295 thoughts on “Potential Terminal Locations for MH370”

1. Rodney Thomson thinks that analysis of the underwater ping signals could suggest a location for a ULB within 500 meters. (He has, or had, 90% confidence that the pings are from a ULB based on his analysis of the audio track of a video posted on youtube.) He also says he is familiar with some of the JACC folks and thinks they are capable of doing such an analysis.

   http://iheartmatlab.blogspot.com/2014/04/automatic-detection-of-mh370-ulb-pings.html

   However, I have not seen any suggestion from JACC or elsewhere that the Bluefin search area is being determined by such an analysis. Instead from reports like this one from the WSJ, it appears that Inmarsat is defining the search “sweet spot.”

   1. Wrong again! From a post 8 hours ago in the WSJ:

      PERTH, Australia— The underwater search of the seabed for debris from Malaysia Airlines Flight 370 is expected to take less time than the previously forecast 6 weeks to 2 months, due to analysis of acoustic data.

      The underwater search area had been significantly narrowed through the detailed acoustic analysis conducted on the four signal detections made by the Towed Pinger Locator on ADV Ocean Shield, Australia’s Joint Agency Coordination Centre, which is leading the multinational search operation, said in a statement.

2. Here is another way to solve the problem that comes to the same conclusion for the north route without needing to make an assumption about the BFO data:

   Suppose we don’t believe we have properly extracted the LOS speeds from the BFO data. Instead, we look at the current search area in the Indian Ocean and find the straight, constant speed path between Duncan’s coordinates at 18:29 and the search site. Once we have this path, we extract the ping radius and the LOS speed at each ping time. We are essentially reverse engineering the data that Inmarsat used without specific knowledge of how to interpret ping latencies and the BFO Doppler data.

   Now that we have the LOS speed and ping radius at each ping time , we find the north path that also satisfies the same LOS speeds and ping radii (ignoring the LOS speed at 18:29 because the plane might be turning). Guess what route you find? The exact same north route that I originally derived. Which has to be the case because using my methodology to find the north and south routes, I predict the last ping occurs very close to the current search zone.

   And that north path at 243 knots ends very close to the Qamdo Bamda Airport in Tibet.

3. Duncan,

   Can you generate a master list of the knowns (or at least identify a string under different categories like reddit) so they can be discussed and vetted? You’ve done a good job with the ping rings. The track, time and altitude to loss of contact is known. The 1815 radar location was given. The altitude of 12000 was given early (Now hearing 45000 feet is being listed which above the listed service ceiling… an altitude which takes a long time to get close to and very hard to control) The initial fuel is reported as 49,100 kgs. Et Cetera.

   You could solicit fuel info from 777 info from commercial pilots. You can solicit team formation to go on Reddit (reduces moderation time delays) and report back findings to the group. This could lead to discussions on possible routes rather than all of the possibilities.

   Your thoughts?

4. Dear Duncan,
   Have you considered the possibility of a northern course across Myanmar at low altitude (to avoid radar detection), and then intersection with air route L888, and following L888 to the northwest until termination of the flight?

   1. Everything considered is in the posts and comments…

5. Duncan

   Aside from the Inmarsat data, there are other news reports with asserted “facts” that seem, if correct, to be provide potential clues. I am wondering if these can be tested now with any known data.

   To me, the two most significant ones are the “claims” that the plane spent 23 minutes at up to 45,000ft.after its initial turn and the claim that the co-pilot’s cell phone “registered” or was searching for cell service for roughly 30 minutes including a cell phone tower in Penang, Malaysia.

   I am wondering whether your group of reader’s and poster’s can provide any science or fact-based comments on these two claims.

   I have read one of your posts where a pilot conjectures that the climb to 45,000ft could have been inadvertent while dealing with an in-flight emergency. However, staying at this altitude for the reported duration of 23 minutes makes this seem less likely. Is there any known information that makes the claim of 45,000ft believable or just conjecture and similarly, the 23 minute duration at that altitude? Accepting these reports as true and the reported speed based upon the radar “data” after the turn, where would this claimed flight segment of 23 minutes have ended?

   The new news report of a 30 minute period (which by necessity means a stretch of more than 100 miles) raises the question of whether the plane, based upon the published radar track was ever close enough to multiple cell towers for a 100+ mile stretch. How does this correlate with the radar “data” of the flight’s altitude as it cross the Malay Peninsula as I understand that above a certain altitude no cell tower would be contacted? Also, is it believable that all the other passengers had turn off their cell phones and so that not one of those was not similarly “searching” for cell service?

