Two weeks ago, I wrote a couple of posts about the strange reboot of MH370’s satcom system that occurred shortly after the plane disappeared from primary radar, and asked if anyone could come up with a reasonable explanation. I drew attention in particular to the left AC bus, which the satcom equipment is connected to. This bus can be electrically isolated using controls located in the cockpit, and this appears to be the only way to recycle the satcom without leaving the flight deck. I suggested that there might be some other piece of equipment that the perpetrator wanted to turn off and on again by using the left AC bus, thereby causing the satcom to be recycled as an unintended side effect.
The readers rose to the occasion. Gysbreght pointed out that paragraph 1.11.2 of Factual Information states that “The SSCVR [Solid State Cockpit Voice Recorder] operates any time power is available on the Left AC transfer bus. This bus is not powered from batteries or the Ram Air Turbine (RAT).”
This is an incredibly interesting observation. Reader Oz fleshed out Gysbreght’s insight, writing to me via email:
We could isolate the Left Main AC by selecting the generator control switch to OFF and the Bus Tie switches to OFF; SATCOM is now dead. What else happens……….the Backup generator kicks in automatically to supply the Left Transfer bus. Here’s what’s so spine chilling; if you now simply reach up and select the Backup Generator switch to OFF………..you now lose Left transfer as well. The CVR is gone! I couldn’t believe how easy the CVR was to isolate!
To recap;
Left Gen Control to OFF
Bus Ties to OFF (Isolate)
Left Backup Gen to OFF.
I now firmly believe your mystery reboot was Left AC power being switched back ON……….. after something that had occurred that the perp or perps didn’t want any possible evidence of on the CVR……whatever was being hidden was done by around 1822; AC back to normal.
Gysbreght notes that the Factual Information also identifies the location of the CVR as Electronic Equipment Rack, E7, in the aft cabin above the ceiling, and suggests: “Later [the perp] could have opened Electronic Equipment Rack E7, physically pulled the SSCVR power supply plug from its socket, and then gone back to the MEC to restore power to the Left AC bus.”
Oz has his own theory: “If you are thinking why the hell you would turn Left AC/Left transfer back on? Flight deck temperature control comes from these…”
There’s a precedent for a suicidal airline pilot depowering the black boxes before flying a plane into the ocean: the pilot of Silkair Flight 185 appears to have done just that before pointing the nose down and crashing in December, 1997. It’s easy to imagine Zaharie reading the accident reports and realizing he should also figure out a way to disable the CVR before implementing his suicide plan. When the moment came, near IGARI, one can imagine the veteran 777 pilot suddenly flipping various switches while the baffled newbie, Fariq, looked on.
It’s certainly an intriguing scenario, but it is not without its flaws. As Gysbreght notes, “I would expect the Captain to know that the CVR only retains the last two hours and overwrites older recordings.” So if Zaharie planned to commit suicide by flying the plane for hours into the remotest reaches of the southern ocean, he wouldn’t have needed to turn the CVR off: the portion between 17:07 and 18:25 would have been erased anyway. This is not in insurmountable problem, however. Maybe he orginally intended to crash right away, a la Silkair, but then lost his nerve.
I’m not quite ready to declare, as Gysbreght has, “Case closed,” but I have to admit that the CVR idea is fascinating. Great work, Gysbreght and Oz!
we have two totally different entities(Sebille and Deltares) that have lot of experience with oceanography and their findings match, how Rydberg comes up with something totally different(although using same data as Sebille) is a mystery to me
I guess if a 3rd party found the plane somewhere off NW Australia those ATSB + IG people would claim it’s a conspiracy and someone moved it there from the south.
@StevanG
Hendrik’s model works on an entirely different principle. It does not predict where debris will go. It predicts where the debris came from based on where debris is found and where debris is not found. It really is an entire different (a posteriori analytics), and in my view, much more powerful approach. Of course, in the absence of any debris one can only use a priori models.
I understand that, but I’m quite sure the experts for oceanography would better know which approach is more suitable here, regardless of the power.
@Henrik, Brock, DennisW, StevanG
Henrik’s model “uses 35 simulations along the so-called 7th arc”. Whereas if it is reported correctly in the NYT,
http://www.nytimes.com/interactive/2014/03/17/world/asia/search-for-flight-370.html?_r=0
then Sebille’s model uses starting points all over the IO. Please correct me if my understanding is wrong Henrik but I agree with Brock’s point 2) that the 7th arc starting point for the model confirms the conclusion 🙁
Henrik,
Just imagine a Brownian particle in a box. Where did it come from? Assume you know its actual position and velocity at a given time.
