MH370: Triumph of the Weird

Well, here we are in cold-case land, scratching our heads. Some $150 million has been spent and no plane. Where does that leave us?

For one thing, with radically altered probabilities of what might have happened.

Imagine that we dial the clock back to August, 2015. You’re Warren Truss and this the story that your data is telling you:

Based on this default story, you’ve concluded that there’s a 97 percent chance that the aircraft hit the ocean within a 120,000 sq km box. What about the other 3 percent? Let’s imagine there’s a 2 percent chance that it’s somewhere else in the SIO, and a small but finite chance — let’s say 1 percent — that, for unkonwn and uncalculable reasons, the plane didn’t go into the SIO at all.

Time goes by. You search out all but the 1 percent of the search box that your sonar equipment can’t image (e.g. seabed crevasses), throw up your hands, and call it off.

So this is how things now stand: Of the orginal 100 percent, 96 percent has been scanned and ruled out. Here’s how remaining probabilities now stack up:

Of course, these are all very rough numbers. The point being, no matter how you slice it, the scenario that was once nearly a dead certainty (flying into the SIO search box) is now less than an even bet, and outcomes that once barely merited an asterisk are now not only possible but probable.

The most probable category, according to this rough calculation, would be scenarios of the second variety. But if the data is valid, how could the default story be wrong? How could the plane have wound up somewhere in the SIO outside the search box? To square that circle, you have to choose one of the assumptions above and bend it. For instance, one might imagine that the 18:40 BFO value was not caused by the plane flying south, but by a plane that descended–perhaps, say, for a descent into Car Nicobar–and then changed plans and flew instead to, say, Antarctica. Or maybe the plane didn’t fly straight and fast, but flew slowly in a curve toward the Cocos Islands, creating a pattern of ping rings that only happened to look similar to those generated by a plane flying fast and straight.

Such eventualities are so unlikely that, back when the search box was being drawn, it was easy to simply discard them. But now that the most reasonable options are off the table, this very geographically dispersed (and hence impractical to search) population of possible endpoints collectively adds up to “very likely.”

Then again, it’s also now significantly more plausible that the plane didn’t go south at all, or that if it did it wound up in the search box but then fell into an unscannable crevasse.

Whatever happened to MH370, it wasn’t the default story told by the data, but rather something that in the summer of 2015 would have been discarded as hopelessly implausible.

215 thoughts on “MH370: Triumph of the Weird”

  1. Jeff you don’t have to go defensive about it since this is yor personal blog and you are free to promote whatever agenda you want, it’s perfectly fine.

    However you also must expect people to say what they really think about it and not take it too personal.

  2. The INMARSAT log shows that the SATCOM link was lost at sometime between 17:07 and 18:03, and that the aircraft initiated a log-on request at 18:25. The most plausible explanation for those events is that the configuration of the electrical power supply changed twice in that time period.

    The primary radar data show undeniably that after the turn-back near IGARI until the loss of radar coverage the autopilot remained dis-engaged. While the high-acceleration manoeuvre that apparently occurred near IGARI explains the dis-engagement of the autopilot, it is strange that the autopilot was not re-engaged after that manoeuvre. Perhaps the autopilot was not available?

    An interesting question could be which non-normal electrical configuration would cause the loss of SDU and autopilot. The autopilot requires the normal mode of the flight control system, and that requires valid air data. Air data is valid when two or more of the three sources agree in the ADIRU, SAARU or both. Presumably that requires pitot heat for at least two of the three pitots.

  3. @StevanG, I’m not sure what you’re referring to. I try not to take things personally but may be guilty of that from time to time. I do tend to get proactive when I think that people are spreading misinformation or are being rude.

  4. I am quite amazed at the widespread interest and concern over MH370, I live between Port Elizabeth and Jeffrey’s bay. I used to be an avionician, and am still involved in a small flying club. I get people asking our club about things they find on the beach, on a weekly basis. We refer all to the CAA. Nice to know with all our worldly problems, people actually care

  5. @Ge Rijn, Rob, Gysbreght, Kenyon and others interested in the significance of right flaperon separation. A little light reading:

    New point: Flaperons should have turned the aircraft left at fuel exhaustion.

    Outcome: A full glide and ditching can be discounted, increasing confidence in the currently recommended search for some.
    It is possible that flaperon asymmetry, which would reduce search width, was not part of the Boeing simulations and if not, that could reduce search width.

