at least, two questionable assumptions:

    1. AP.
    2. The time of FMT.

    It appears that the trajectories based on the “free flight” dynamics can also fit BTO/BFO data reasonably well. So far I reached <3.5 Hz BFO residuals, and <10 km BTO-equivalent residuals, and still improving.

    In brief, the airplane is forced by the:

    1. Gravity,
    2. Thrust,
    3. Lift,
    4. Coriolis, and
    5. Wind drag force.

    The tricky part is to realize that the heading of the aircraft and orientation of the fuselage are two different things. The wind drag is also not that simple term as it may seem to be: it has to account for the impact of cross-wind. Also the mass of the aircraft has to be directly modelled.

    I took the approach to find unknown coefficients by the minimization of BTO, BFO and mass residuals functional. Interestingly and coincidently, I am getting quite reasonable output coefficients.

    A number of interesting observations:

    1. Despite a large number of 3D environmental forcing, which is sourced from ARL NOAA GDAS, 1 deg x 1 deg resolution, interpolated, there is still a class of solutions that fit BT/BFO reasonably well.

    2. The predicted location of the aircraft at 19:41 is ~150 km south of Kate's yacht at 19:25, at 4.2-5 km altitude, subject to parameterization of various terms.

    3. The terminal location mostly converges to around 100E, 28S, which is coinciding with the original high-probability ATSB area. Moreover, as I see, Fugro has even conducted the extended-width survey there, but left some gaps in that area.

    4. The second phone call BFO residual is "in line" with other BFO residuals. No need to 'invent' different BFO bias for the respective channel.

    5. Coefficients found in a result of the minimization are fairly reasonable. Coincidence?

    6. Low altitude (4-6 km) explains the lack of radar tracking by Indonesians.

    7. The terminal location is consistent with the origin of the sound recorded by Curtin University if one considers it on the 7th arc (assuming it was the sound of the seabed impact, about 1 hour later).

    Due to lack of time I cannot post comprehensive details of the model right now (system of ODEs), but will try to do it in 2 weeks or so. I can copy/paste trajectory 19:41 to the end if somebody is interested in this scenario. The major problem remains how to connect 18:22 and 19:41 (4-5 km altitude) positions.

  3. @Victor:

    “rather than bash the technical folks”

    A “bash” was not the point of the writing.

    “ask yourself why there are no independent investigators trying to solve this mystery.”

    I don’t know that there aren’t independent investigators trying to do that. I used the qualification “crowdsourcing perspective”.

    @Chris Butler:

    Re documentary time:

    21:30 local (AUS) on Sunday on Seven

    Another link with info:
    https://t.co/Pv1QCifARR

  4. Hudson Posted February 14, 2015 at 3:35 PM:

    Hello Hudson ~

    Welcome to the board. Your post is delightful. Nice to meet you!

    @Ron [Black] explained (on his Google+ page), “It was easy to call MH370 by satellite phone, just dialing:

    +870 5 35200217

    where the phone number “35200217” is the ICAO 24-bit Aircraft Address [AES ID] of 9M-MRO in octal notation.”

    It’s my understanding that the two attempts via satphone did not complete the circuit to the point of an audible ring in the aircraft.

    ~LG~

  5. Hi Lauren,

    On the 70 cm figure; I suppose this is to some extent a PR opportunity for Fugro and they naturally quoted the cross track resolution for the tool, which is the resolution perpendicular to the direction of travel. That’s the easy direction, it is stitching together those lateral slices and stacking in the direction of travel to make an image which is trickier. The resolution along track (i.e. the width of each lateral slice) is usually poorer (even with synthetic aperture techniques) and will very likely be worse than 70cm, especially at the edges of the swept area.

    Slightly academic as it should be good enough to spot an engine, unless the terrain is rugged enough to put such an object in shadow.

    Would be very nice to see some more images from the search, but I suspect they are afraid of the Tomnod Effect, no doubt the phones would start ringing as people started seeing all kinds of objects on the sea floor.

  6. Jeff: Thank-you for your expert guidance and supervision. I completely agree with your comments on the past performance of the CH7 reporter and ATSB. Hope the latest in both cases is improved.

    Hudson: Thanks for the acknowledgement and kind words.