Re accuracy. Let’s start from a simple thing: resolution. In my understanding “adrift.org” model has the resolution of 1 x 1 deg, which roughly corresponds to the cell size of 100 x 100 km at the equator. What happens inside of a cell depends on numerical schemes applied. In your conclusions you specified the area of the debris origin, which is two orders smaller than the representative area of a numerical cell. Not to say that such an accuracy is not possible, but it is very unlikely to achieve.
Re “chaotic behavior”. Any floating ‘particle’ on the surface is subjected to current and wind forcing, which are turbulent in their nature. The longer a time interval is, the higher degree of uncertainty is in the estimation of the end point. How does “adrift.org” model take into account the actual conditions during 2014-2015: currents, tides, surface waves, wind conditions, and respective characteristics of the turbulence? What calibration coefficients are used, and how sensitive are results to those coefficients? They would be different for a small plastic bottle and a flaperon of 2×1 m size. The site says the model is based only on buoys data. I did not find details of the model as the links appear to be broken, but it looks like “adrift.org” feeds some kind of time-averaged current field and into a simple Eulerian tracer model. Is it the case?
But a really curious thing is that if you smooth your curves, the peak probability would be somewhere around 30S.
Henrik,
By a chance I have noticed the following article, which may better explain what I tried to say above:
noc.ac.uk/news/insights-mh370
@AM2
I don’t find the 7th arc to be an unreasonable constraint for starting points. It is all we have really.
@StevanG
Suitability is modified by what you have to work with. If the ATSB comes up with their refined drift model (I suggest you don’t hold your breath), and it resembles Henrik’s approach, I will cry foul. They can’t legitimately use a posteriori methods. But now, as debris data becomes available (and I assume it will accumulate over time), Henrik’s approach is unassailable.
@Henrik
Correction… the general conclusion that people might assume from reading your report, that the ATSB is searching in the correct place. Your specific model is conditional on that section of the 7th arc being the crash site and given one found item so far estimates probability within the arc?
@AM2
I think based on the small amount of data available for Henrik’s model, it would not be reasonable to say that the search is being conducted in the right place. The flaperon could have come from anywhere on the 7th arc. There are no probability nulls. I think if I had reviewed Henrik’s paper before publication, I would have suggested he remove that inference. But hey, it in no way detracts from the quality of the model construct.
@StevanG: it is straightforward: if you are interested in TESTING the ISAT data’s veracity, you REFRAIN from feeding it into your model as an axiomatic truth, and try to develop an UNconditional back-drift distribution, which you then COMPARE to arc7. If they are disparate, then you have (yet more) circumstantial evidence that the ISAT data is bunk.
If, on the other hand, you are already convinced the ISAT data can’t POSSIBLY be wrong, you hard-code its impact point indications into your model as though it were gospel truth, and use the drift model to find the optimal impact point WITHIN the tiny universe to which it has already been restricted by assumption.
Dr. Rydberg has opted for the latter approach.
The extent to which the two methods yield different results is the extent to which the latter method should be ignored – because the difference represents the degree to which our faith in the ISAT data SHOULD be tested; if an UNconditional drift analysis predicts probability densities nowhere NEAR arc7, then we have yet another piece of evidence whose “a priori” incongruity with their data “ISAT believers” must chalk up to bad luck.
@Brock
You have calm down a bit. Just because the IG, ATSB, and various PhD’s made some truly horrible assumptions in conjunction with the ISAT data, does not mean the data is wrong or invalid. The data is what it is.
Benaiahu – barnacle distribution will also depend on it’s buoyancy and posture in the water and amount of irrigation the surfaces got along with the heat of the sun. They punctures are likely spots for attachment. If you open them up they can make useful bait too so I wonder if anyone was tempted? Would be just as likely as someone attacking it?
http://www.watoday.com.au/world/mh370-search-barnacles-on-piece-of-planes-wing-could-help-unravel-the-mystery-20150803-giqrd3.html
VictorI,
That is interesting what Miles has, thanks for posting it. I wish someone would clarify the simulator once and for all.
We’ve so far had:
1. FBI says “no red flags”
2. Zaharie’s brother-in-law states on 4 Corner piece that simulator was broken and doubts he used in this year
3. Landings on SIO “islands” on simulator
4. Now the entire SIO route on simulator
Matty,
Time to call in “Barnacle Bill” I guess!
