Flight MH370

[Chimpy]

The scariest thing about this thread is that Ghambit is pretty much the voice of reason. 

[Lantyssa]

Heh.  That's so true.

While a hijacking might give some 'meaning' to this tragedy, a fire still makes the most sense.

[MahrinSkel]

Since "serious" news outlets are considering theories up to and including the literal hand of God, I think we're all comparatively reasonable here.

--Dave

[Khaldun]

You know, it's pretty clear that even the relatively cautious mainstream media thinks it wasn't an accident, though the plane may well have met with an accident later.

[Goreschach]

No, the mainstream media doesn't want it to be an accident.

[Merusk]

More accurately, they don't want an answer because it lets them speculate, fuel the usual bullshit with talking heads, including anti-muslim nonsense and raise ratings.   The latest one I saw was  how the oldest pilot was a political exrtremist because they found a picture in which he was wearing a "Democracy is Dead" t-shirt.

[01101010]

If this was hijacked and taken to completion, then what happened to all those passengers? I'd think by now there would have been something. Just armchair throwing it out there.

[Ghambit]

Thought this was an interesting point of view: https://plus.google.com/106271056358366282907/posts/GoeVjHJaGBz

MH370  A different point of view. Pulau Langkawi 13,000 runway.

A lot of speculation about MH370. Terrorism, hijack, meteors. I cannot believe the analysis on CNN - almost disturbing. I tend to look for a more simple explanation of this event.
 
Loaded 777 departs midnight from Kuala to Beijing. Hot night. Heavy aircraft.  About an hour out across the gulf towards Vietnam the plane goes dark meaning the transponder goes off and secondary radar tracking goes off.
 
Two days later we hear of reports that Malaysian military radar (which is a primary radar meaning the plane is being tracked by reflection rather than by transponder interrogation response) has tracked the plane on a southwesterly course back across the Malay Peninsula into the straits of Malacca.
 
When I heard this I immediately brought up Google Earth and I searched for airports in proximity to the track towards southwest.
 
The left turn is the key here. This was a very experienced senior Captain with 18,000 hours. Maybe some of the younger pilots interviewed on CNN didn't pick up on this left turn. We old pilots were always drilled to always know the closest airport of safe harbor while in cruise. Airports behind us, airports abeam us and airports ahead of us. Always in our head. Always. Because if something happens you don't want to be thinking what are you going to do - you already know what you are going to do. Instinctively when I saw that left turn with a direct heading I knew he was heading for an airport. Actually he was taking a direct route to Palau Langkawi a 13,000 foot strip with an approach over water at night with no obstacles. He did not turn back to Kuala  Lampur because he knew he had 8,000 foot ridges to cross. He knew the terrain was friendlier towards Langkawi and also a shorter distance.
 
Take a look on Google Earth at this airport. This pilot did all the right things. He was confronted by some major event onboard that made him make that immediate turn back to the closest safe airport.
For me the loss of transponders and communications makes perfect sense if a fire. There was most likely a fire or electrical fire. In the case of fire the first response if to pull all the main busses and restore circuits one by one until you have isolated the bad one.

If they pulled the busses the plane indeed would go silent. It was probably a serious event and they simply were occupied with controlling the plane and trying to fight the fire. Aviate, Navigate and lastly communicate. There are two types of fires. Electrical might not be as fast and furious and there might or might not be incapacitating smoke. However there is the possibility given the timeline that perhaps there was an overheat on one of the front landing gear tires and it blew on takeoff and started slowly burning. Yes this happens with underinflated tires. Remember heavy plane, hot night, sea level, long run takeoff. There was a well known accident in Nigeria of a DC8 that had a landing gear fire on takeoff. A tire fire once going would produce horrific incapacitating smoke. Yes, pilots have access to oxygen masks but this is a no no with fire. Most have access to a smoke hood with a filter but this will only last for a few minutes depending on the smoke level. (I used to carry one of my own in a flight bag and I still carry one in my briefcase today when I fly).
 
What I think happened is that they were overcome by smoke and the plane just continued on  the heading probably on George (autopilot) until either fuel exhaustion or fire destroyed the control surfaces and it crashed. I said four days ago you will find it along that route - looking elsewhere was pointless. 
 