  6. How it will end

    Even on land some crashed planes have been located after 50 years or more. Google for “Star Dust”. There are a couple of instances of Indian Air Force transport aircraft being found after 30+ years. Many US transport aircraft on the India-China hump route crashed in the Himalayas (maybe Indian, Chinese and Burmese territory) and have never been traced.
    In this case, it is quite possible that there are no large pieces of wreckage (other than the engines) so there will be no large pieces to be found on the ocean floor (as there were in AF 447). Interestingly, human remains were recovered from the wreckage of AF 447 even after two years and were identified through DNA tests and other means.But this “preservation” may not happen everywhere. The FDR and CVR were recovered in AF 447. But even if they are recovered in MH 370 it is quite likely that they will not show much that is useful. The FDR may reproduce the path, but not why the path was taken. The CVR may have been switched off by the perpetators (Z or whoever it was). And if it was a ghost flight, there will be nothing to hear in the last half-hour except the sound of the crash.
    So how will the mystery end? Advances in sensing technology may find something on the seabed a few decades from now.
    Or: Some people in the Malaysian government and/or military knows what happened and will break their silence some time.
    Or: US intelligence must have been monitoring events in real time and someone involved will write a book about it sometime, even though it takes a few decades.
    Or: some kind of private investigation is able to access information suppressed by Malaysia. Perhaps someone here is working in that direction.

  7. @Jeff Wise

    To set things straight a bit about your percieved ‘Russian connection’ to MH17 and then to MH370. I don’t doubt your integrity in this matter but your objectivity.

    There still is no proof Russians shot MH17 under direct command of Russian authorities.
    There is full proof Russia delivered a BUK-system to the Ukraine pro-Russian rebels most probably with a Russian unit-commander which was responsible for it’s operation.

    I rather believe this commander made an awfull mistake fed up by his Ukraine counterparts.
    Russian involvement in MH17 is obvious.
    I agree with you. But in another way.
    An embarrissing way I think.
    The stupid thing is Putin and his government chose to cover their big mistake and kept doing this till the point of no return.
    Still there still is a point of return.

    If Putin would choose to admit the mistake afterall and put everything on the table about what happened he would still gain back some respect he lost and sure would relieve the suffering of all the NoK from MH17.

    Speculation about a Russian link to MH370 would also vanish with it.
    This would be helpfull for IMO it doesn’t help to find the plane now.
    It’s distracting from the current data available.

    This data is all we have right now. You can ignore them, think they are false and manipulated but still it remains all we have.

    Till real evidence shows up pointing to a Russian- or any other scenario the SIO remains the place to search.
    And I believe North close to the discarded search area.

    Their calculations where not wrong regarding the data they had at the time IMO.
    They only refused to adjust this data soon enough after it became obviously clear they were probably off almost 10 degrees lattitude after the flaperon find and after all info that came from other debris and drift-analyzis.
    And the ATSB knows this regarding their final statements just before the ending of the search.

    I hope too a private project takes over taking in account every valuable objective thought and option that’s presented here and on other blogs.
    Just looking at the data!

  8. @Gysbreght
    Loss of AutoPilot at IGARI is an interesting theory but it is a new idea, right? I do not think IG or ATSB or others are saying that.

    Without raw military radar data, which Malaysia is holding as a state secret, I have to throw my hands up in the air in frustration re: IGARI area maneuvers. Seems constant 0.84 Mach and fairly constant altitude fit the path timing quite well, as one scenario.

  9. This post follows on from a previous post;

    (In the explanation that follows, I do not list every button press that
    would occur, but instead give a broad outline of the way reprogramming of
    MH370’s Flight Management Computer was actioned.)

    When MH370 flew-by waypoint IGARI onto M765, the currently Active Waypoint
    in the Flight Management Computers route became BITOD. The Active Route
    in the FMC had the DEPARTURE airport of WMKK and the ARRIVAL airport of ZBAA.

    When the circumstance occurred which necessitated MH370’s turnback after IGARI,
    the Pilot Flying diverted the aircraft to fly towards a closer airport and
    may have expedited reprogramming the changed flightpath into the FMC so as to
    concentrate on attending to the circumstance.

    The number of routes that are stored in the FMC memory is limited by the
    amount of data e.g. waypoints, airways, alternate diversion route(s), radio
    navigation aids, etc., that make up each route, or each partial route, that
    is stored. The memory space is limited compared with current day standards.
    Routes can be input by e.g. hand, or by e.g. download from Malaysia Airline’s
    Operations Dispatch Centre. Routes previously stored in the FMC memory space
    may eventually be overwritten by more newly input routes, or routes from
    previous flights may be deleted (e.g. at the completion of the flight) if
    the pilots choose to do so.

    The procedure to hand reprogram a flightpath for the autopilot to follow can
    be performed in several different ways. One method can be to utilize one of the
    other (non-active) Routes that may be present in the (memory) Route ‘pages’
    of the FMC. In MH370’s case, obviously a route which facilitated sending MH370
    to, or near to, the pilots flying desired airport, would be a prefereable route.