@DennisW
Just wanted to say that I’m OK with the current search area at the moment as it has been derived from the only data we have but am continuing to sit on the join of several fences (ouch).
@Matty – Perth
Thanks for your comments. It appears that the barnacles are on most of the structure, (all depths) hinting to me that although buoyant enough to float and drift, it did not ride high in the water at all. There also seemed to be sufficient ‘volcanic grit’ on sides and interior to suggest that some period of time was spent ashore (minutes to months for all I know).
Another observation is that there are reddish spots on the top. It appears to me that these were spots where barnacles were anchored but were recently removed, perhaps from rubbing on shore/rocks? If there was a fisherman with a boy, I can imagine time passed away with barnacle removal activities waiting for that fish to bite. The ideas are limitless, but I’m very confident the investigators can solve this. Aeronautical and structural engineers, biologists, and safety incident professions will be key. Hands on inspection is the key, analysis by photo is just too difficult to achieve high levels of confidence.
The photographers that took high resolution shots at the scene prove to be highly valued resources by providing public access to the debris and giving us folks who want to help out a chance to discuss the possibilities. I find that the more I learn from the capable folks helping out on this mystery, the more I’m able to grasp and potentially contribute. There are some really talented folks that post on this site, I’m grateful they share their thoughts publicly.
For me, the high quality photos also provide a timestamp and boundaries to which the official analysis MUST satisfy intelligent factions of the public. And *Barnacle Bill* 🙂
@Cheryl: I know for a fact that Miles is working hard to determine what useful information was on the simulator. He understands how important this information is. On Twitter, he said that he had two sources that confirmed the story.
I see that now just about everybody’s drift analysis takes the wing from the SIO middle-of-nowhere location.
But it didn’t take a model to predict that something that crashed in the SIO would drift westward to Madagascar. Even the simplest drawings show a circular drift westward.
Did we ever figure out how the ATSB came up with Indonesia as the landing spot?
@AM2
I am OK with it (the current search area) as well. Finding the aircraft makes little practical difference if the flaperon is positively identified to belong to MH370. That finding provides a good measure of closure for the NOK, which is the only truly meaningful thing at this point.
The CVR will likely be blank, and the DVR will likely show that the plane was flown to where it was found, ran out of fuel, and not much more.
In fact, if the flaperon is proven to be from MH370 the motivation to continue the search actually diminishes IMO until more debris information becomes available for Henrik’s model or the equivalent.
I have studied the radar data that is available for MH370 and it suggests that after the turn the plane flew at M0.84 and FL340. The plane might have accelerated during the sharp left turn as it descended 1000 ft.
Here is a summary of the work. I’ll work on producing a short write up that will include some other anomalies I found.
https://www.dropbox.com/s/w3uh3mrsw08dlr7/2015-08-03%20Flight%20over%20Malaysia.png?dl=0
Victor,
What are the potential implications of this information? In other words, what is your take on what this data sheds light on specifically?
@JS
The ATSB used a consulting firm, Global Environmental Modeling Systems, to do the work. Their site is linked below.
The site lists recent work, but the list does not include the MH370 debris study, and the list seems to stop about 10 years ago.
http://www.gems-aus.com
@JS
Like several others, you seem to believe that the ATSB forecast that the debris would end up in Indonesia. Actually they said that it would most likely drift west.
“The ATSB continues to receive messages from members of the public who have found material washed up on the Australian coastline and think it may be wreckage or debris from MH370. The ATSB reviews all of this correspondence carefully, but drift modelling undertaken by the Australian Maritime Safety Authority has suggested that if there were any floating debris, it is far more likely to have travelled west, away from the coastline of Australia. It is possible that some materials may have drifted to the coastline of Indonesia, and an alert has been issued in that country, requesting that the authorities be alerted to any possible debris from the aircraft.”
http://www.atsb.gov.au/media/5496949/MH370%20Operational%20Search%20Update%2029%20Oct%202014.pdf
@Dennis: yes, that was the group which is supposed to have authored the study which sent debris to Indonesia.