This pilot, as I say, was a hero struggling with an impossible situation trying to get that plane to Langkawi. No doubt in my mind. That's the reason for the turn and direct route. A hijack would not have made that deliberate left turn with a direct heading for Langkawi. It would probably have weaved around a bit until the hijackers decided on where they were taking it.
 
Surprisingly none of the reporters , officials, other pilots interviewed have looked at this from the pilot's viewpoint. If something went wrong where would he go? Thanks to Google earth I spotted Langkawi in about 30 seconds, zoomed in and saw how long the runway was and I just instinctively knew this pilot knew this airport. He had probably flown there many times. I guess we will eventually find out when you help me spread this theory on the net and some reporters finally take a look on Google earth and put 2 and 2 together. Also a look at the age and number of cycles on those nose tires might give us a good clue too.   
 
Fire in an aircraft demands one thing - you get the machine on the ground as soon as possible. There are two well remembered experiences in my memory. The AirCanada DC9 which landed I believe in Columbus Ohio in the eighties. That pilot delayed descent and bypassed several airports.   He didn't instinctively know the closest airports. He got it on the ground eventually but lost 30 odd souls.   In the 1998 crash of Swissair DC-10 off Nova Scotia was another example of heroic pilots. They were 15 minutes out of Halifax but the fire simply  overcame them and they had to ditch in the ocean. Just ran out of time. That fire incidentally started when the aircraft was about an hour out of Kennedy. Guess what the transponders and communications were shut off as they pulled the busses.
 
Get on Google Earth and type in Pulau Langkawi  and then look at it in relation to the radar track heading. 2+2=4  That for me is the  simple explanation why it turned and headed  in that direction.

Smart pilot. Just didn't have the time.

That's a great summation (more eloquent then I could provide); thanks.  I'd only add that the author is assuming they had instruments (or even controls) to run the approach to Lankgawi.  At night, near impossible to fly that thing VFR (visual) over a dark countryside and ocean.  They had good weather though, so maybe he could pull it off... maybe not.  As a pilot, he may have weighed the risks and decided to kill time until daylight.

Recall, the last military radar contact was right near Lankgawi airport also.  Then the track goes haywire.

Also, logic dictates the pilot would've preferred a lower cruising altitude assuming he was instrumentless.  The airways are busy and midairs are always a risk at normal cruising altitudes.  Inflight TCAS (traffic awareness) systems typically only work off of transponder; so if that's off, nearby traffic wont be able to avoid and vice versa.  Some models have magnetic disturbance modes, but that's a crapshoot and short range.  The smarter move would be to stick below 5k feet for safety reasons, and away from approach/departure vectors.  The caveat being radar would have trouble tracking.

Using that logic, it's not hard to see how MH370 might have gotten mixed up in another aircraft's flight plan even though they never flew it.   The ACARS ping data doesnt concur though, assuming it's valid.  Youd have to assume he crashed the plane in the sea somewhere between west and east Malaysia.  By 8am the data stopped.

[MahrinSkel]

That scenario has it crashing into the ocean well after dawn and leaving *nothing*.  No survivors, no debris, no witnesses, in some of the most active seas in the world.  Not to mention being inconsistent with the satellite data, which is the one piece of the whole thing I actually trust.  You are not going to dump a plane in the South China Sea or just off northwest Australia and not find anything.

An accident scenario only works on the presumption that the initial problems took out all communications, that flight control was retained for a while and then was also lost before anyone was able to use a cell phone to communicate, and the plane went straight and level heading roughly southwest for a couple of thousand miles with no controls, then *somehow* enough control was restored for a landing soft enough that the plane did not break up, in a manner that few or none made it off the plane into the water with a flotation device.  It's plausible that there is debris in the southern Indian Ocean that hasn't been spotted yet.

It's even possible that there was a perfect storm of coincidence such that nothing will ever be found, like could have happened with AF447 if they had somehow flown a thousand miles south before going in the drink (there was almost no debris, they only found the crash site because they were in communication almost to the end).

--Dave

[lamaros]

An improbable chain of events stemming from an accident is still far more likely than an improbably chain of events stemming from some expertly conducted human actions than have no discernible value to anyone as yet.