    The previous flights of MH370 appear to be, (going back in time);
    (WMKK to unknown -disappearance flight, did not terminate at ZBAA.)
    WMKK to ZBAA (Peking, China), & return.
    WMKK to VHHH (Hong Kong, China), & return.
    WMKK to VGHS (Dhaka, Bangladesh), & return.
    WMKK to WADD (Bali, Indonesia), & return.
    WMKK to VABB (Mumbai, India) 4th March 2014 Departure 12:30PM UTC (08:30 PM MYT),
    & return.

    While we can’t be sure that any of the programmed routes, or partial routes,
    of these previous flights remained stored in the FMC memory, it is known
    that the route WMKK to VABB uses N571. The waypoints of the route to VABB
    can have the following sequence;

    I consider that one of the flightcrew selected in the CDU the route WMKK to
    VABB, or its partial route (its LEG), and added waypoints to it such that
    the routes amended waypoint sequence was;
    (WMKK etc….) GUNIP IGARI KENDI VAMPI MEKAR NILAM IGOGU (..etc. ..VABB),100.95941162494388&chart=304&zoom=4&fpl=GUNIP%20IGARI%20KENDI%20VAMPI%20MEKAR%20NILAM%20IGOGU

    (It is possible that other waypoints may have been added, such as ENDOR
    placed before KENDI, or one of the waypoints north of Kota Bharu airport
    placed after IGARI, however, minimally, the sequence as shown is sufficient
    for the flight crews purpose of diverting to WMKP.)

    The ‘amended’ route, or partial route, of WMKK to VABB was then made the
    Active Route and the aircraft was steered on an intercept course to the
    route line between IGARI and KENDI, by means of one of the HDG/TRK modes.
    Pressing LNAV would ‘arm’ the LNAV mode and the autopilot would fly MH370
    so as to to merge onto that amended route. About 2.5 Nm from the flightpath
    line LNAV would engage and MH370’s autopilot would thereafter LNAV along the
    sequence of waypoints.

    I consider the flightcrew member that reprogrammed the FMC intended the
    autopilot to LNAV MH370 to the KENDI waypoint, whereupon (after having had
    the time to deal with the circumstance) the flightcrew would disengage the
    autopilot and fly the aircraft to a landing at WMKP from the KENDI waypoint.
    The route programmed was merely a means to that end, and the remainder of
    the route after KENDI was regarded as non-relevant by the flightcrew member.
    People, even in an emergency, do not expect to die. The flight management
    computer simply continued to action the remainder of the route.

    Refering to the FI, page 3, the turn back after IGARI appears to have
    commenced about 17:21:13 UTC and at 17:30:35 MH370 was stated to be at height
    of 35700 feet. That height is close to the additional plus or minus 500 feet
    the Malaysian eAIP specifies should be adopted in its “IN-FLIGHT CONTINGENCY
    PROCEDURES”…for aircraft requiring..”TURN-BACK OR DIVERSION” (as cited in
    my previous post). By 17:30 UTC, flight at ~35500 feet had probably been
    occurring for several minutes, and that level is consistent with operation of
    the required eAIP contingency procedure.

    The FI, page 10, figure 1.1E – Primary Radar Targets (plots), shows MH370
    as north of WMKC at 17:36:50:32, and at about that time page 3 states MH370
    was exhibiting “height fluctuation between 31100 and 33000” feet. When MH370
    passed north of WMKC it flew through three or more airways which converge
    on that area, and for which heights of MH370 crossing at 31100 to 33000
    would place aircraft in danger of collision. Operation of MH370 between
    these levels suggest that the flightcrew by that time were non-conscious and
    the autopilot was attempting to fly the aircraft under difficult conditions.
    (It should be understood that a factor that could contribute to those
    conditions is that once MH370 began to follow the reprogrammed Active
    Route, I understand that the previous {different} settings for that route
    would apply. This may mean that e.g., different settings for height, for
    wind, etc., possibly for speed, could apply.)

    The theory in this post explains why the path from IGARI was taken but
    does not address further aspects of the events that impacted upon MH370,
    such as what the circumstances were that resulted in MH370’s degraded flight
    behaviour (e.g., ‘coast and drop’ with probable overspeed), and such as
    what factor resulted in the FMT. At present, I have no strong hypotheses
    for those events.

    (If attempting to replicate the reprogramming of the FMC in a flight sim,
    I suggest not simply starting the sim in the South China Sea – instead,
    ensure there is minimally, the active route WMKK IGARI BITOD ZBAA in the
    sim, and the active waypoint is BITOD.)