Here are the conversations I’ve had with official “drifters” (including GEMS) since last December:
https://drive.google.com/file/d/0B-r3yuaF2p72MXhnS0N2ZGY0YzA/view
the problem with the Henrik’s model is not the model and the calculation itself which was professionally done, it’s the assumption that’s contradictory to facts
if the plane really crashed there we would have lot of debris, if there would be lot of debris something would have to be seen by numerous planes and ships that extensively scoured the area only 10-15 days after the impact
VictorI,
Well then Go Miles! I have a lot of respect for him and maybe he can do the necessary gumshoe digging. If the whole SIO route is there what more red flag did they need? And if it is there with the simulator being inactive for about a year, does that mean the route was put on a year or more circa the timing of the Youtube videos? It sounds damning, who the hell would want to go to the SIO even in some simulation? Well let’s see what transpires with this and the flaperon analysis. Why would they say no red flags and then not divulge the biggest red flag of all?
>extensively scoured the area only 10-15 days after the impact
There were lots of gaps in the original search, which concentrated further North from late March. The Northern part of the current search zone was poorly covered.
https://en.wikipedia.org/wiki/Malaysia_Airlines_Flight_370#/media/File:MH370_SIO_search.png
I don’t think the word ‘scoured’ applies.
Yupp I see it now, I wonder why would they leave such a big gap there…
Richard, StevanG
The current sub-sea search zone was given two days attention, 17th-18th March. 1 aircraft on 17th & 3 on 18th allowing an approx total of 16hrs on search task to cover 380,000km^2 with visibility reported to be poor, rain with low cloud. I’d suggest ‘barely glimpsed’ rather than ‘scoured’.
The modelling work published by AMSA’s Drift Working Group contributors describes start points in S20°-S30° latitudes: that was the area of interest at the end of the air phase in Apr 2014.
:Don
Cherly, VictorI, If Miles is correct, then Malaysian investigators have deliberately suppressed a key finding in the case. While I have not been particularly impressed by Malaysia’s handling of the case–indeed it fills me with glee that France has sidestepped the entire ICAO-mandated process by investigating the Reunion debris as a criminal matter–I’m not ready to believe they are that nefarious.
@VictorI:
Interesting contribution as always. I’ve something similar with the same results. I believe the radar altitude data are probably accurate to +/- 1000 – 2000 ft.
Surely a lot of practice is needed for a 20,000 hrs B777 captain before he can undertake a 6 hr flight across an ocean with virtually no pilot activity.
@Jeff, my inner “nefariousness-meter” tells me that there’s nothing I would put past the Malaysian government. I’m more interested in the reason to suppress something like that. Captain Shah being the sole perp is a get-out-of-jail scenario for the government. So, if that simulator information was really suppressed there must be other things which have been suppressed, too – like what Captain Shah’s reason for staging that crazy stunt. Maybe, he did have contact with the authorities during his first leg after the turnaround. If Miles O’Brien is right on this, there has to be more. And it would be indeed a very shocking development. A lot of factions would not look good at all.
I think, the idea that the simulator was broken for over a year isn’t correct. Journalists spoke to the brother in law and he said it was probably inactive for three months.
@Gysbreght, of course Captain Shah doesn’t need practicing to fly into the SIO, but he might’ve tried to figure out how far a certain amount of fue fuel would get him and which area might be suitable if the wreckage shouldn’t be discovered for a long time.
@Jay: Probably the biggest implication is the fuel performance model we have been using is probably okay. Originally, I set out to prove the opposite.
@Gysbreght: Yes, the altitude data seems accurate within several thousand feet and the ground speeds to within 16 kn.
The altitude of FL340 is interesting. If a higher altitude is assumed, then the Mach number gets higher than M0.84 (due to the drop in temperature). At FL340, the pilot is flying at an “even” FL consistent with westerly direction, and the descent allows for a rapid acceleration at the sharp left turn. We don’t know the true altitude, but FL340 does fit nicely.
I think if Miles O’Brien’s sources are really correct on this, then we have to accept that the captain did it. Shah practicing out-of-fuel runs into the SIO would be in the same league as Andreas Lubitz’s googling of how to block access to the cockpit. But many questions would arise in the wake of this information – like, why he did it this way, did he have helpers, did he contact anyone during the early flight. And was this the only thing which got suppressed?
That being said, we have to be careful with this rumor. There’s still the puzzling story about the copilot’s cellphone, which was apparently spread through anonymous US sources through CNN.
We would need to know a lot more details about this in order to assess credibility.
@jeffwise: On the subject of relevant data found on Shah’s simulator, I no longer know what to believe, and that’s why Miles’ work is critical. I hope he has more to say. If the contents of Shah’s hard drive were misrepresented by Malaysia, it would have to be with the tacit approval of the US.
It also is possible that the simulator runs lead to the general area of the current search and so the data is more relevant to the criminal investigation than the search in the SIO. The US gently nudged (pushed?) the search towards the SIO in the early days and has been relatively quiet since then.