[MahrinSkel]

The accident scenario has the advantage that we don't have to try and figure out the who and why, which takes you off into tinfoil territory if you lack the resources of the CIA.  Ignoring that, you can also describe a plausible scenario where the plane was deliberately diverted to Kyrgyzstan, Tajikistan, or Turkmenistan for reasons unknown by parties unknown.  Either scenario can only be definitively dismissed by proving the other, since both are improbable and involve unknown antecedents, only in the aftermath can we find any answers.

Playing Sherlock Holmes is fun, either way, but only because nothing we spitball here is going to have any actual effect on either the progression of the search or us personally.

--Dave

EDIT: This is interesting.  At the same time they are backing off of reports that the ACARS was disabled before the last radio communication (which I ignored because I couldn't find the primary source or an exact timestamp, but many took as 'confirmation' of foul play), Malaysia authorities are now saying that the last ACARS data received shows the Flight Management System (a more sophisticated verson of an autopilot) being given instructions to make the turn over Malaysia.  The report says that this is being seen as another indication of foul play, but I am not so sure.  If the FMS uses different control runs than the flight controls, you can easily imagine that the pilots, dealing with a fire or other damage, would attempt to use it as a workaround.

[satael]

An improbable chain of events stemming from an accident is still far more likely than an improbably chain of events stemming from some expertly conducted human actions than have no discernible value to anyone as yet.

I think that it's an accident with some one in a million chance thing (that will get changed/fixed once they figure out what it is) mucking up the picture. Making a plane deliberately disappear like this would take a mastermind and planning on level seen only in movies and still have a really high chance of failing.

[Ingmar]

Oceans are big. Really big. It doesn't really stretch credulity at all for the plane to go down without a trace, I don't think.

[KallDrexx]

Thai radar shows plane turn-around.

However, the more important part in my eyes:

"The unknown aircraft's signal was sending out intermittently, on and off, and on and off," the spokesman said. The Thai military lost the unknown aircraft's signal because of the limits of its military radar, he said.

If true it definitely adds another data point to equipment failure

[Jeff Kelly]

I'm not clear what they mean by 'unknown aircraft's signal was sending out intermittently'.

They can't mean the transponder signal because then it would not be an 'unknown aircraft'. The transponder - even if it was only intermittently active - would have shown that it was MH 370 as well as direction height and airspeed. So they must mean radar contact.

[IainC]

I'm not clear what they mean by 'unknown aircraft's signal was sending out intermittently'.

They can't mean the transponder signal because then it would not be an 'unknown aircraft'. The transponder - even if it was only intermittently active - would have shown that it was MH 370 as well as direction height and airspeed. So they must mean radar contact.

But military radar works on active pings rather than passively receiving. I suspect that it's simply a case of hedging the language in this case. The military radar was showing a blip in the same space as the intermittent transponder signal. It's reasonable to assume that the blip was MH370 but not a 100% certainty without other corroborating data - especially in such crowded airspace.

[Jeff Kelly]

"The unknown aircraft's signal was sending out intermittently, on and off, and on and off," the spokesman said. The Thai military lost the unknown aircraft's signal because of the limits of its military radar, he said.

This doesn't sound at all like they had a transponder signal, it's more like a 'normal' radar contact they lost because the aircraft got out of range

[Ghambit]

redux:
ACARS only 'pings' like once every 30mins and it's only a few bits for a handshake.  Not a lot of power needed.  On the contrary, I'd assume the sat-comm would ping more often.  It's still confusing which piece of equipment they were using to triangulate.

I'll go on the flipside here and surmise the only possible 'hijacker' scenario I might give credence to.  I've learned that apparently the pilot can de-pressurize the plane as well as control O2 flow from the oxy generator.  Note, the pilot's airsupply is separate typically (I believe it's on a bottle) and lasts much longer, whilst the cabin O2 is run via oxy generator which doesnt last very long - 15 mins or so? 