    (As to the SDU reboot, I suggested a cause for that in this post;)

  10. Apologies for the delay in posting the above theory, which I had intended
    to post several days back.

  11. Clarification Note;
    When I refer to the “previous flights of MH370” above, please understand
    that as, in fact, ‘the previous flights of 9M-MRO‘.

  12. @TBill: You need to look at Figure 4.2 in the DSTG report. These are smoothed (filtered) primary radar data every 10 seconds. The three-sigma variance of the raw data is shown by little dots.

    Until 17:20 (IGARI) the track is straight as a ruler. The groundspeed varies only slightly due to wind variations at a constant M.82.

    After 17:20 the speed, and to a lesser degree the track angle, vary much more than they would in auto-piloted flight. The 10-second radar data end at 18:02. The remainder of the curves is extrapolated.

  13. @TBill, @buyerninety:

    To make it easy, here is the figure I’m referring to.

    @buyerninety: “If you had only the radar data for flightpath & height of 9M-MRG; would you undeniably be able to state when it was that the the human pilots were flying or when it was that the autopilot was flying?”

    In the case of MH370 (9M-MRO), I have track and speed. Those parameters show undiniably that the airplane was not controlled by the autopilot.

    In the case of 9M-MRG it depends on the track, which is not shown in the ATSB report. I wouldn’t trust the radar data for altitude without more information on its accuracy.

  14. @buyerninety: If there were radar data for 9M-MRG it would show groundspeed, and that would definitely identify when the autopilot disengaged.

  15. @Gysbreght
    How would you be able to tell when the autopilot disengaged in 9M-MRG,
    as opposed to when the pilots were flying it?

  16. @buyerninety. About SDU reboot, you alluded to some quotes earlier about it being part of ELMS load shedding and restoration.
    “”The load shedding is galleys first….followed by individual equipment items powered by the Main AC busses”. When…”the loads
    decrease, ELMS restores power to {the} shed systems”.”

    The ‘individual equipment items’ include various fans, seat equipment, chillers and hydaulic pumps but not the SDU. On loss of AC from IDGs and APU the SATCOM (and TCAS) drop off line – the backup generator powers just ‘essential equipement’, which does not include them. Getting an IDG or the APU going will revive them.

    In case you do any more looking into FMT and subsequent, or the “circumstance”.

  17. @buyerninety. ‘Getting an IDG or the APU going will revive them’ was a bit loose. Getting one powering the left main AC bus one part of me meant to say.

  18. @buyerninety Posted February 11, 2017 at 10:50 PM:

    How would you be able to tell when the autopilot disengaged in 9M-MRG, as opposed to when the pilots were flying it?

    In the case of 9M-MRG the heading trace shows a constant heading, a U-turn, and again an approximately constant heading for about 10 minutes. Although the transition between the U-turn and the following constant heading is not quite what would be expected for an A/P-controlled manoeuvre, that discrepancy is perhaps too small to allow a conclusion that the A/P was disengaged.

    However, the groundspeed trace shows variations that are not compatible with any A/P mode.

    This is part of Figure 6 of the ATSB report on the incident:

  19. @TBill

    As promised, I have now put together my thoughts on a route following waypoints all along and ending north of the previous area. My argument is that there is only one such route possible on the least of necessary assumptions (one change in speed at the penultimate waypoint), assuming that there was no unexplained loitering and taking into account the so called 1912 NW point. I also try and connect this with a speculative route around Sumatra, bypassing Indonesian radar. Finally, I compare the result with peer-reviewed drift studies, arguing that there is a particularly good match with the one drift study that does include debris washed ashore in Africa.

    I’d very much welcome any feedback on this. Please feel free to tear into pieces.

  20. @Buyerninety
    Regarding your scenario, the FMC is designed to make it is easy to select an alternate airport such as WMKP to land. Therefore why bother pulling up an old Route2 over N571 and modify that for the purpose? From FlightRadar24 it looks like N571 is heavily traveled, so chances are an MAS pilot would be quite familiar.

  21. @Nederland
    OK got it…the first time I tried the link did not seem to work. Thank you for the reference but I am trying to make sure I agree with what I said. One thing is I am mostly using FS2004 (aka FS9) right now. FSX is the final MSFS software release which is needed for the more rigorous PMDG 777 model. But that should not impact flight paths. One thing I was saying is for a “curved” or slanted path such 170S to NOBEY, it can be hard to see big differences in end point between a great circle, and TRUE or MAG heading for shorter segments (eg; after BEBIM), not accounting for wind.

  22. @Nederand

    Looks like it hangs together from my pass through. I think your route relying mainly on waypoints or the LNAV routes using Cocos (overflown) or McMurdo are similar, plausible, and representative of a family of routes that can end in the 26S to 30S area on the 7th arc. I don’t know that there is a way to prefer one over another. Simply showing feasibility is sufficient.