I have speculated for months that the US may be tolerating Malaysia misinformation out of fear of undermining and weakening Najib’s government, which is viewed as friendly. Malaysia has important strategic interest to the US to fight against terrorism and to counter expanding Chinese influence in the region.
I know that Miles is working to get more info from his sources. Hopefully we learn more in the coming days. Even a better understanding of the reason for the leak at this time would be helpful.
@littefoot: I had the same thought. If the story of flight paths to ths SIO on the hard drive, which was first floated and then dismissed, is now proven to be true, do we also have to consider other stories that were reported and later dismissed? The cell phone connection at Penang is one. The rapid changes in altitude is another. I tend to dismiss both of these, but who knows?
@dennisw
”The CVR will likely be blank, and the DVR will likely show that the plane was flown to where it was found, ran out of fuel, and not much more.”
False. The data stored on the Flight data recorder will show altitude,speeding,temperature etc. Will for sure indicate if flight path was deliberate or not.
@GuardedDon
interesting, (admittedly with hindsight) they could do more visual search before starting a lot more expensive underwater one, they probably believed the plane is really sunk intact somewhere
A pilot who wants to figure out how far a certain amount of fuel would get him doesn’t sit motionless behind a computer screen showing an image of a featureless ocean during six hours. He picks up the FCOM or QRH, a pencil and a piece of paper. His home-built simulator wouldn’t tell him which area might be suitable if the wreckage shouldn’t be discovered for a long time. The whole idea of finding a trace of such a boring flight on his HD is just plain silly.
@IR1907
Like I said, the FDR will show that the plane flew to where it was found. Of course the flight path was deliberate. Planes don’t fly themselves.
The FDR records only about 2000 parameters, including but not limited to the positions of pilot flight control inputs, positions of elevator, stabilizer, ailerons, rudder, autopilot modes and selections, electrical systems configurations, pitch and roll attitudes, airconditioning and pressurization, selected thrust settings and engine parameters, etc, etc.
@Gysbreght
Sure. Pretty much common knowledge. So what?
@Gysbreght, the matter is fairly simple. Either there were practice runs into the SIO or there weren’t. If they were found on Shah’s simulator he obviously must’ve had a reason for creating those simulations. So, let’s wait and see if something more transpires. Since Miles O’Brien has confirmed personally via twitter that he has more than one credible source I’m not discounting that possibility lightly.
Well, if a certain Miles O’Brien confirms it personally on twitter it must be true. Silly me.
@littlefoot
I find it hard to believe, what info could he obtain by flying to SIO on the home simulator?
Unless he practiced landings on Cocos or Christmas island, as Telegraph has published.
http://www.telegraph.co.uk/news/worldnews/asia/malaysia/10917868/MH370-captain-plotted-route-to-southern-Indian-Ocean-on-home-simulator.html
“The pilot of the missing Malaysia Airlines plane which disappeared in March with 239 people on board had plotted a flight path to a remote island far into the southern Indian Ocean where the search is now focused, investigators have discovered.”
so is Telegraph purportedly lying or he really practiced those? Because if he really did then how could anything except Christmas Island be perceived as his final goal I just can’t fathom
@dennisw
I meant change of altitude,speed after the FMT.
@StevanG,
So, CI is really “far into the SIO” … Really? I must have another look at my atlas.
@DennisW,
Planes fly themselves pretty well if left their own devices, until the fuel runs out. A course must be set, and I agree that this requires manual input. But it is entirely possible that a course was set inadvertently, rather than by deliberately entering waypoints into the FMC. Consider the possibility that one or more orbits were carried out somewhere near MEKAR using the heading bug as the crew struggled to get things under control. They expired, and the plane rolled out of the turn on the last heading and continued on its way into the SIO.
@IR1907
Yes. I assumed that. My point is that the FDR is very likely to show a perfectly functioning aircraft being flown until fuel deletion, and falling into the ocean. It is very unlikely to tell us anything particularly useful. Does anyone care about what altitude or speed or any combination of control service orientations that were used?
Of course, people who believe in any number of bizarre failures such as depressurization, fires, etc will at least be quieted. Without the FDR we would be hearing about these theories for the next few decades. In that sense it would be worth finding it, and then all we would have to do is convince the tin hats that the FDR data was not somehow faked.
As far as the flight simulator, I share Gysbreght’s view. I think it is a complete fabrication. My only surprise is how long it took to creep back into the public consciousness.