Since October, 2014, the search for MH370 has been guided by the assumption that sometime after it disappeared from Malaysian radar screens over the Malacca Strait it turned south and flew straight and fast into the Southern Indian Ocean on autopilot. The ATSB, which is conducting the search, has always been agnostic as to why exactly the plane would have done such a thing—maybe the pilots succumbed to hypoxia, or fire, or committed suicide—but the underlying assumption is that the plane would have flown its last few hours without control from a human being: that is to say, it flew as a “ghost ship” until it ran out of fuel shortly before 0:19 on the morning of March 8 and spiralled into the sea.
Analysis of the satcom signals received up to that point, combined with understanding of how 777s fly, indicate that a “ghost ship” plane should have wound up somewhere in a box 40 nautical miles wide and 400 miles long. As I’ve described earlier, the highest-probability areas of this box have already been searched and no aircraft wreckage has been found.
Previously, I’ve suggested this means that the plane did not fly to the current search area. On January 5, several members of the Independent group published an article on Duncan Steel’s website that agreed with this premise:
we now have a new piece of information. Simply: the aircraft has not been found within the priority search zone. If that continues to be the case then we must consider other possibilities which might conform to the known data (Inmarsat BTO and BFO values, and the fuel limits which can work either way, either setting a range limit or else requiring fuel to be burnt more quickly per unit distance) and lead to a revised end-point for MH370 that is outside of the search zone, and north of it (given that the fuel limitation prohibits end points further south).
The question I’d like to address today is whether the absence of MH370 from the current search area means that the plane COULDN’T have flown to its endpoint on autopilot alone. The reason such a suspicion might arise is that to reach an endpoint north of the current search box the plane would have to have flown a course that was either curving steadily to the left, or slowly decreasing in spead, or a little of both. But the 777 autopilot cannot be programmed to fly in a curve or to steadily decrease the thrust of the engines.
At first blush, then, the answer would be: no. MH370 couldn’t have flown to its endpoint without a human at the controls. That means that one of three things might have happened: 1) The perpetrator took the plane on a slow, curving course to the northeast; 2) The plane hit the 7th arc over the current search area but held it in a glide so that it wound up beyond the current search; or 3) The plane was commandeered by someone who managed to spoof the signal so that it wound up going north instead of south. The first two scenarios presupposes a suicidal pilot, most likely Zaharie; the third requires demonically clever perpetrators. Which of these scenarios is more likely should become more apparent if and when we get to see the results of the examination of the flaperon held by French criminal investigators since its discovery on the island of Réunion last July.
Coincidentally, this would also rule out “hero pilot” scenarios that have remained proven remarkably popular despite the vast weight of evidence against them.
However, the case is not closed.
Our own demonically clever Victor Iannello points out that there is a way to program a 777 autopilot such that it takes a curving path of variable speed. He points out that if MH370 loitered near the Andaman Sea for awhile, its autopilot could then be set to follow a magnetic heading as the plane descended at a steady, one-tenth of a degree flight path angle. As it descended with its engines at constant thrust, the increasing thickness of the air would have caused its speed to decrease. And the winds aloft, combined with the steadily changing magnetic declination, would have caused its path to curve. In such a scenario, the plane could have wound up more than 500 nm from the search box. (See image and table below).
Now, I should hasten to point out that Victor doesn’t think that this actually happened. He believes, as I do, that such a solution is incredibly arcane, and it’s hard to imagine why anyone would want to program a plane’s autopilot is such a way. But it’s not correct, technically, to say that a ghost-ship scenario is impossible.
By the way, several flaws with the suicidial-Zaharie scenario have been pointed out before, but I’d like to add a new one. If Zaharie wanted to disappear, he could have done so more safely and effectively by waiting until BUNTA (the boundary between Ho Chi Minh and Sanya AOR) or IKELA (at the boundary between Sanya and Hong Kong) to go dark and then head out over the open ocean. (You can see the route of the originally filed flight plan here.) From either spot he could avoid flying over land, with the risk of detection and interception: there’s a 200-mile wide gap between Taiwan and Luzon, and beyond that lies the open Pacific beyond. If you really want to vanish, there’s no beating the Marianas Trench.
To my mind, the only reason you’d head west at IGARI is because there was someplace specific you wanted to go.
139 thoughts on “Can We Rule Out a Ghost-Ship Endgame for MH370?”
@carla – it may have been possible to remotely fly it to the north hitting all the arcs perhaps using Victor’s previously mentioned hack/spoof of antenae orientation.