    Victor: I promise to take your recommended medicine next time.

    Brock: Your arc width probability spreadsheet is a nice piece of work. I see a lot of thought in setting up the model, but I could not follow every detail. I’m not sure how to model such a complex set of possible paths much better in a spreadsheet. I’ll just make some general comments about the considerations, based primarily on the 777 simulations.
    1. It is very likely that the plane started turning within 1-2 minutes after fuel exhaustion, and ended within 4-8 minutes.
    2. It is virtually certain that the turn radius, bank angle, speed, ROC/D, and AOA all changed considerably, over the descent, increasing and decreasing and repeating…and not in a smooth continuous way.
    3. Phugoids probably started and stopped 1 or 2 times while the turning was in also going on…very complex.
    4. The bank angle may have been as low as 25 degrees a good part of the time (larger radius) and as high as 80-90 degrees near the end (small radius).
    5. It would not be surprising to see it roll over inverted in a near vertical spiral dive near the end, in which case some breakup before impact could be expected (like loss of some control surfaces and landing gear doors, etc) See Chinese Air 006 case for a relevant real life case (https://www.youtube.com/watch?v=E__FDKNbczs). But this is just one of several possible ways the flight could have ended. It may have impacted the water at high speed with the wings near vertical (70-90 degree bank), but not so much nose down. We observed just such a case in the simulator. Many other possible endings, but all high energy impacts, which is probably the main reason why debris has been difficult to locate.
    6. It probably completed no more that 2-3 complete 360 degree turns, if that many.
    7. At the time of impact, the heading could have been anything (0-360 degrees).
    There’s more, but I think this is enough to illustrate the difficulty of modeling the probability problem with a simple spreadsheet model. If I had to weight all these complex issues and possibilities in my hand dandy right brain computer, I come out with a probability distribution that peaks near the N-W and S-E sides of a -20/+30NM band, not the middle. The peaks would not a lot greater than the middle…maybe twice as high as the middle. The probability would drop off fairly rapidly outside the 50NM width, and reach zero around ±100 NM (corresponding to a best glide straight and level). But the integrated area outside the -20/+30 NM band is 95% inside the nominal band.

  7. Lauren – It was JS who made that beer can quip but I get you, I had one as a souvenir, now gone.

    Mike – don’t be deterred ever. To me you look married to those numbers and the crunchers can be obsessive but they have framed a lot of what we discuss here. We all articulate things differently and I was actually used to the way John Fiorentino went about it in the end. I’ve been smacked down with contempt but my eyes are forward. Just a note – trying to type with an English Bulldog pup on your lap is bloody hard. I take my hands off her and type and she (Phoebe) attacks the keyboard.

  8. The last sentance in the previous post was somehow trashed in the editor. It shuld read:

    But the integrated area outside the -20/+30 NM band is 95% inside the nominal band.

  9. Jeff: The same thing happened again. Some how, part of the sentence is lost when I post it.

    I’ll retype it here from scratch:

    …outside the -20/+30 NM band is 95%.

    Hope it works this time.

  10. Jeff:

    Something weird is going on. I’ve tried 3 times to post this sentence and every time part of the sentence is lost. Must not like some characher, or thinks I am using HTML codes.

    trying to say….less than 5% outside the nominal band and more than 95% inside the nominal band.

    Mike

  11. Victor – I agree the Maldives thing is tantalizing but one problem is the way the authoritarian govt over there jumped on it. They arrest people for anything as it is. Ditto the lack of an acoustic detection from the seabed devices here in Perth and Cape Leeuwin. Submariners say such an event would be audible from long range but nothing, and they seem pretty well situated?

    Jeff – the older we get the blunter we get – the crankyness is often mellow hearted. The beers however is a pretty good concept. It’s not yet 9.00 am here but the breeze is drifting through the gum trees that hang around, and the grass is cut as the Cockatoos lope by with casual wing beats and the veranda is empty, so the scene is set.

  12. @All

    Hmmmm….?

    Mike-X & JW have delivered, along with the IG group, have delivered some of the BEST info going. Not to mention the rest of the group.

  13. @ALSM

    I was perplexed by the same thing only to find out that my Greek Spy wife was sneaking & answering & or deleting my thoughts regarding MH370 . I was “Obsessed” & needed a break.