So essentially, the pilot (or whoever took over the cockpit) asphyxiated the passengers purposely at altitude (they're dead, so they cant use their phones, storm the cockpit, etc.)  After the cabin was asleep permanently, he rushed to a lower altitude before his independent air supply runs out (O2 bottle flowrates at altitudes above 15k are super high) and/or possibly to evade radar.  Since ACARS was off at this point, the ground stations wouldn't know what was happening.  Pilot diverts to Lankgawi to blend into commercial traffic on his northbound trek, which takes himself over very sparsely populated countries with little to no radar coverage.    He eventually lands the plane for whatever use (likely for sale), and since the passengers are already dead they're no trouble to deal with.   I'd put this scenario on the younger pilot (the 29 yr. old copilot) only because he would likely be over-stressed with probably a <$30k salary and student/flight-loans in excess of $100k.   

As for the transponder.  They routinely glitch (happened a few times to me).  The flow says to cycle them on/off, and usually that fixes it  -  this is very common dialogue between ATC and pilot.  If the xpndr isnt fixed, sometimes the controller may divert the plane depending on the traffic.  Slick controllers won't stress the pilot, so sometimes they say nothing until you're on the ground and they're all like "hai, just FYI your transponder wasnt working on approach"     Controlled airspaces require a transponder except in odd cases.  The most common glitch is altitude, because the transponder is usually reading altitude from a transmitting altimeter inside the plane - if both aren't calibrated perfectly you'll get a false reading but usually not more then say 500ft.  If you've got clogged pitot-statics like in AirFrance, the xpndr will transmit improperly as well.

[Jeff Kelly]

OK now it's getting even more ridiculous:

- Several outlets are reporting that allegedly several inhabitants of a small island in the Maledives saw a 'large commercial jet flying low over the island' on saturday march 8th. Allegedly they reported the incident to local authorities on the same day.

- taliban officially denied involvement in the disapperance of MH 370, a 'commander' is quoted as saying that 'we wish we had the opportunity to hijack such a plane'

[Jeff Kelly]

redux:
ACARS only 'pings' like once every 30mins and it's only a few bits for a handshake.  Not a lot of power needed.  On the contrary, I'd assume the sat-comm would ping more often.  It's still confusing which piece of equipment they were using to triangulate.

The 'all-knowing landfill' (wikipedia) states that ACARS routes messages depending on connectivity over either VHF or Sat-Com. As to the triangulation, all sources I've read talk about satellite communication used to ping so it's either a dedicated satellite-trasnceiver for a dedicated data reporting system or the Sat-Com link of ACARS.

It also seems that they've received the signall at two receivers. If they had only received it on one receiver they wouldn't have been able to triangulate anything, with three it wouldn't be a geodesic line.

re. power: It's not the amount of power needed to send/receive a ping. That would probably negligable in flight when the generator is on. It's the fact that you couldn't power such transmitters from a small compact internal energy cell. We were talking about whether or not the pings could have still been received even after the plane had crashed/sunk. I deem that to be unlikely. A dedicated locator probably could have, the power requirements of a standard VHF or Sat-Com transmitter are too high for that though. Such a device is connected to mains or a dedicated backup power source that is able to supply the amperage needed to power the transmitter. It's unlikely that it would still function after a crash or major defect. There are also other emeregency systems (portable radios, locator beacons, flight data recorder etc.) on board for such an event so it very likely wasn't a design requirement either..

If the ping was answered by ACARS or another standard reporting system then the plane either landed mostly intact or was still in the air. Which wouldn't exclude the theory that a defect or major incident caused the erratic behaviour.

[IainC]

The Thais have cast some doubt on the 'climb to 45k then dive to 23k' manoeuvre. They suggested that those readings could be data artifacts and not necessarily valid readings. Partly this is because such artifacts are fairly common but also because there's some doubt about being able to lose that much altitude that fast and still have wings on the airliner afterwards.

[Jeff Kelly]

Ok so let's suppose for a moment that what those islanders on the Maledives saw was true and it was really a very low flying MH 370.

How's that for a theory:

At the point where air traffic control lost contact the plane ran into some kind of trouble, a major defect perhaps. The pilot starts a descend and wants to turn back to land. The defects are so severe though that the instrumentation is affected. Only option is VFR. Even worse some of the instruments still work but now the crew has to deal with conflicting and possibly incorrect info.

After the descend and with no way to contact air traffic control the crew gets lost and flies west instead of east. Maybe by the time this happens the crew is no longer conscious/alive and the plane flies on due west over the indian ocean until it is seen last by the islanders.