I just spotted a complaint of a passenger of the abnormal EY440 on Facebook. Original complaint is in German. His name is Xaver Soellner. We need to bring him here.
Add https in front.
Can’t see it, can you copy it here?!
@Oleksandra, Like @StevenG, I can’t see this page. Perhaps you could somehow take a screenshot and put it in Dropbox, or something? Or email to me and I’ll do that.
@MH and @Carla, The oft-noted problem with a remote-control hijacking is that the plane had no external communications between 17:21 and 18:25. So, we can be pretty sure that didn’t happen.
@Nederland: The paper at Duncan Steel very much underplays this point, but none of the various alternate autopilot modes work without a human pilot; for instance, in the magnetic heading mode, you’d need someone to manipulate the engine thrust settings to vary speed. (The exception being Victor’s very special case described in the post above.)
@jeffw – a fully awake pilot who landed the aircraft is still on the table !!
Trying one more time to post the link:
@MH, Yes, very much so.
@Oleksandr, No joy, thanks for trying again. Maybe it’s block for users from the US.
If link doesn’t work by click, you still can copy/paste it into browser. This does work for me.
I can repost the conversation between the passenger of EY440 and “template” response from Etihad if needed. Anyhow we know name of one of the passengers, who can be contacted, which is great.
He wrote this case deserves media attention, so I think Jeff is just the right person.
We need details of this flight, as much information as this passenger can recall. What announcements were made, when, by whom, whether inflight maps were working, whether passengers noticed cycling over Malacca.
No success, link is broken or something.
I wrote this guy a message 5 minutes ago on FB. Hope he replies. I´m German.
@LouVilla, @Oleksandr, Thanks for getting on this, guys. I’d be more than happy to reach out to himy.
This is conversation between Facebook member Xaver Soellner (facebook.com/xaver.soellner) and Etihad Airways with regard to EY440.
I hope I do not violate copyrights etc.
Wahnsinn was wir heute mit euch erleben mussten! Erst fliegt der Pilot in die falsche Richtung! Statt von Ho Chi Minh City über Indien nach Abu Dhabi geht es Richtung Malaysia und der dortige Flugraum klärt ihn dann gerade noch rechtzeitig auf bevor er weiter Richtung Südpol unterwegs ist. Kleiner Tipp: nächstes Mal nach dem Start kurzer Blick auf den Kompass und nicht erst nach zwei Stunden, daran lässt sich leicht ablesen wohin man unterwegs ist. Aufgrund dieser komplett sinnlosen Schleife langte der Sprit dann nur noch so gerade nach Mumbai. Wir mussten teilweise offensichtlich recht tief dorthin fliegen, so dass es ziemlich geruckelt hat. Was für ein schreckliches Gefühl, wenn einem der Flugkapitän gerade gesagt hat, dass das Benzin nur noch bis Mumbai reicht. Nachdem dann der Tankstop dort drei Stunden inklusive komplettem Crewwechsel gedauert hat, haben wir natürlich den Anschlussflug in Abu Dhabi verpasst. Dafür gab es dann einen Frühstücksgutschein für den Gegenwert eines kleinen Kaffees. Wow Danke! Der war an einem einzigen Café einlösbar, das wir dann selber suchen mussten. Und natürlich bekamen wir zwar ein Ticket für einen Weiterflug allerdings über Istanbul (statt für den nächsten Direktflug), der uns dann 13 Stunden später an unsere final Destination München gebracht hätte. Nach ewigem Hin und Her bot man uns eine Alternative über Rom, die uns nur 10 Stunden gekostet hat. Wow ein erneutes Danke! Trotzdem hatten wir noch Glück, weil das männliche arabische Flughafenpersonal in Abu Dhabi mit mir als Mann wenigstens verhandelt hat. So viel Glück hatten zwei allein reisende Frauen nicht. Die wurden einfach immer nur müde belächelt und dann wieder weggeschickt. Dass auf dieser schier endlosen Schwachsinnsreise natürlich unser Gepäck nicht in München angekommen ist, brauche ich eigentlich nicht zu erwähnen. Hat aber auch sonst niemanden so richtig interessiert. Drei Mal hintereinander Airline des Jahres kann man sich nach so einer Erfahrung schlicht gar nicht vorstellen!!