    Routes intersecting 19:41 further North, say around 8N, so allow higher ground speeds in the range of 480-500 knots and are more consistent with a loiter, but that does not mean your ground speed of 425 knots is unreasonable.

    I also agree that the added refinements you mention will make little difference to the terminus, and no difference at all with respect to feasibility.

  23. @TBill

    Thanks, I shall update this accordingly.


    Can you kindly provide a link to your Cocos route?

    I found the presumed ground speed in agreement with the ATSB’s assumption of a route leading into that area (although that’s the one still further north):

    “One analysis showed that the best fit for the Doppler frequency was at a ground speed of 400 kts, with slightly ‘less’ best fits at 375 and 425 kts. A Monte Carlo style analysis, using a number of different starting positions on the 2nd arc also gave a best fit at 400 kts. A most probable speed range of 375 to 425 kts was selected. … Flight planning carried out by MAS independently showed that there was sufficient fuel onboard the aircraft to reach the positions determine by the analysis.”

    I should perhaps provide a clearer reference to this:
    p. 46.

    My impression was that it is a) difficult to reconcile a route overflying Cocos with further waypoints en route b) does not satisfy the ATSB “1912 NW point”.

  24. @DennisW

    Thanks. I think one of the advantages of the route proposed above is that it does not require an unexplained loiter near Car Nicobar airport, only for the aircraft then flying back into Sabang radar range. If the ATSB was correct at one stage to eliminate a point north of the 1912 NW point from consideration, then those routes wouldn’t work either.

    Also, I found it striking that a route to IGOGU/APASI would work well with a descent rate of 2000 ft/min as it means MH370 would drop off the radar (whether or not it was operational) by the time it reached a target altitude of ~ 20,000 ft. That would not work out as nicely if the plane was going to Car Nicobar or a similar destination first.

  25. @Nederland

    Repeat the Indonesians declared diplomatically they did not see the plane in their airspace.
    They probably did not lied with this statement.
    But another statement tells they saw it among the Andaman Islands.
    I don’t believe their radar stations were down that night or they were sleeping.

    I think their statements tell the plane went around Indonesian airspace or went below their radar as you suggest.

    A diving under radar could fit with a Kate Tee scenario. Coming from The North near Car Nicobar decending below Indonesian radar with a shut-off burning right engine that flamed-out and triggered the APU to start, for the left engine IDG was isolated before at IGARI, forcing the SDU re-logon at ~18:25.

    I’m seriously thinking something crucial happened around 18:22 when the plane suddenly disapeared form radar.
    I think there must have been a steep descent there to vanish so quicly.

    And later climbing back to altitude of max. ~25.000ft on one engine could explane a slightly slower speed and endurance ending up North of the discarded search zone.

  26. @Ge Rijn, Recall that, according to the most recent report that we have that references the subject (“Bayesian Methods”), the plane didn’t vanish from radar at 18:22; rather, it appeared for the first time in twenty minutes.
    I’ve suggested this brief appearance might have been due to the plane dipping its wings toward the radar head in the course of a turn, then vanishing again as the wings are leveled off as the new direction of flight is established, but of course this is pure speculation.

  27. @Nederland

    My point is that it really does not matter. There is good reason to believe the aircraft terminated North of the current search area. I think that conclusion is approaching a consensus. The fact that plausible routes relative to the ISAT data can be made to work is simply icing on the cake. I am not prepared to create a probability distribution as a function of latitude or endorse 30S over 27S.

    In any case, the ball is not in our court at the moment. Not that it has ever been. I do not see a new search starting on the basis of analytics – yours, SSWG, mine, Victor’s, DrB’s,… Those days are over.

  28. @Nederland it’s a good job but “no loiter” is still an assumption… we have no idea if there was a loiter or not.

    @Ge Rijn

    I believe Kate Tee might have seen MH370, too bad she couldn’t be investigated properly as it didn’t fit official pet theory.

  29. @Ge Rijn

    I’d take the “Andaman Islands” mentioned by spokesman Agus Barnas as referring generally to the Andaman and Nicobar archipelago or the Andaman Sea. That would cover a path such as between IGOGU and APASI – otherwise MH370 would have been close to Port Blair, and it clearly doesn’t make sense as Indonesia can’t look that far and then also there is the so called NW point as a terminus.


    It’s just another approach. One never knows what happens to anyone’s ideas in several years time.

  30. @Ge Rijn
    “I don’t believe their radar stations were down that night or they were sleeping. I think their statements tell the plane went around Indonesian airspace or went below their radar as you suggest.”