  14. @nihonmama: The “rather than bash the technical folks” statement was not meant to be directed specifically to you, although it did read that way. I am agreeing with the thrust of your argument that the “street detective” work should be valued. I am also lamenting that I don’t such much substantive detective work occurring. In fact, the technical work is well ahead of the street detective work. If there were more data for the technical folks to work with, we might have more agreement among us.

    And as far as I can tell, Malaysia is doing a much better job at burying the evidence than any independent investigator is doing at uncovering it. Part of the burying of the evidence includes using the Malaysian police to “investigate” (code word for “intimidate”) any accusing the government of engaging in a conspiracy. The fact that this is being done so openly indicates to me that the Malaysian government is deliberately sending a strong message that an open discussion about MH370 will not be tolerated.

  15. @Nihonmama

    I am not frustrated in the least, and I do appreciate your concern.

    Getting banned for asking (perhaps not politely enough) for an IG member to provide the compelling evidence he claims to have, is just something I am certainly willing to accept.

    The assumptions relative to flight dynamics after the FMT are absolutely central to the validity of the SIO hypothesis. It is important the we (the collective) have a clear understanding of the basis for the assumption. If it proves to be incorrect, it changes everything.

    In any case, my posting here is finished unless I have a major eureka experience that is worth sharing.

  16. Hello,

    I was reading the comments and noticed that gentleman is having a problem making a commentary post.

    I began to read to read a little further back and also noticed that same person is swamped in hypotheticals in the SIO and claims them to be a fact and provided a list of facts?

    Correct me if I am wrong, but isn’t a fact an event that is indisputable and shown as to “Actually” occurred versus presuming that it occurred? Thats how it is interpreted in the High Courts, so why is it a fact and not?

    Oh, I see now as it must have never been a fact to begin with and still being sold as fact a peddled into world and misleading all of the people who really want the facts.

    Thank you for your time and your missing post may be in the SIO too..

  17. @dennisW – please do share any eureka moments that do come along. Agreed that the assumptions for the FMT need further clarity …

  18. In watching the link above. It’s mentioned that the satellite was running out of fuel & for the lack of another term wobblie? This could have a measure of miles of distance,couldn’t not? The BFO & BTO data could push the location further from the 7th arch toward Diego Garcia.

  19. @jeff
    >How on earth could their confidence be growing? That seem outlandish to me.

    I can’t speak for what the ATSB investigators told the Australian journalist but it would be a consistent position. There are two quite separate issues. The first is the signal analysis to define the core search area (7th arc and which part of that arc), the second is the aircraft performance analysis to assess the crash site position with respect to the core search area. The investigation may have increasing confidence that the core search area is correct. That the aircraft has not been found could just mean not enough area has been covered and/or the performance of the aircraft took it further away from the 7th arc. Since we know almost nothing about what happened onboard, wide error bars from this part of the analysis are unsurprising.

    On another point, discussion of the signal analysis on this forum has concentrated on the autopilot models which is where the IG have concentrated (at least from published materials). From the second ATSB report (and my analysis amongst others) the majority of the search area follows from the data optimisation models which do not assign high weight to the time of the FMT or to the radar data. The excitement has been in the South with two (now three) Fugro vessels. The Northern area is less well covered. I would not be surprised if another season of deep tow scanning is needed.

  20. A few questions to pilots (I hope they will not be banned on the basis of incompatibility with the AP hypothesis):

    1. What steps would a descent procedure include in order to land in Maimun Saleh airport starting from NILAM, 9-10 km altitude (app. 200 km away)? I am interested in both AP and AT modes.

    2. What will happen if after the initiation of the descent, the aircraft loses further human’s input?

    3. In case of aborted landing by engaging TOGA over Maimun Saleh, what would be a standard procedure to repeat landing, and what would be typical flight parameters, such as the rate of turning, CW or CCW circle, altitude?

  21. @Richard Cole @all

    Thank you for the information about the FUGRO event. Did those people say anything about what they expect to see?