Etihad Airways Flight EY440 from Ho Chi Minh to Abu Dhabi on Thursday 7 January 2016 was diverted to Bombay to refuel due to route changes, we apologise for any inconvenience caused and as always, the safety of our guests and crew remains our top priority *GW
Are you kidding me?? The pilot flew in the wrong direction without noticing for more than 2 hours! That’s why we had to get redirected!! That’s a case for media coverage
We sincerely apologise for any inconvenience and disruption caused on your flight, we value all guest feedback and will send your comments over to our team and create a case for you. To do this, we would ask you to kindly send us an inbox message with your booking reference please. *GW
We will decide what to do with this matter on Monday. We are still missing our luggage by the way…
Do you have your PIR reference number, we would like to help with this too. We are sincerely sorry for this experience and hope to hear form you soon *GW
What the hell, man.
Pilot didn´t know he was flying the wrong way. First as EY440 was in malaysian airspace the ATC KUL was contacting the pilot he is in the wrong airspace.
Never ever read something like this, lol.
Sounds like EY now trying to throw the pilot under the bus.
No, not EY, a disgruntled passenger of EY.
I can´t believe what this man complains. I don´t mean this man is lying, no, i mean it seems impossible that the HCM ATC let the pilot flew in the totally wrong direction without contacting him.
It´s very hard to believe that both the HCM ATC and the pilots had absolutely no idea what was going on.
The good news is it seems KUL ATC did. Good job.
Thanks for contacting him.
Etihad’s response was essentially the same as Jeff posted a few days ago. Can’t they find a better explanation to avoid inconvenient questions? But frankly, I thought Etihad agreed with passengers on everything a-priory. Apparently they didn’t.
@Nederland: There are several shortcomings of the theory that MH370 was seen in the Maldives, but fuel exhaustion is not one. See for example the following:
Others here have independently reached the same conclusion.
The strangest thing about this is that apparently all passengers were satisfied by the explanation that the airplane needed refuelling because of a ‘rerouting’, and only one passenger complained about a coffee voucher and delayed luggage.
2 out of 227 passengers travelled using false passports. That’s almost 1% for this flight. I wonder what the worldwide average is? I cannot imagine 1% being typical. Perhaps the 2 carrying the false passports were not just trying to begin a new life in the west.
Could the reboot have been caused by a passenger ordering something and swiping a credit card from the in-seat IFE terminal?
@Brian – http://whotv.com/2014/03/27/credible-lead-mh370-search-shifts-680-miles/ shows many locations of debris close together, spotted by satellites from a variety of countries but never found by ship. I cannot believe this is just garbage and not debris from MH370.
@Jeff – I believe that Victor explained that as the plane gets lighter after burning fuel, eventually the speed will be reduced in order to reduce lift. Increasing altitude would also reduce lift but I understand that an altitude change requires ATC approval so the A/P will not change altitude without pilot input.
Assumptions used in the model that I do not remember being discussed here:
The distance calculated between the satellite and the A/C is shown as a straight line. As the plane got further south of the sub-satellite location, a light beam from the satellite to the A/C would refract when it reaches the earth’s atmosphere, wouldn’t it? I have checked and found that radio waves within a certain frequency range do not refract. Both the uplink and downlink frequencies are within this range If this is not correct, a slight refraction would decrease the ping arc radius.
Similarly, I suspect the influence of gravity on the radio waves is very small and can be disregarded.
All fuel range calculations are based on the reduction of the gross weight by the burning of the fuel. No passengers and/or cargo were unloaded or jettisoned.
Many of us have used various graphs and tables to try to determine the range of MH370. A common speed being assumed is the LRC speed of M= 0.84 for most of the flight. Victor’s August 18, 2015 observations support this speed. But what if the perp’s wanted to go as far away as possible? They might of dropped to the MRC speed that I believe is about M=0.821. Makes sense, doesn’t it? Yes, when you are at the start of the journey. However, what is the furtherest MH370 could have traveled after the FMT? Here is where I made my mistake:
I though that if we picked a point near the 7th Arc that was reached at LRC speed, the distance reached at MRC speed would be a little bit further. But that not correct. Let’s say the plane was powered from 18:22 (GW around 210,000 kg; 35,600 kg fuel) to 00:17, or about 5.92 hrs. If it continued at 471 KTAS (I think that’s M=.821 at FL350) the MRC would be 2790 nm. The LRC for the same endurance at 482 KTAS would be 2853. That discrepancy is that prior to the journey the MRC speed should give a greater endurance than LRC but in the case of MH370, the endurance for all speeds is 5.92 hours. That means if there was enough fuel the longest range would be 3019 NM at its fastest speed of about M=0.89. For it to travel 3019 NM instead of 2854 NM it would have gone further southwest on the ARC. Obviously, this range needs to be reduced for the 3 to 6 minutes it flew with one engine INOP and to account for the left curve it flew after flameout of the second engine.