    Obviously Nederland and I tend to agree with you, but the ATSB/orig IG search area in essence assumes FMT at 18:40 in Indonesia’s back yard. What is their rationale for ignoring the Sabang radar? Hate to say my concern is that they wanted to search in a certain area for various reasons unrelated to finding MH370.

  31. @Ge Rijn “A diving under radar could fit with a Kate Tee scenario”
    @StevanG “I believe Kate Tee might have seen MH370, too bad she couldn’t be investigated properly as it didn’t fit official pet theory”

    The proposed track based on Freddie’s motive fits Kate Tee’s GPS track as plotted very well with the aircraft overhead just after 19:20.
    MH370 goes out into the Anderman Sea towards LAGOG then comes back to NOPEK while descending to a lower altitude to facilitate an approach and landing at Banda Aceh. With no confirmation from the ground of successful negotiations it heads out to BEDAX.

  32. @StevanG

    There is at least the indirect argument that the 1912 NW point probably precludes any extended loiter in that area. For example, the path proposed by Godfrey/Iannello has the great circle route to McMurdo starting at 19:41 (and involving ~ 45 min of loiter around Car Nicobar). In that case, MH370 would have been spotted by the Singapore radar asset.

    Even in a negotiation scenario, I find a loiter unlikely as there would have been several landing sites in reach up to waypoint BEBIM. It would be more logical, I feel, that MH370 inteded to clear the Sabang radar zone asap.

  33. @Nederland

    “It’s just another approach. One never knows what happens to anyone’s ideas in several years time.”

    Sure enough.

    My interest has been rekindled on a couple of issues JW highlighted on this blog over the years.

    A big one is the reboot at 18:25. There is no satisfactory explanation for it. The fact that the BFO data fits so well to other observables at that time is quite inexplicable. It really does rule out a prolonged absence of power to the AES oscillator, and most certainly rules out a depressurization event. I would even be so bold as to say power was never removed from the AES. At the risk of fanning Jeff’s flames, it is far more probably that someone simply was disconnecting things like RF cabling as opposed to removing power.

    The hole in the LIDO radar track is another. I cannot think of anything that really hangs together for that. Don Thompson speculates that it is the result of combining two radar tracks with the latter track being associated with Western Hill. That actually feels pretty good to me assuming that Western Hill was dormant and advised to resume active status.

    In the background there is the bizarre behavior of the Malays in the hours and days following the diversion. There is a lot bad smell there for sure.

  34. @Nederland
    “… the 1912 NW point probably precludes any extended loiter in that area. For example, the path proposed by Godfrey/Iannello has the great circle route to McMurdo starting at 19:41 (and involving ~ 45 min of loiter around Car Nicobar).”

    I believe Victor, myself, probably DennisW are not thinking the 1912 siting is inconsistent with a loiter, rather it is consistent with a loiter, whereas a loiter is defined by not making an immediate FMT at or before 1840. Exactly what that loiter looked like is probably up for grabs. Victor/RichardG are not the only persons to independently discover that the McMurdo waypoint apparently used in the simulator seems to also fit the actual flight Inmarsat pings (@pigdead over on Reddit also discovered that). Right now the Iannello/Godfrey path is about the best thing we have, although admittedly it seems there could be other possible variations on the McMurdo theme.

  35. @Tbill

    Yes, I do endorse the notion of a loiter. It is consistent with my theory for the motive of the diversion. At the end of the day people do things for a reason. I not believe suicide was that reason. There is no fit – in particular the diversion of a Beijing bound flight for the purpose of dropping the plane in the SIO.

  36. @Ge Rijn

    “Repeat the Indonesians declared diplomatically they did not see the plane in their airspace. They probably did not lied with this statement.”


    General Sutarman, Indonesia’s chief of police at the time of MH370’s disappearance, is alleged to have made the following comment in a September 2014 Huffington Post article;

    “I spoke with the Chief of Police of Malaysia Tun Mohammad Hanif Omar, I know what really happened with the MH370,” he is cited as saying by Indonesian news site Kompas.”

    This needs to be added to the “triumph of the weird”

  37. @TBill

    I don’t think it was a suicide either.

    I should perhaps set out more clearly in the updated version why the loiter assumed by Godfrey/Iannello (and others) is not very likely imo. I feel it’s for three reasons:

    a) it requires a very special position of the Singapore radar asset in relation to the NW point:
    The ASTB said that this NW point is approximate only, but defines endurance limits. My understanding therefore is that the ATSB excludes the possibility the plane was north of the NW point and this precludes the earlier route.

    b) It requires that Indonesian radars in Aceh were switched off, definitely Sabang, probably also Lhokseumawe, despite indications that they were not turned off

    c) the pilot didn’t care at all whether or not Indonesian radar was turned off neither during the loiter segment near Car Nicobar nor on his way to McMurdo.