    I got a shock, when being told that the resolution is at 70 cm. If the IG calculations are right MH370 should have nose-dived into the SIO in a much steeper fashion than Swiss Air which caused the plane to be torn to millions of small piece no bigger than 30 cm. Since the gravitzational forces on MH370 should have been considerably bigger, the pieces of the plane would be even smaller than that, maybe 20 cm , and the debri field would look like one piece every 7 meter. How could they possibly find anything else than the engines? Someone spoke of the coke can, they could find. just nonsense witout sense again? Is this search a fashion show?

  22. @ALSM – thanks for your thoughts. We’re WAY off on bottom line (I say 95% after 16nmi, you say 95% after 50nmi), even though I’d thought we were pretty close on key INPUTS.

    I suspect the gap is due primarily to two things:

    1) simplicity of my model – as you suggest, a full-scale mathematical model would be ridiculously complex. I think we need to think of my first graph as a distribution NOT of (initial/final/average) TURN RADII, but rather of absolute CROW’S FLIGHT DISTANCE between point of flameout and point of impact. If the first graph is redefined thus, how would you recommend refining its distribution?

    2) ALSM, we are so rarely in MY neck of the analytical woods, I’m going to have to ask you to indulge me, here. I stand on decades of experience analyzing multivariate statistical distributions when I say that non-statisticians – regardless of right-brain acuity – tend to combine uncertainties conservatively. For example, the +/-5nmi BTO error (I’d thought it was +/-3nmi a couple months back – did I miss something?) is almost as likely to DECREASE as INCREASE the distance from measured BTO arc to impact point. That’s why my final distribution is much more heavily weighted toward its centre than yours, DESPITE having as a contributing component the double-peaked distribution you describe. Think of it as the “aggregation benefit” of summing independently distributed errors – the whole will be significantly less than the sum of its parts.

  23. Brock:

    Why do I believe the probability distribution should be weighted more heavily towards the edges of the search width? Assume some type of circling or spiral descent. The true radius is changing, but assume a circle for now. To illustrate the effect of circling (vs. random), consider a simple model. Aircraft is circling down with a 10NM radius centered on the arc. Then the probability distribution looks like this:

    https://www.dropbox.com/s/y3f1ughbn16ziz9/Probability%20of%20postion.pdf?dl=0

    I am not suggesting this is what happened, just pointing out that in a generally circling path, the plane spends more time near the edges of the width, whatever that width is. Of course, the radius probability has to be factored in too, but this illustrates the nonlinear effect of the circling path for one radius assumption.

  24. @all

    Between a wobbly satellite running out of fuel, BFO’s & BTO’s. All of the Kentucky wind age, drag, drop & spindrift ain’t got sh*t on MH370.

  25. @ALSM: What I’ve been trying to point out (badly, no doubt) is that the “outer edge” concept you describe is already a built-in component of my model. While Column A models a randomly distributed DISTANCE (flameout–>impact), Column B models a uniformly distributed ANGLE this vector makes with the 7th arc, and then turns these 2 values into a distance from the 7th arc via basic trig.

    If I replace my random distribution of distances in Column A with your hypothetical KNOWN 10nmi distance – and set the BTO error back to zero – my simulation replicates your probabilities. I’ve added a second tab to demonstrate this (see “10nmi fixed dist” tab, cells AW:BB26).

    https://drive.google.com/file/d/0B-r3yuaF2p72a2R2MTB0UWVia2M/view?usp=sharing

    I think a comparison of the two tabs shows clearly that, if you replace your 10nmi fixed distance with a DISTRIBUTION of distances between 2 and 15nmi (with an average of 5), the density pulls in to the centre quite rapidly.

  26. Nihonmama – Oh, those Russians

    A RUSSIAN-SPEAKING gang of computer criminals has stolen millions of dollars since late 2013 from banks in Russia, Eastern Europe and the US, according to a report from Kaspersky Lab ZAO, a Russian computer-security firm.

    It is unclear exactly how many, or which, banks were hit. Some US financial-services executives have been briefed on the findings, people familiar with the briefings said.

    US government officials were aware of the report on Sunday, though some said they were sceptical about how much money may have been lost at US banks. Kaspersky earlier briefed a computer-security trade group for US banks, which didn’t immediately raise its “threat level,” a person familiar with the matter said.

    The report offers one of the more detailed looks at how organised crime has evolved in the computer age. The thieves captured video of what was happening on bank computer screens to learn how to mimic the way bank employees access their systems. They hacked into computers that control ATMs so that they would dispense cash when criminals walked by. They were also able to move funds from one account to another, seemingly without being noticed.