The burn rate for this scenario is 6013 kg/hr and might be achievable for 9M-MRO either for a pilot initiated step cruise or possibly starting at FL400 at the FMT.
The strangest thing is that media keeps silence.
the plane descended from ~FL350 to ~FL260 prior to FMT according to radar calculations, I don’t understand why would it climb again to FL400…
“I believe that Victor explained that as the plane gets lighter after burning fuel, eventually the speed will be reduced in order to reduce lift.”
What about adjasting (reducing) AoA?
“I suspect the influence of gravity on the radio waves is very small and can be disregarded.”
If I recall correctly red/blue shift has effect of order 0.2 Hz on the BFOs. We discussed this in detail with Yap.
The other interesting aspect, which affects range, as noted by LouVilla, is whether aircon was functional or not. Especially in connection with a sub-optimal flight if only one engine was working.
Quick questions :
FI Page 1 :
“The Captain ordered 49,100 kilograms (kg) of fuel for the flight that gave an endurance of 07 hours and 31 minutes including reserves. The planned flight duration was 05 hours and 34 minutes.”
FI Page 30 :
“Based on MH370 ATC flight plan dated 07 March 2014, the take-off fuel recorded was 49,100 kg.”
Question 1 : How much fuel was in the tanks BEFORE the Cpt ordered 49800 kg of fuel ?
Question 2 : Why is the take-off fuel weight equal to the ordered fuel ? Didn´t MH370 burn fuel on it´s way to Runway 32R ?
Lauren H said: “@Jeff – I believe that Victor explained that as the plane gets lighter after burning fuel, eventually the speed will be reduced in order to reduce lift.”
If the autopilot was set to maintain speed and altitude, as the plane burned fuel and got lighter, the pitch would decrease, thereby decreasing the angle-of-attack (AoA) and maintaining a balance between lift and weight.
@VictorI, Interesting, thanks for clarification.
@jeffwise. I agree that it is arbitrary to define the southern turn as the point exactly when the ghost flight commenced. It is quite obvious that MH370 was in control until final radar contact at 18:22 (as there is evidence for changes in direction, altitude and speed). This has also been suggested early on by Razak.
Notwithstanding this, I am not convinced that failure to find wreckage in the current area implies pilot control until final impact and does preclude the ghost flight scenario in its entirety. The IG paper, for example, entertains a scenario of a missed landing approach and subsequent pilot incapacitation after the southern turn. The early ATSB reports also suggest a great variety of routes, straight but variable from arc to arc (implying further change of direction), which amounts to an impact area far greater than the one being searched. Other scenarios of pilot incapacitation (at whatever stage of the flight, but before final impact) may include: suicide by self-induced hypoxia (perhaps because something went wrong or there was a change of plans; passenger/crew insurrection or a fight/conflict on board, perhaps along with a broken window/fuselage/decompression scenario; a warning shot gone wrong while trespassing foreign airspace, involving a decompression scenario; any other technical issue, to name just a few. Nor do all of these require suicide intention or a completely different route.
At any rate, the ATSB claims that the current search area includes most, but not all of the possible impact points derived from the underlying assumptions. The search may not have been thorough enough or the wreckage sits in a trench or is submerged in silt (or will be found in the next few months). Most drift studies on the flaperon suggest the flight ended north of the current search area, but certainly in the SIO.
It just seems strange that any suicidal person would keep on flying, expecting their death, for hours on end or glide/ditch (just to end up freezing in the ocean) rather than crash, with no known motive, psychological or personal issues. Or would reboot the SDU if their intent was to cover up the deed (and if not, why bother flying that far in the first place). And if political (but not terrorist) motives were at stakes, suicide is just as futile as unheard of.
remote control of an airplane does however shut down alle pilot communications and as i’ve read also turns off the responder.
i don’t see a reason for this option to be ruled out by the period without external communications.
@Victor &@Oleksandr – Yes, I agree that AofA can be reduced to reduce lift but I thought that after a significant weight reduction, either speed or altitude must be adjusted? For example, at a constant Pressure Altitude of 35,000 ft the LRC speed drops from M=0.84 when the GW is 220,000Kg to M=0.796 when the GW is 180,000Kg.