The only mystery being how radar installations and air surveillance units from several different countries could miss a Boeing 777 flying straight on.

Oh and of course what the hell the Malaysian authorities have been smoking to lead us onto a wild goose chase over half the earth

[Jeff Kelly]

and yes I only just realized, that this is exactly what the wired article was saying, even though I've read it.

[edit] It would fit quite well. The Dhaalu atoll where the islanders allegedly saw the low flying plane could be reached by flying due west in a straight line from Langkawi on Pulau, where the writer of the wired article said the pilot crew would be heading.

They would have passed south of Sri Lanka though so the air traffic control of Colombo should have noticed it. They deny any contact with MH370 though

[Ghambit]

The Maldives angle doesn't coincide with the last known (810am) satellite ping data though.  That's what's got this entire investigation bottlenecked.  Ironically, the satellite is in geosync quite near the Maldives; and the siting there would quite literally be the closest piece of land possible to that satellite.

[Jeff Kelly]

There are several ways you can triangulate a position via a radio signal, all of them are very susceptible to interference and/or resolution of time measured. Especially when the distances get large. That's because radio signals travel at speeds that are near c.

- time of flight: measurement of the time it took a signal to reach different receivers.
- angle: measure the angle at which different receivers received a signal
- received signal strength; if you know the transmit power of the transmitter and the received signal strength at the receiver you can estimate the distance
- signal phase: measure the phase difference of the signal you sent out and the reply
- round trip time of flight

All of those triangulation methods measure values that are susceptible to interference or measurement errors and can give you results that can be off by miles if the data is distorted.

If your signal travels at c you'll be off by 1 meter if your time measurement is only off by 3 nanoseconds.

[MahrinSkel]

There are several ways you can triangulate a position via a radio signal, all of them are very susceptible to interference and/or resolution of time measured. Especially when the distances get large. That's because radio signals travel at speeds that are near c.

- time of flight: measurement of the time it took a signal to reach different receivers.
- angle: measure the angle at which different receivers received a signal
- received signal strength; if you know the transmit power of the transmitter and the received signal strength at the receiver you can estimate the distance
- signal phase: measure the phase difference of the signal you sent out and the reply
- round trip time of flight

All of those triangulation methods measure values that are susceptible to interference or measurement errors and can give you results that can be off by miles if the data is distorted.

If your signal travels at c you'll be off by 1 meter if your time measurement is only off by 3 nanoseconds.

Time of flight is what is used by radar, and is limited in precision to a scale determined mostly by pulse width.  Angle depends on the design of the receiver, and can be very precise in some cases (this is what was detected in the satellite pings, accuracy appears to have been fairly rough with an error in the neighborhood of +/-100nm around the lines in the maps I linked earlier).  Received signal strength is nearly useless except as a *very* rough indicator of distance when transmitter strength is known and interference isn't a significant factor.  Signal phase is only useful in narrow circumstances when you are trying for sub-micrometer refinement of an otherwise known distance (in radar signal phase is used to distinguish moving targets, but that's a whole different thing).  GPS uses extremely precise clocks to measure tiny differences in time of flight of signals that are changing in pre-determined ways.

--Dave

[Jeff Kelly]

Yeah they even have to compensate for time dilation effects due to gravity, otherwise clocks would shift so much that you'd be off by continents.

I seriously wonder how they triangulated the path and how reliable the info really is.

[Jeff Kelly]

I've looked around but I couldn't find any detailed info on how they calculated the positions.

What I could find out is that they probably used multilateration to get to the line they've released. This means that they'd need to calculate the time difference of arrival fpr three receivers or antennas/antenna segments.

You basically measure the time difference between the time the signal hits the first antenna/receiver and the time it hits the second antenna/receiver. If you know the exact location of both antennas/receivers you can estimate the location of the sender.

One TDoA measurement (two receivers) will only allow you to place the transmitter on a hyperboloid though, you'd need a second TDoA measurement (a third receiver) to get to a geodesic (the intersection of those two hyperboloids) and three TDoA measurements (four receivers) to pinpoint an exact location in 3D space.