    Also, there are some BFO variations in the Car Nicobar to McMurdo route, beyond 2 standard deviations, statistically very unlikely (unless there are other factors to consider).

  38. @Gysbreght
    You said in reference to 9M-MRO, after 17:20, “…track and speed. Those
    parameters show undiniably (sic) that the airplane was not controlled by
    the autopilot.”

    In the case of 9M-MRG, you said “…groundspeed trace shows variations
    that are not compatible with any A/P mode.”

    In the case of 9M-MRG, you would now be aware that looking only at the
    radar data track and speed can be misleading – as the example of 9M-MRG
    shows, the incorrect conclusion could be drawn that the autopilot
    was off and that it was the pilot flying the 777 like ‘a jet fighter’.
    In fact, it is known that it was the autopilot that was flying that 777
    when those large variations in track and speed were occuring, due to the
    autopilot receiving incorrect indications from a fault in the equipment.

    Therefore, your assertion as to when the autopilot was flying, depends
    on an assumption that it is ‘normal conditions, i.e autopilot flying an
    aircraft with no degradation in the equipment (and with a human pilot
    monitoring the autopilot every second).

    You should allow the possibility that there may have been damage to the
    equipment and/or cabling on 9M-MRO, and there was no human pilot able
    to intervene to correct any instability.

    The FI does state that MH370 was fluctuating in height between 33000 and
    31100 north of WMKC, (although perhaps you attribute that to the slant
    range from WMKC radar? – fair enough), but also the language of the FI,
    stating that MH370 would ‘coast’ and it would ‘drop’ (page 7), that
    wording can allow an interpretation that the autopilot was encountering
    an unstable condition (such as an overspeed or/and in combination with
    an attempt to maintain a particular height level).
    I seem to recall that VictorI maintained (at least in the past) that
    there was overspeed of MH370 in parts of the track before WMKP. Until
    more information is available about MH370, I don’t share your certainty
    in the matter of the autopilot vs human pilot question.
    (Note; also the DSTG book graphic is derived using the assumption of
    {page 20} “a simplified almost constant velocity model” for MH370. If
    MH370 was e.g. encountering a periodic overspeed condition that was
    being ameliorated by the autopilot, their assumption would not apply.)

  39. @JeffWise @DennisW

    coincidence radar disappearance and SAT logon

    I thought about an explanation since october 2014 but i feel the assumption of Jeff Wise, that the last radar contact was not a disappearance in the first place, which happened about 20 minutes before, but was really a short reappearance because of a flight maneuvre is the best i could think of. Its a real explanation from a practical aviators view and sound simple enough to withstand the famous “Razor”.

    But then, we really can only guess, that this radar signal was indeed MH370. Its strange for me, that the Investigation was so sure about that signal being linked to MH370. I would guess, when you didnt have radar contact for 20 minutes you lost it. Fullstop. Exclamation mark. The last contact signal could hav e been anything fantasy can imagine. its more than speculative.

    One of the fantasies is, that it was indeed MH370 (and maybe this fantasy is suported by some insider knowledge from indonesian sources). If that is true, it can only be for the reason Jeff was speaking about in another thread. A flight maneuvre, that produces a short lived radar return. If that is true, the coincidence of SAT logon and flight maneuvre has to be analyzed.

    Jeff Wise presumed, that the flight maneuvre was a right turn towards N and the SAT logon was timed to this final deviation, not to the loss of radar range.

    Also since DennisW made very healthy observations about the oscillator behaviour at the re logon, it can be concluded, that the logon was part of deliberate act in the course of a capture scenario.

  40. @buyerninety:

    I maintain what I wrote: The MH370 primary radar data for track angle and groundspeed show undeniably that the airplane was not controlled by the autopilot between approximately 17:20 and 18:02.

    I think that in your attempt to save your theory you are clutching at straws.

    In the case of 9M-MRG the autopilot pitch mode failed 2 seconds after the start of the upset. The autopilot was manually disconnected 6 seconds after the start of the upset. After the autopilot was disconnected, the envelope protection features of the flight control system and pilot control inputs controlled the pitch and thereby the altitude and speed variations. So also in that case the variations of groundspeed show undeniably that the airplane was not controlled by the autopilot after the start of the upset plus 2 seconds.

    You then consider possible failures of equipment ands/or cabling on 9M-MRO. Are you suggesting that these failures occurrec near 17:20 and were repaired later?

    I’ve stated that my position is not based on FI primary radar altitude information. The groundspeed variations shown in the DSTG report suggest that there were also altitude variations.