    The location of the hackers is unclear. Some of their servers are based in China and some of their Web domains appear to be registered to Chinese nationals, the report says. But, as Kaspersky notes, such clues can and have been faked in the past.

    The US Secret Service and Federal Bureau of Investigation didn’t immediately respond to requests for comment. The New York Times earlier reported on Kaspersky’s findings.

    The vast majority of victimised banks appear to be in Russia, according to the report.

    The firm’s researchers were able to access three of the hackers’ servers and found internet Protocol addresses of apparent targets. Kaspersky found 178 such IP addresses linked to Russia on the servers and 37 linked to the US, according to the report. Each address may not be a different bank, or a bank that lost money.

    The report says one bank lost $US7.3 million to ATM fraud to the group and another lost $US10 million to hackers exploiting its “online banking platform”. Kaspersky estimates cumulative losses could reach $1 billion.

    Perhaps more concerning, the hackers in this case consistently used previously known security holes in Microsoft software, for which Microsoft had issued “patches,” to break into the banks. If the companies had updated their software, the hackers would have had to find another way in, the report said.

    Kaspersky has tracked the group of criminals back to 2013 and said there is no indication they have gotten out of the bank-robbing business.

    The Financial Services-Information Sharing and Analysis Center continues to investigate the report’s contents, people familiar with the matter said. Cybercriminal attacks against Western targets have been common but the shift to mostly Russian banks and their customers is new and raises questions about the hackers’ change in tactics, a threat intelligence industry expert said.

  27. Brock: Thanks for that clarification and confirmation.

    All: For the record…There is absolutely no satellite “wobble”. That term is a very misleading description of the orbit shape and character, which is actually a near perfect ellipse, precisely known and predictable.

    All satellites have a finite inclination and eccentricity causing some finite “figure 8” ground track for every geosynchronous communications satellite ever placed in service. We are extremely lucky that I3F1 had a 1.7 degree inclination. Without it, there would be no north-south discrimination possible.

    It is not out of hydrazine. It has a limited, EOL supply that is being preserved for E-W station keeping only, required for license compliance (separation from neighboring satellites). The N-S station keeping is unnecessary for operations or compliance. It requires much more hydrazine than E-W maneuvers, so more years of service can be squeezed out by stopping the N-S burns before running completely out of hydrazine. This is SOP, not anything unusual or broken.

    In summary, the I3F1 orbit is very precisely known at all times. There is nothing about the inclination or eccentricity that introduces any “wobble” or path prediction error greater than would be the case with a perfect geostationary orbit.

  28. @ChrisButler:

    Thanks for the link. That’s the Miles O’Brien piece for PBS/Nova that aired back in October. Great stuff.

    3000 pieces of debris found from AF447.

    @Matty:

    Oh, yes those Russians. Just don’t get me talking about hacking. You know…

    But again, if you ever get a chance to see the Summer Palace, the Winter Palace (now the Hermitage), the Moscow subway and all of that architecture, nothing will surprise you.

    Kaspersky Labs gave the NYT an advance look. Their report on this hack comes out today. I’d wager it’s the tip of the iceberg and worse than what they can disclose publicly.

  29. @Matty:

    On another (Russian) note – do you know the story of Sergei Magnitsky? If not, the best, most gripping article IMO is from July 2012, in FT Magazine — “The Magnitsky Law”.

    Bill Broder, who’s at the center of this insane story (and Putin’s still after him), has a new book on the Magnitsky case entitled Red Notice.

    An adaptation for a feature film is already in the works.

    An interview with Broder today:

    https://www.youtube.com/watch?v=YbgNeQ_xINM

  30. @ALSM: you’re welcome.

    Since, in a corkscrew scenario, turn radius should effectively proxy (crow’s flight absolute distance, flameout->impact), and since my distribution of turn radii was the one aspect on which you explicitly signed off as being broadly consistent with your simulator results, then I’m going to conclude from this conversation that the correct 95% confidence interval search width for ruling out the “AP to flameout, no pilot intervention, crossing arc7 between Diamantine Trench & e87” scenario is much closer to the 16-17nmi already searched than to the 50nmi you postulated yesterday.

