A Couple of MH370 Things

There have been a number of interesting developments in MH370 land:

NEW MH370 PATH ANALYSIS by frequent commenter sk999 has impressed a lot of the old hands. Using somewhat different statistical techniques than the ATSB and IG before him, sk999 analyzes the Inmarsat data to assess where the plane would wind up under various autopilot modes. His results generally jibe with his predecessors’ work and add more weight to the idea that if the ATSB really believes that the plane was flying on autopilot-only, they would be better served by searching further to the north along the arc, beyond the limits of the current search box (though not north of Broken Ridge), rather than further away from the 7th arc as currently planned. What’s also notable, in my opinion, is sk999’s very clear elucidation of the problems with the routes that he assesses; for instance, he points out that all of the routes have problems accounting for speed inconsistencies in the 90 minutes between the fifth and sixth ping. These discrepencies are too large to be easily explained away as being due to inaccuracies in the winds-aloft data. Sk999’s frankness about these issues is refreshing; in the past, there has often been a tendency by those describing possible routes to adopt a position of, “Hey, here’s a route I came up with, it works really well, take my word for it.” (I’m probably as guilty of this as anyone.)

NAJIB IS IN TROUBLE and at last it looks like he may have to go. Is it possible that his ouster will lead to disclosures about what really happened in the aftermath of MH370’s disappearance? In a report last year, ICAO offered an uncharasterically harsh assessment of Malaysian government interference in the search process. Among their most glaring sins: allowing the search to proceed in the South China Sea for a week even though the military had spotted the plane turning toward the Andaman Sea the night of the disappearance; refusing to pass along crucial Inmarsat data to Australian officials who were tasked with searching the ocean for the plane; and lying about the determination that the flaperon had come from MH370 (it did, but that hadn’t yet been determined at that point). What the heck??

THE ATSB zinged airline pilot Byron Bailey, who wrote an error-filled article in the newspaper The Australian arguing that the only possible explanation for the disappearance of MH370 was pilot suicide. The ATSB had never before gone after an article in such detail before; they didn’t even touch Clive Irving’s piece in the Daily Beast, which was much worse (but which, on the other hand, was friendlier to the ghost-ship scenario that the ATSB still favors.) Personally I think it’s great to see the ATSB engage with the media coverage in this way; there’s too much nonsense about MH370 being peddled in the general media. Bailey responded to the ATSB critique with a second piece in The Australian.

THE ATSB ALSO perked up my ears with their response to an inquiry from reader Susie Crowe, who asked ATSB spokesman Dan O’Malley whether the Australians had received information from the French regarding their investigation into the Réunion Island flaperon. O’Malley replied, “The ATSB looks forward to receiving the report on the flaperon from the French judicial authorities, once it is completed.” In other words, Australia is spending over $100 million in taxpayer money to dispatch search crews to one of the most difficult and dangerous stretches of ocean in the world, and the French have not even shared with them information about the flaperon that might indicate whether or not they are looking in the right place! To which I might add: !!!!!!

183 thoughts on “A Couple of MH370 Things”

  1. Does this flight mode work for very small bank angles (0.1deg)? A threshold of 5deg is mentioned and a bank angle that size means a 12deg/min turn and the aircraft goes round in circles. Or am I missing something?

  2. @Richard Cole:

    The threshold of 5 degrees applies for the autopilot roll mode ATT. What RetF4 describes applies to the plane flying with A/P off, controlled only by the fly-by-wire flight control & stability augmentation system described in FCOM Chapter 9 Flight Controls.

  3. Gysbreght, RetiredF4,

    Re EY440. I have converted data into UTM-47 projection and zoomed to the loop area. The loops are egg-shaped, not round. Here is the result:


    I have also made a plot of ground speed vs heading. The assumption of constant airspeed of 456 kts gives a very good fit at wind speed of 26.5 kts. However, such a wind speed would result in the center displacement by ~50 km over 1 hour. Obviously this did not occur. The altitude was also constant during this period.

    So, what was the lateral navigation?

  4. Oleksandr:

    Just guessing again after some reading in the A330 FCOM:

    ATC directed the flight to hold at an assigned waypoint.
    Pilot selected the HOLD page on the MCDU and engages NAV mode.
    When there is no defined holding pattern in the database for the assigned waypoint, the system proposes default holding pattern data.
    Default time on outbound leg is 1.5 minute (above 14,000 ft)

    Does that fit?

  5. Oleksandr,

    I’ve no idea why the north-bound and south-bound turns have different radii after the ‘teardrop’ entry into the holding pattern. According to the Wiki article referred to earlier, ICAO mandates holding turns at 25 degrees bank for airspeed greater than approx. 170 kTAS.

  6. Gysbreght,

    At a glance it appears there are no way points inside of the loop area – I will check this accurately.