   I recognized that I am referring to news reports un-verified “data” and I do not recount them as facts but it seems to me that it is possible to compare these unverified reports to see if there is some truly fact based item that makes these claims mutually exclusive such that working hypotheses can be considered for the earlier part of the flight.

   Thanks

   John

   1. Thanks John.
      What you say is good. Yes, any media reports might lead one to useful information. It’s just – as you recognise – that one should not accept all (or any) such reports uncritically.

   2. The way I heard it was that the plane climbed to 45,000 feet, then dropped suddenly to 23,000 feet. That would be plausible if there were a sudden upset that disconnected the autopilot.

      Consider this scenario.

      1) the previously damaged wingtip bends and/or breaks off.
      2) this induces a sudden roll toward the damaged wing
      3) in response to this roll, the autopilot disconnects and the plane begins to climb (TWA 800 initially began to climb after it started to break up. And AF 447 initially start to climb when its autopilot disconnected).
      4) under a sharp, sudden climbing turn near the service ceiling, the engines flame out.
      5) with the engines stalled, the plane loses both electrical power and cabin pressurization.
      6) with no electrical power generation, the plane automatically begins electrical load shedding, dropping off the transponder for example.
      7) the plane aerodynamically stalls and begins to dive.
      8) the ram air turbine automatically deploys allowing the pilot(s) to re-start the engines.
      9) with the cabin having been depressurized, the passengers suffer hypoxia.
      10) with the cabin depressurized the pilot(s) go to the oxygen masks, possibly too late, because they prioritized regaining engine power and flight control over personal safety.
      11) with the engines running again, he levels off
      12) suffering the effects of hypoxia, the pilot(s) do the best they can to try and reach the nearest diversion airport in the Strait of Malacca.
      (see also high altitude cerebral edema http://en.wikipedia.org/wiki/High_altitude_cerebral_edema )
      13) with the mask on and suffering from the effects of hypoxia, the pilot mumbles in response to a radio hail by a Tokyo bound flight 30 minutes ahead of him.

      It is not a high likelihood scenario, but it is based on the nominal “facts” as reported shortly after the plane went missing. It does not explain the events over the Strait of Malacca, which would seem to require some additional efforts by the pilot(s) before they lost consciousness and the plane continued on autopilot to termination of flight.

      Do you have a better scenario to cover the climb, dive, mumbling radio transmission, loss of transponder etc?

      At least it is based on a fact, the aircraft’s wing had previously been damaged and repaired. A damaged plane having a structural failure due to inadequate repairs is not unprecedented, see Japan Airlines flight 123.

      1. A good question would be why would the repaired wingtip fail? Using the Aloha Air accident as a model, we could consider corrosion as the culprit. Malaysia is a very hot, humid, salt air environment. That is conducive to corrosion.

         We don’t need any elaborate terrorist plots. Just old fashioned corrosion eating away at the mechanical integrity of the wing.

6. @ Alain,

   Many thanks for the update on fuel consumption. I stand corrected.

   I had mistakenly assumed that at 480 knots and 35,000 feet the aircraft engines would be working efficiently at near 100% throttle and a good Thrust Specific Fuel Consumption (TSFC) and at 250 knots and 5,000 feet the engines would be working inefficiently at less than 50% throttle and a poor TSFC.

   Richard

7. It is hard to know the reason for this course that is nearly due north. Default values of autopilot are a good guess, and I am not abandoning this possibility. I think choosing a route that minimizes the number of countries that are flown over might be another. Does a slow north trajectory minimize the chance that a radar operator is alerted? Clearly, there is evidence to minimize radar detection.

   But don’t you think it is a bit of coincidence that the due north course ends right over an airport? The longest paved airport in the world with few flights per day in a very remote part of Tibet? Do you think a crash this close to the airport would go undetected?

   1. I did not start with the assumption that the landing was at an airport. Rather, as I began to use Google Earth to find a potential crash site based on my predicted north path, I stumbled upon the airport. In much the same way, when I mathematically calculated the northern path that matched the Doppler data, I stumbled upon a near due north heading. The math is leading me there without assumption other than a direct path (rhumb line or great circle) from your coordinates at 18:29, constant speed, and matching the (derived) LOS speed data.

      I know you have been overwhelmed lately. I think you are wise to allow yourself the time to rest and think clearly.

      Victor

      1. Thanks Victor. One might also mention that we had reasons to suspect a due north route before you arrived at your predicted north path… That is an argument in favour of your scenario. That is, your preferred northerly route was not predicated on other things such as a default 360 degree heading for the autopilot.