    (To save one back & forth: superimposed fugoid oscillations would materially affect neither path [as viewed from above] nor final impact point; as such, I do not believe their absence from my model moves the needle.)

    @all: I hope all this mumbo jumbo does not obscure the importance of this is argument; I feel it is NOW time to challenge search leadership on this, their latest search debacle, whilst @ALSM is arguing we should wait several more months – until the current search width has been TRIPLED – before challenging the status quo.

  31. Brock,

    As I mentioned already several times, apparently there is class of “free flight” models that fit BTO & BFO data (except 18:40), subject to the parameterization of the thrust and fuel consumption. My latest run already converged to |BFO residuals| <2.5 Hz. All trajectories end up around 100E. The flight mode would likely correspond to either TOGA, or engine 'failsafe' mode (in case engines loss communication with FMC). Earlier I requested ALSM whether he is able to check what would happen in case of the sequential flame out of the engines at 6 km altitude. In contrast to the AP mode, I would expect the aircraft to plunge to below 3 km after the flame of the first engine. But so far I did not get any comments from him on this issue.

    The second issue is to explain the reason of this mode of the flight, and whereabouts of the plane between 18:22 and 19:25 (19:41). But this is a common problem.

    In either case, I see the probability of the crash location pointed by IG as very low. And not only because of the post-flame-out scenario.

  32. @Nihonmama: thanks for the link.

    It is odd to me that such a high-budget video – released February, 2015 – would fail even to MENTION the fatal flaw in the scenario the NTSB/FBI/Inmarsat appears still to be selling: that it requires surface debris, which drift experts say should have begun hitting Australian shores by August, 2014.

    SUSPICIOUSLY odd…

  33. Nihonmama – the hacking – it will be an iceberg for sure. Whether MH370 is a hack or not we are cruising in an age where bedroom dwelling malcontents sit with contaminated brains looking for a thrill. What would be really dangerous if they actually went outside and joined up with each other. Seems to be the case in Russia.

  34. @Oleksandr: I appreciate your efforts – and am a kindred spirit re: doubts about the establishment theory – but describing your model as a beautiful fit to every BFO value “except 18:40” is, to me, like describing a trip over Niagara Falls in a barrel as a beautiful ride “except for the landing”. Either it fits ALL reliable BFO values to within a well-defended tolerance, or it doesn’t.

    If it DOES fit all values – but requires deviation from common flight modes to achieve the fit – then I for one don’t rule it out, but DO consider the deviation at least a strike against.

    If we as a group have any influence on the search whatsoever, I would much rather we use it to leverage public demand that the JIT turn out its pockets (FULL disclosure of ALL models & inputs) than to demand we spend years running everyone’s pet theory to ground.

    I agree with Victor & Matty: what’s needed now is an investigation of the investigation, not a re-re-rehash of the dodgy data.

  35. @Brock:

    Mark the key word you used: “selling”. That doc left out a whole lot. And it seems to be aimed at an audience that doesn’t have much (granular) information.

    Matty:

    “it will be an iceberg for sure” YES.

    And, there appears to be something else strange going on out in the world concerning some Russian groups and money laundering. With cohorts beyond Russia. More on this perhaps later.

    And all: from the same ATSB that is conducting the increasingly smelly search for MH370, here’s the latest:

    Did Australia mislead ICAO over the PelAir crash?

    “The core issue is that the Pel-Air ditching in 2009, and the botched accident report produced by the Australian Transport Safety Bureau, the ATSB, doesn’t appear to have been reported to ICAO.”

    http://t.co/IFVWqZrmJS

    Time to go eat some shrimp.

  36. Hi Brock,

    I think I was not sufficiently clear. What I meant is that I intentionally left out BFO 18:40. I am not as sure as others that it corresponds to the same final mode of the flight. Generally, I think it is possible to make an assumption of a single FMT to fit 18:40 BFO in the manner IG did it, but I am running into the same problem as with AP theory: such a single turn makes no sense.