My key point is if you change to a conscious pilot until 00:11, in addition to a curved track, the possibility of a step cruise exists.
@Victor – I trust you have seen the recent TV commercials with Oliver Smoot’s cousin George.
StevanG – Per statements here by experts, I believe that the height reported by radar is not reliable. Also, the last height reported in the FI was 32,800Ft at 1739:59 UTC. If the intent of the FMT was to fly as far away as possible, a Flight Level above FL350 would have been chosen. At 210,000Kg (estimated GW at 18:22), the optimum LRC altitude per the FCOM appears to be around 38,000Ft. Why choose anything lower? All I’m saying is if MH370 went to FL400 after the FMT, the point of crossing the 6th arc would have been further to the southwest. I believe the reported “fuel limits” are for a given altitude and not the absolute maximum range of MH370.
@Lauren H: Constant altitude, constant speed conditions are common for long cruises, for instance, to maintain spacing for transoceanic flights. LRC speed drops as fuel is burned, but a reduction in speed is not a requirement.
I had not seen George Smoot’s commercial, but I did find it online. Thanks for telling me about it.
Decrease of AoA results in decreased drag coefficient, and hence drag. Decrease of airspeed results in decreased drag proportionally to squared airspeed. I guess there is optimum depicted in the table you cite.
With regard to FL400, my opinion is that the aircraft would re-appear on several radars, but it didn’t. Furthermore, recently I lean to think that either dual flameout occurred around 18:20, or some manipulations with power busses followed by shut down of the left engine took place. Inevitably following dive explains dissapearence 18:22, while power interruption explains SDU reboot. In other words disappearence and reboot are results of some event, rather than coincidence.
While it is true that altitude is not measured accurately by radars, one has to differentiate between accuracy and reliability. I was always surprised by willingness and stubbornness of some groups and individuals to discard radar data in favour of assumptions.
“If the intent of the FMT was to fly as far away as possible”
well that’s the point, we don’t know if there was such intent and I would gladly bet there wasn’t
“I was always surprised by willingness and stubbornness of some groups and individuals to discard radar data in favour of assumptions.”
We don’t have radar data. What we have are cell phone pictures of a messed up and amateurish presentation made at the Lido Hotel by people who did not understand what they were presenting in any case.
I think I wrote that, not Lauren.
Accurately speaking you are right: we don’t have radar data, and all the information we have is at the level of rumours. However we have some indications consistent with this:
1. There is a gap near VAMPI. I am not satisfied with terrain-based explanation until I can see details.
2. Disapperence 18:22. I am not satisfied with the explanation that meteorological conditions have contributed to “premature” disapperence.
3. I am not satisfied with the explanation that 4 or 5 radar systems, which were supposed to track MH370 if it was FL350 or above, were not functional at that time.
4. Initial rumor had it that MH370 flew at FL400 shortly after the turn at IGARI. Moreover, I have seen altitude animation from either BBC or CNN. I was unable to find/retrieve it later, but where did they get the data?
5. Low altitude is consistent with Kate Tee’s testimony.
On the other hand we have the assumption of FL350 made by IG and ATSB based on the extrapolation of the last known FL and cruise altitude, mispresented as a fact. Indeed, cruise altitude is consistent with FL350 and vice versa. They refuse to understand that flight was already abnormal after 17:25, and these assumptions may not be applicable.
FL350 was backed by the high average speed. It is true, but simple arithmetic shows that the aircraft could temporarily descent to 7 km without impacting its ability to reach the location 18:22.
Finally FL350 assumption was also supported by fuel consideration. The reality is that fuel is still sufficient to reach the 7th arc if you consider its middle part rather than its two extreme ends (I mean 40S and 17S).
So what do we have? Rumours consistent with several observations against the assumption of a normal flight.
Does anyone have an idea what would happen in TO or TOGA mode with only one (right in this case) engine working?
As aircraft gradually gains speed rudder has to be deflected; but in order to be deflected the feedback system has to recognise change in heading or rotation first, and adjust the rudder accordingly. Will it attempt to return aircraft to the previous heading, or will it just dump rotation (yaw)?
Oleksandr, thanks for the cites. Here’s the response from Martin Dolan I mentioned.
“Bruce Lamon said…
Mr. Dolan, thanks for taking these questions.
Early on the Chinese (March 22) and the French (March 26) published satellite images of floating objects of significant size found in the vicinity of the initial Australian search area. Were any of these objects ever relocated or otherwise analyzed to see if they came from MH370?