   2. Yes. Also, maybe the path to the satellite was partially obstructed, resulting in a partial ping. The elevation angle above the horizon was about 42 degrees for the satellite at this time. The path might have been blocked by a mountain during descent or a building once on the ground. I would think that any satellite communication would require an unobstructed LOS.

   3. Victor, as you commented before, “The north path solution ends in the Tibet mountains, about 80 nm from the Qamdo Bamda airport.” There appear on Google Earth to be 2-3 uninhabited ranges of the Kangi Garpo Mountains between your end point and the airport, so I can easily imagine a crash there going undetected.

      I have not heard of any searches in Tibet. I speculate that China would prefer that the plane not be found than be found in China because it would reflect poorly on its military radar, analytical capacity and criticism of Malaysia. I don’t see a motive for flying to Qambo. There is precedent for “zombie” flights, and I infer from Alain B.’s comments that the B777 is particularly capable of such.

      Aviation disasters are more commonly the result of malfunctions and/or errors than sinister intent and here what little we know seems to me to fit better with the former.

      To me the main issue with the autopilot northern route theory is altitude, because we have reasonably possible explanations for default heading and speed. Annette has suggested the possiblity of a default altitude of 10,000 feet. I do not see a route on Google Earth where a plane flying lower than 17,000 feet can get to your end point. At 10,00 feet it looks as though it would hit the mountains at about 29 degrees,13 minutes N in the vicinity of the Dibang Wildlife Sanctuary in India.

      As Duncan has been saying for weeks it seems, testing the autopilot theory cries out for input from someone who understands B777 autopilot default/reboot operation.

   4. Dear Victor et al, I was nosing around online, and found this list of details about Chinese/Tibetan airports — including their PCN numbers, which have been surprisingly hard to find. http://airfieldcharts.com/chinaairports.htm If you click on each airport code it gives more detailed info. The 777 ER has a medium PCN of 85 on a rigid surface. http://en.wikipedia.org/wiki/Aircraft_classification_number As far as I understand it, Qambo Banda has a rating of RB32, so it has a rigid surface, with a medium subsurface, which is fine. But it appears to only has a PCN tolerance of 32 which is very low — I guess it could be a typo, but all other sources are locked. http://en.wikipedia.org/wiki/Pavement_classification_number . As far as I understand, the ACN has to be equal to or less than the PCN. Even with an emergency landing, the difference between the 2 is so great here that the aircraft, (and runway) would be unlikely to survive the landing, let alone be intact enough to hide in a hangar, or re-fuel and take off again. But other airports nearby might be fine (Chengdu for example is RB98)– so I am sending you the list so you have something to refer to. I hope the links show up as live links, if not, I apologize — I am doing this with my left hand and a sleeping baby in the other arm!

8. Duncan, you queried the default altitude:-
   Altitude window of the Mode Control Panel on page 12/13 of the B777-Automatic_Flight document:-

   “Displays 10,000 feet when power is first applied”

9. Any southern path will lead to a crash in the water. No debris on the ocean surface has been found. Because of the complexities of underwater searches, to find the black box in any reasonable amount of time will require a very precise estimate of the actual crash site; hence, the considerable effort by Inmarsat to refine the Doppler data. A search for a land-based termination does not require this level of precision.

   A northern path can either be terminated with a crash or a successful landing. If there was a crash, at some point debris would be found. I do not know if a thorough search of potential crash sites for northern paths has been performed; however, the longer it is that we do not find debris on land, the more the evidence points to a successful landing.

   According to my calculations (which I have not seen refuted), a path almost due north from Duncan’s estimated coordinates at 18:29 matches the Doppler data derived by Mike Exner very well. My predicted southern path matches very close to the current search area. This leads me to believe that Inmarsat is using a similar set of assumptions as me (even without final refinements that they are applying to pinpoint the exact southern crash site).

   Much the way I believe we cannot discount a northern route, I think we cannot discount an intended destination for the northern route. I believe it is more than a coincidence that my predicted southern route ends in the current search area. I also believe it is more than a coincidence that my northern route ends at Qamdo Bamda Airport in Tibet, a long runway in an isolated part of Tibet with very few flights a day. Runway heading 320.

   In light of all of this, I think we have to seriously consider the possibility of an intended and successful landing at the Qamdo Bamda Airport, especially if we see no crash debris for the northern or southern paths. This is where the math and logic are leading me.