    However, the intriguing feature is that the position 19:41 predicted by my latest “free flight” model is 93.6E, 4.4N, 5.5 km altitude. According to Kate (the sailoress), she saw some unusual aircraft from her yacht at ~94.5E, 6.7N, 3 km altitude, 15 minutes earlier. And it was heading exactly as it would need to move to reach 93.6E, 4.4N according to her witnessing. Coincidence? May be. But given a number of other coincidences, I think 18:40 BFO could mean something else, descent for example, or circling around Maimun Saleh airport.

  37. Quote victor//

    Here is a short list of data that would be helpful:
    1. All raw radar data from all sources.
    2. All ATC communications, including between ATC centers and between KL and the RMAF.
    3. Surveillance video of the plane while on the ground at KLIA.
    4. The complete, unredacted satellite logs, including the payload data (not just the signaling data).

    End quote Victor//

    Sorry, I could not resist. How about?

    5. The “compelling circumstantial evidence” that ALSM has to support the AP assumption after the FMT.

    That would help me a lot.

    The reality is that none of the information you list would change anything in the context of the IG mindset.

  38. Dennis: Every 777, 767 and 747 pilot we have discussed this question with believes (strongly) the autopilot would be the only way anyone would fly the plane for 5 hours, regardless of where they were going, or what their motive. The AP is the standard interface everyone uses to fly the plane, not stick and rudder. It is very difficult to control the plane manually for any extended time at cruse altitude and speed. We have discussed this dozens of times, so please stop asking over and over just because you don’t like the answer.

  39. Yes, at cruising altitude and speed, the plane would be flown by AP. However a human pilot could still be actively controlling the plane by changing heading, speed, and altitude every few minutes. Or, a pilot could choose to fly at, say, 10,000 feet where stick and rudder is quite practical. He could even be trying out various acrobatic maneuvers just to see what he and the plane could manage. Once we allow this kind of thinking, a large number of paths can fit the BTO/BFO data, and at lower altitudes endurance can be much greater than at LRC altitude, allowing many paths to fall within a “performance limit.” I don’t think this sort of scenario is likely, and worse, if it did occur, we would have very little idea of where to look.

  40. @Dave Reed

    Thanks for the clarification/illumination. It is precisely this kind of scenario that some seem so eager to dismiss. For what’s to be done with said set of conditions?

    Yet, in Shah’s own words vis a vis his newly installed flight sim motor: “Now we can do some REAL flying”.

    But surely he was just whittling his thumbs, frozen in time, on his 5 hour southern beeline.

    It’s time for a new approach, IMHO.

    Spencer

  41. @ALSM

    OK. What you have is what you have. I am off it. But you know how I feel about it. No hard feelings.

  42. @ALSM: thanks for pointing me to a means of calibrating the BTO error, which I have now done (turns out I was pretty close).

    I also re-platformed the whole thing on nautical miles (because I was annoying MYSELF with statute), and made a few improvements in layout and documentation. Latest version is linked to below.

    Note that, in addition to widening the BTO distribution slightly (to match the 50/95/99 tail criteria you provided – thanks!), I’ve also effectively widened the crow’s flight distance distribution by roughly 15%, by virtue of switching from statute to nautical without re-parameterizing the distribution. This was a nod to your general comment that we should err on the conservative side, here.

    After making all adjustments, the density ALREADY COVERED drops to 92%.

    https://drive.google.com/file/d/0B-r3yuaF2p72STFqZ09ua0xCQjQ/view?usp=sharing

  43. @N-mama

    That was great, Thanks!

    @All

    IMO…he was flying with & w/o the autopilot to throw off the landing area/s.

  44. Saw the doco and apparently no one had terror links and it’s off the table? Not sure where that narrative is really coming from as everyday young westerners with no terror links arrive in Iraq/Syria to fight with IS. Some crew members do indeed fit that template.

    The AP question might come back to the state of mid of the pilot? Holding the stick might be annoying but no more arduous than holding a steering wheel while you head off for a summer vacation with the kids in the back. It may boil down to where the pilot’s head was. What does a hijacker do once he sets the AP? Play sudoku? What did the 9/11 mob do?

    On the search I’m now half inclined to think that Abbott may shell out a bit longer as he probably now thinks this thing owes him something. If he could be convinced that it would bear fruit? But he may also be skeptical by now and by May it’s getting nasty down there and it may need to stop.

    Spencer – an earnest inquiry from me: Do we have anything official about that simulator, or another unnamed source?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.