JULY 14, 2014 09:14
Martin Dolan, Chief Commissioner (author) said…
No recovered debris has been identified as originating from MH370, and the search effort has not found any debris associated with the satellite images or other visual detections.
JULY 17, 2014 11:27”
Your logic employed to narrow down the possibilities is flawed insofar as you omit one critical possibility: that the BTO is wrong. It never ceases to amaze me that folks are prepared to rule out obvious, highly-plausible possibilities simply on the basis of the BTO. Since when must this doctrine be regarded as infallable? Perhaps we have all simply got (too) used to this being “the only real evidence around”.
@sk999 I have requested from DSTG the results of their BTO and BFO fit/errors from the 10 flights, 60 datapoints sampled. I have received replies, but not tabular data showing how well the models performed against known positions. The “Bayesian Book” provides results only in graphical form, with errors plotted along the range arc as if BTO were always correct. But by their own report, the BTO was not always correct – and most especially for the the accident aircraft, 9M-MR) on the 7th March when it was flying from Beijing to KL.
Has ANYONE seen how well BTO actually performed in the “validation experiments”?
I think some major information is being overlooked, and clear assumptions can be gained from what has been discovered, in addition to what has NOT been discovered.
First and most importantly, lack of debris from a crash into the water. The lack of any meaningful debris, luggage, interior flotsam, cups, bottles, floatation devices, points directly to someone having manually controlled the plane down to the surface for a soft, controlled landing.
Second, was the goal to achieve maximum distance south? or was the likely goal to not be found or located? My suspicion is the latter: not be found, never be located.
If this is the case, and if from #1 we conclude that someone was manually flying the plane down to the surface, then the priority was not to achieve greatest distance south, but rather, to simply not be found. ie, land the plane before it ran out of fuel, when it was just about to exhaust its fuel supply, to minimize a surface slick. Further, cast uncertainty on the auto-pilot theory, by heading east or west, or even reversal of heading and going north for the final 5-10 minutes of flight. If there are steady westerly winds, a 120 or 130 degree turn to the NE, from a due south heading, for the final ten minutes of descent and controlled landing would run against the assumptions being made by crash investigators.
As Alfred Hitchcock stated, when all reasonable possibilities have been eliminated, then only the unreasonable remains.
No debris field, no interior items floating: This means the plane landed intact, and sank in one piece.
This means it was not a ghost flight, and someone trained and skilled controlled the plane down to a soft landing.
If the intent was to make a statement, then a crash into water or crash into a political building on land would have made sense.
If the intent was to make a statement but hide it in a cloud of uncertainty, then fly the plane AS IF IT WERE ON AUTOPILOT, but then at the end of the plane’s journey, regain control and DIVERT the plane from the continued direction of autopilot flight!!!
The person in control might not have been aware of the handshakes, but could have suspected something was being used to track the flight. If the pilot was fully confident in the deception, he might have performed the descent in a steady southern heading, without change. If the pilot wanted to throw a final curve ball, even five minutes flight time on final approach in a due-east or due-west heading would result in 20-30 miles of course deviation from the due-south bearing.
no debris, no crash.
no crash, not a ghost-plane.
not a ghost-plane, had to have been flown.
Flown for publicity? or flown for mystery?
Clearly, two years on, flown for mystery.
mystery is complete as the plane was flown as if it was on auto-pilot, even though the lack of floating debris clearly indicates this was only a ruse.
If flown for mystery, then make unexpected bearing change in the flight’s final fifteen minutes, heading east, west, or even north.
ie, current search area is too narrow, and too far south.
About 1:30am, JAL750 & MH88 may have spoken with Copilot Hamid on VHF emergency radio.
At 1:43am, MH370 may have issued a Mayday SOS broadcast.
At 1:50am, the a/c’s transponders briefly squawked the location “06.9298N, 103.5901E”, i.e. the plane’s last known location after IGARI, immediately prior to the sudden sharp turn to port.
At 1:52am, Copilot Hamid’s cellphone attempted to place a call.
At 2:25am, somebody rebooted the SDU and (?) associated electrical systems on that circuit.
Thus, there is very arguably a lot of evidence for life aboard the plane, for more than one hour after the flight deviation event.
RAT may supply power to critical FC systems & the AP, if so AP may remain active throughout most of any unpowered glide (AP may not function below 200′ unless specific Land2 or Land3 controls are set)