    But I have a difficulty to decipher the egg-shaped nature of the loops. I think it is somehow related to wind. If airspeed is fixed and bank angle is fixed, then turning radius would be constant, and hence the center would drift. But it does not happen, meaning that bank angle is varying.

  7. Oleksandr,

    A holding pattern is usually defined relative to a ‘fix’ at one of the four start- or endpoints of the turns in the ‘racetrack’ pattern.

    “bank angle is fixed, then turning radius would be constant, and hence the center would drift.”
    Not necessarily. Either the pattern designed by the FMGC is flown automatically by the A/P in NAV mode, which would compensate for wind to follow the designed ground track, varying bank angle, turn rate and radius as necessary to prevent ‘drift’, or the pilot could have seen this as a welcome opportunity to maintain his hand-flying skill, having nothing better to do while waiting for clearance to proceed.

    As I said, I’ve no idea why the radii are different. The parameters that define the FMGC-designed holding pattern are the ‘fix’, the course of the inbound leg, the turn direction (L or R), and the time for the outbound leg.

  8. Nothing suspicious there.

    The anchor point of the holding can be any coordinate given or selected. In this case it is imho the upper left crossing point. From this point the holding direction looks like 070° (+ or – 10°).
    The wind is blowing from the upper left corner to 142°, therefore the groundspeed is highest on this heading. The holding is flown with constant bank angle and constant IAS or Mach, drift is correct by correcting the opposite leg, which is the leg on the bottom of the pic. That is the reason that it differs in heading.
    It is now obvious that the turnradius on the right side has a tailwind component and thus a bigger turn radius, whereas the turn on the left side has a headwind component and thus a smaller turn radius. I´m sure you guys can compute this bank angle somehow.

    The entry leg like shown is in order to cross the holding fix at the first time with the correct outbound heading. When cleared to exit the holding, you can proceed directly to the holding fix, which the last turn shows.
    Sure this is all guess work based on my expierience from some year ago, some details might have changed. We had to do those holds without the help of the “puters” as we had none, on a basic HSI and flown manual.

  9. RetiredF4,

    Your explanation makes a lot of sense, but would imply that the pattern was indeed flown manually, without the help of “puters”, each turn at constant bank angle (probably less than 25°). In that case the hand-flying was quite accurate.

    As to the “anchor point”, I would expect the pilot navigates towards it, then starts turning to join the pattern. That would put it at the south-west ‘node’ of the racetrack. But I’m not a pilot, just guessing.

  10. @Gysbreght

    “In that case the hand-flying was quite accurate.”

    as expected from experienced 777 pilot eh

  11. @ Gysbreght
    Let´s assume the upper left crossing point is the designated fix. A direct approach to the fix would bring the ship during the turn into the holding well outside the holding airspace, therefore a teardrop entry within the holding airspace is done.The lower point as fix makes no sense, as there is always a outbound leg from the fix and not an immidiate turn.

    I´m not sure it was handflown in the pure sense, more something like : The speed is controlled by the autothrottle, those jets are almost never flown without it. The nav gadgets present the holding pattern once entered pretty nice and for the holding entry you just track to the inbound leg opposite to the fix(like in the graph) and initiate the right bank. The system most probably displays the correct point to initiate the bank. After that its just banking and rolling out at the appropriate points. The bank is maintained by the FCS.

  12. RetiredF4,

    Thanks for your reply. I agree that speed would be controlled by the autothrottle, if only for reasons of passenger comfort.

  13. RetiredF4,

    Remember that this is probably not a normal holding area within protected airspace. I imagine that ATC instructed the flight to proceed to XYZ and to hold there until further notice. Would ATC have provided the holding pattern parameters to be entered in the FMGC? If not, what would you do?

  14. @Bruce

    You said:

    “This is certainly music to my ears. I’ve been a strong proponent of a curved flight path for almost two years now but haven’t gained much traction. I’ve heard nothing but True This, Magnetic That, and Waypoint-Waypoint-Waypoint.

    It’s long past time we left the SIO.”

    I feel your pain. No matter. it is what it is. Does not matter how much lipstick you put on a pig it is still a pig. The IG is a bit of an enigma to me. They pride themselves on fact based analytics, and then conclude that the constrained AP flight dynamics are a fact. Very strange. Maybe someone appointed them as fact arbiters, and I missed the memo.

    While I applaud the role they have played in being a beacon for the physics of the problem, I have lost a lot of respect for them relative to problem solving.

    Hard to recover from a pin in the map at 38S with your name on it.

  15. Gysbreght, RetiredF4,

    So far none of us is correct. There is one waypoint, GUNIP (99.530556E, 4.4980556N), but it is either loop entry point, or the last navigation point before entering the loop:


    RetiredF4, just assume for a moment aircraft knows nothing about about wind and sets constant bank angle. Airspeed is constant. That result in a constant turn radius in the relative coordinate system, which moves with the wind. In this coordinate system the shape of loops would be perfectly round. Now move back to the fixed coordinate system. The shape would be distorted, but the center would drift with the velocity of wind. This does not occur, meaning bank angle was adjusted.