   1. Victor: I read and understand what you are saying.
      For me, the jump to Qambdo Bamda is too much. You are right that it could be a remarkable coincidence. I might count as an argument against it the due north route! That is, if the aircraft were under human control, I might anticipate some other heading being taken (i.e. the aircraft would not begin at a point due south of that airport and over five hours flying time away).
      But I really hope that I am wrong.

   2. One can not discount the reports early on that RR said their data showed a full and proper shutdown of the engines. That was quickly discounted by the media, but never proven to not be true. Then there is the fact that Boeing never said that the statements from McInerny, that the plane was in Pakistan were not true.

      Sometimes you have to look at the early data and reports and try to find the truth before it is covered up.

10. WRT to the WSJ statement:

    “Investigators relied on the principle that the satellite’s temperature distorts the way it receives radio waves. The investigators analyzed changes in the temperature of the Inmarsat satellite on each occasion that the plane made contact and then made calculations about the jet’s trajectory, according to Warren Truss, Australia’s deputy prime minister.. Investigators also incorporated aircraft performance calculations from Boeing to fix the sweet spot, according to Mr. Dolan, developing the most precise projection yet of the jet’s probable point of impact.”

    …Once again, the press takes useful engineering information and tries to convert it to English, but in the process, distorts the information almost beyond recognition. In this case, the statement “… the satellite’s temperature distorts the way it receives radio waves….” should be read: “The spacecraft payload (transponders) master reference oscillator, from which the outbound and inbound LOs are derived, has a finite temperature coefficent which can be calibrated with the C band pilot subsystem. By measuring and taking into account the oscillator temperature coefficent, the range estimates and Doppler values can be fine tuned.” This is like translating English to Chinese and back to English, but I am pretty sure that is what the WSJ would say if they were engineers talking to engineers. That’s a logical refinement that should improve the estimates at the margines.

11. Take it all cum grano salis.

    How appropriate that the stock market values and trends of the companies were included in that article.

    This is (or should be) all about the lives of the passengers and crew, and the emotions of their families and friends; and also finding the flight recorders so that an understanding of what really happened can be developed, so as to try to minimise the chance of it happening again.

    My own opinion is that the media (including the WSJ here) has failed badly in not trying to question the ‘official’ investigators, instead lauding them as “experts” when the reality is that their ineptitude has been evidenced to all.

12. Duncan
    With your permission, I will make a break in the technical analysis to come with some basic reallity.
    – I guess my scenario is right until Malacca straight (Electric problem at IGARI and decision to land in emergency to Langkawi – for us, pilots, it’s so obvious).
    – I have no one reason NOT to believe Inmarsat reports and all your calculations based on these data (what a job you did !)
    In this case it’s remaining one thing to explain : The Heading.

    Philosophy of the evolution in the aviation field
    Today, industries want to manage the world using more “machines” (robots, computers, softwares ..) than “human capabilities” with their brain.
    In aviation, 20 years ago decisions had to be taken in regards of that.
    Pilot will become:
    (1) a “push button” (computerized flights with the help of human) or
    (2) a “driver with the help of new computers and technologies”.

    Boeing took the decision (1) with the B767 which was a trial. And definitively with the B777. I can say that because I was flying B737-series then B747-series and when we saw the B767 , it was a revolution for us.
    Today the B777-series has the most sophisticated autopilot born in the brain of computer system engineers (thus who don’t fly everyday). On the paper it’s fantastic, however when you think about it’s stressing. Today on B777, there is NO possibility to be out of the “autopilot control” even if you put it “officially off”. Proudly they call that “Uninterruptible Autopilot System project” and B777-series are the first step.
    Imagine dear Duncan being in your car at 200 km/h, the night, with heavy rain, no wiper and you can’t control anything … and a little voice say you : “Don’t worry Duncan, the software controls the situation”.
    From their side, Airbus with the same capabilities, decided to let the final decision to the pilot … until today! You can really fly a A 380 if you don’t want to use the autopilot.

    The question is : The heading has been changed by a human or by a software !?

    1. As always Alain, your input is vitally important.

       What you describe about ‘sophisticated’ new aircraft looks very similar to what has happened with cars/automobiles. That can avoid some sorts of accidents; but it can cause others.

       As you wrote, the core thing to explain is the heading.

    2. Hi

       Is there any evidence that Boeing has had their “Uninterruptible Autopilot” as described in their patent application implemented and certified for use by the FAA/EASA for airliner use?

       Just imagine some of the issues the FAA/EASA would have in terms of certification of such a system. Ground based hijackers hacking into the system to control the aircraft etc.

       Cheers

       1. The answer to your question is ‘no’. This has been covered in various comments and replies. Search for input from LGHamilton.

Comments are closed.