    I’m not sure what you’re trying to say in your final note. The DSTG authors state a reservation with regard to the speed estimation during the first turn. That does not invalidate the speeds obtained from the radar data after the first turn. Those speeds do not indicate an overspeed condition. If MH370 encountered a “periodic overspeed condition” how would the autopilot “ameliorate” it?

  41. @David
    You didn’t give a reference for your statements, but I seem to recall this
    was covered previously – the ‘individual equipment items’ listed were not
    stated as being a complete (exhaustive) list, so it can be interpreted as
    merely an incomplete (representative) listing of the total items –
    as it says in the ‘B777-Electrical.pdf’ …”typical loads”.

    You said “Getting an IDG or the APU going will revive them.” This is not
    quite how it works. The ELMS decides whether to reenergize an item or system
    it has previously depowered, and it does this on the basis of whether there
    is at the time sufficent power available to reenergize the item or system.
    If the circumstances aboard the aircraft do not allow sufficent power to
    reenergize the item or system, the ELMS will not reenergize it.

  42. @Gysbreght
    If you believe that the estimated primary radar data for track angle
    and groundspeed show undeniably anything, then there is little point
    in anyone engaging with you on this matter.

    “undeniably” is a statement of opinion, not fact.

  43. @all

    There has been much discussion about autopilot modes versus manual flying, but not much about 9M-MRO flying itself with the autopilot not engaged and relying only on the FBW system without any pilot input to the yoke. I raised this in an earlier session and asked if anyone could provide more specific guidance on 777 lateral control laws, but got no response.

    My understanding is that in Normal FBW Mode, turning the wheel to a specific angle commands a specific rate of roll (P) around the x-axis of the aircraft. When the wheel is centered to command zero P the system maintains the bank angle. It is possible that the P rate per angle of wheel displacement is scheduled with airspeed for handling qualities. But the question is whether this FBW roll axes would automatically return the bank angle to zero if there is no pilot input to the yoke.

    Can anyone confirm that the basic 777 FBW system in normal mode without any A/P navigation mode engaged would keep the wings level with no pilot commands to the yoke? A non-FBW aircraft with A/P off would enter a spiral dive and be gone very soon.

    Clearly 9M-MRO could have flown the entire flight using the autopilot, or manually, or in some combination – this is all I have seen discussed.

    But could 9M-MRO just flown itself without the autopilot and with an unconscious?crew.

    Does the roll axis hold bank angle at zero if there are no pilot inputs? Some FBW systems will do this – but I do not know the exact algorithms used by the 777 roll axis.

  44. Let me clarify my comment. If the pilot returned the wheel to neutral after achieving a bank angle of 30 degrees, it would hold the 30 degrees and not return to zero. But if the pilot had set the bank to zero and did not touch the wheel again, it should hold zero.

  45. @Hank
    I am not expert enough, but I once proposed the PIC may have switched the A/P off and allowed the plane to continue flying on that heading. Victor indicated that would probably not maintain control in a real plane with winds, like it does on FS9. Come to think of it, I have not tried it lately now that I am using the PSS777 model.

  46. @TBill

    When flying a light plane and you take your hands off the wheel, the plane will bank and the nose will drop and this will progress until you end up in a tight spiral dive. Easy to correct by just holding wings level with the wheel or even using the rudders to hold a heading.

    One of the simplest autopilots is called a “wing leveler” because it only does that. It controls the ailerons to maintain the wings level and it is easy for the pilot to override it to make a turn. This is great in instrument conditions because it keeps the plane from getting away from you while you are checking a chart.

    This is not about course or heading – only keeping the wings level.

    I believe the 777 control system – without any autopilot navigation engaged – will hold the wings level.

    Keeping the wings level is the most essential “control” feature. You need to distinguish between where the airplane goes (navigation) versus just not flipping upside down and crashing (control).

    As long as the airplane remains level, small upsets will cause the heading to change and winds impact true course versus heading. So with a wing leveler the airplane could drift almost anywhere. It could fly mostly in a straight line or fly in a circle.

    If the 777 FBW system does provide this basic wing leveling capability, the plane would not roll over and crash, but it would turn and randomly change headings based on any atmospheric disturbance that momentarily tilted the wings one way or the other from time to time.

    My understanding is that the 777 model does provide roll rate command and bank angle hold functionality – I just can’t confirm the algorithms on the actual 777.

    If you set up your simulation to have some small level of turbulence that would allow the bank angle and/or heading to be disturbed, the bank would return to level, but the heading would remain changed. If you just let the plane fly, its course would just drift around. The actual course would depend on the winds aloft model for the simulation and level of gusts or shear that you could simulate.

    It would be interesting to see what kind of flight paths a 777 could take over seven hours by only seeking to keep its wings level.

Comments are closed.