    Is there any mode, some kind “go around”, which returns aircraft to original location after manually predefined/set time? Combined with the fixed airspeed, this would result in predefined distance to fly to reach original location.

    The other observation: EY440 flew via GUNIP fairly accurately; deviation is a way smaller than 1 km. In case of MH370 the deviations from VAMPI and MEKAR were much larger: 5 to 10 km. Thus it is more likely that MH370 did not follow these waypoints. It either flew “straight”, or it was heading to some other waypoint further.

    Other thoughts on EY440?

  16. StevanG,

    Why would an experienced pilot be manually cycling at FL360 for almost 1 hour? And why passengers did not start asking questions? And why does media keep silence?

    Earlier I was asking you and Dennis whether you have any motive in mind. Any thoughts?

  17. StevanG, Dennis,

    In the previous post I meant thoughts on such a bizarre route diversion, but not subsequent need in refueling after spending 2-3 hours wandering around the Malay Peninsula. These 2-3 hours at 450kts would be sufficient to reach Abu Dhabi in case of a regular flight.

  18. Oleksandr,

    “That result in a constant turn radius in the relative coordinate system, which moves with the wind. In this coordinate system the shape of loops would be perfectly round. Now move back to the fixed coordinate system. The shape would be distorted, but the center would drift with the velocity of wind.”

    Perhaps this is how the Flight Management & Guidance Computer (FMGC) calculates the holding pattern. Doesn’t seem to be a computation that a pilot could do easily without a “puter”.

  19. @Oleksandr
    The distortion of the pattern and the different turn radii develop, because the outbound leg from the fix is flown wind corrected, while the turn itself is not, as it is is flown with constant bank. The inbound leg to the wind is again flown wind corrected, but again with comstant bank. Relative to the air the turn radiii would be the same, but not relative to the ground, as the groundspeeds differ in the right turn to the south (tailwind component, highr groundspeed) and the right turn to the north (headwind component, smaller groundspeed).

    But the discussion gets academic. It does not mattter wether the hold was at a designated fix or just a coordinate, wether it was flown manual or on autopilot, and wether it was exact or somewhat off. Holding airspace offfers a degree of freedom, as long as you stay on altitude. The hold itself has nothing to do with MH370.

  20. Gysbreght,

    I also think so. Another thought is that the first loop was done manually, and then pilot manually or automatically followed GPS-recorded track for several times.

  21. RetiredF4,

    The distortion can be explained by what you wrote. But how to explain that track was nearly constant during subsequent loops? And how to explain minor differences, unless the shape of trajectory is affected by minor variations in wind field (i.e. path was computed in real time based on actual wind conditions or actual position)?

    The apparent relevance to MH370 is a potential proof of another stable flight mode, which allows for curved paths. Earlier I asked IG to comment on this, but so far none of its members responded.

  22. @jeffwise
    Jeff, I’m guessing you have received no response from Etihad regarding EY440?

  23. @Oleksandr
    As priviously explained by Gysbreght I think,the holding airspace is defined by a fix, from which the aircraft tracks outbound (track not heading), the holding speed (IAS) and the turnrate = bank angle. If no holding is published in this area the instructions of ATC will include some details and the crew will hold like they are used to (standard holding procedure).
    At least after a complete holding the crew or the computers will update the navigation and try to hit this fix at the exact spot.

    Therefore any subsequent holding is started at the correct position. The complete flightpath over ground differs somewhat. There are some reasons for this, manual navigation, manual flying, TLAR (that looks about right) method used, changing winds in strength and direction…….
    But it does not matter. They flew well within the holding airspace.

    As to the mentioned relevance to MH370: We know that there are many possible variables involved concerning the navigation, modes and disturbances which could have affected the flightpath of MH370. I do not see your point that micro measuring this holding pattern will make any difference.

  24. @Oleksandr

    I really don’t know and I don’t want to waste my time thinking about this flight because I think it has nothing to do with MH370 at all.

    If I had to make a guess then it would be a very broad one – someone fcked up being it pilot ATC or a third person.

  25. Mr. Wise, I am not an aviation expert nor have anything to do with aviation but I have been following all the developments in this case from day 1(mostly). Now, what I fail to understand is that you find articles from Byron and Clive bad and worse respectively, but what about your article about Russian theory published earlier in 2015. What category does that belong to? If you ask me as a reader, that was worst.

  26. @Ashish, Why, where there specific inaccuracies about it that bothered you? Or did you just have a gut-level reaction that you didn’t like it?

  27. Putin is going to us Malaysia to fly a EMP to u.s and have a low altitude for a crucial effect on our country it’s the only thing that makes sense

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.