As it seems they only received the ping on one geostationary satellite though (marked on the map, right over the maledives) which means that the ping would have needed to hit three different antenna segments or spot beams on the same satellite, they'd need to know the exact distance of those antenna segments from each other and they'd need to be able to measure the time difference of arrival for each of those segments with a certain precision. It's not that hard because you can do a wave count and measure phase shift between those signals, which is easier than if you needed to precisely measure times with nanosecond resolution but still.

Another problem I have with the picture is that it doesn't show a geodesic line. If they had two distinct TDoA measurements the result would be basically a straight line that connects the two points where both hyperboloids intersect. On a spheroid a straight line is always a geodesic line (shortest line following the curvature). The image looks as if it is the line that is created when one hyperboloid intersects with the surface, though. Which implies only a single TDoA measurement. That would mean that, at least theoretically, the plane could be anywhere on the 35° circle the red line is on. They've probably only marked the area the plane could realistically have travelled given its remaining fuel supply.

I still would have liked for them to disclose how they came up with that line. I hope that at least the nations that take part in the SAR operation have the data and can verify the accuracy. I also wonder why they didn't calculate similar intersections for the other data pings they must have received. Maybe they could have shown the direction the plane was travelling.

[Korachia]

Here is an elegant yet simple theory of this whole debacle: http://www.wired.com/autopia/2014/03/mh370-electrical-fire/

An electric fire, and a pilot's attempt to handle it by getting to the nearest airport might be the explanation.

(Edit: oh it's already been posted - sorry for that. Carry on)

[Ghambit]

I've looked around but I couldn't find any detailed info on how they calculated the positions.

What I could find out is that they probably used multilateration to get to the line they've released. This means that they'd need to calculate the time difference of arrival fpr three receivers or antennas/antenna segments.

You basically measure the time difference between the time the signal hits the first antenna/receiver and the time it hits the second antenna/receiver. If you know the exact location of both antennas/receivers you can estimate the location of the sender.

One TDoA measurement (two receivers) will only allow you to place the transmitter on a hyperboloid though, you'd need a second TDoA measurement (a third receiver) to get to a geodesic (the intersection of those two hyperboloids) and three TDoA measurements (four receivers) to pinpoint an exact location in 3D space.

As it seems they only received the ping on one geostationary satellite though (marked on the map, right over the maledives) which means that the ping would have needed to hit three different antenna segments or spot beams on the same satellite, they'd need to know the exact distance of those antenna segments from each other and they'd need to be able to measure the time difference of arrival for each of those segments with a certain precision. It's not that hard because you can do a wave count and measure phase shift between those signals, which is easier than if you needed to precisely measure times with nanosecond resolution but still.

Another problem I have with the picture is that it doesn't show a geodesic line. If they had two distinct TDoA measurements the result would be basically a straight line that connects the two points where both hyperboloids intersect. On a spheroid a straight line is always a geodesic line (shortest line following the curvature). The image looks as if it is the line that is created when one hyperboloid intersects with the surface, though. Which implies only a single TDoA measurement. That would mean that, at least theoretically, the plane could be anywhere on the 35° circle the red line is on. They've probably only marked the area the plane could realistically have travelled given its remaining fuel supply.

I still would have liked for them to disclose how they came up with that line. I hope that at least the nations that take part in the SAR operation have the data and can verify the accuracy. I also wonder why they didn't calculate similar intersections for the other data pings they must have received. Maybe they could have shown the direction the plane was travelling.

I had similar thoughts, but since the satellite is right over the maldives the geodesy is near symmetrical (more like an ellipse really) from there.  e.g. the spotbeam is nearly the same distance all the way around.  We are assuming the beam geometry is exactly the way the media presents it though; which is unlikely.  For all we know that beam may actually be obliquely pointed towards the east.

Whole thing seems fishy to me because there's so much missing data.  Like you said, there should be other data pings (every 30 mins minimally).  The fact that there isn't tells me that Inmarsat might be looking to backpeddle from their calculations... or arent that confident in them.

edit:  I'd been following a tad on reddit and I dont believe even they are attempting to parse the possible data.  It'd be an interesting project; if I werent so hammered with school and applications I'd fool with it for fun.  But, I'm sick like that.

[IainC]

It was explained previously in an article about the ACARS thing and satellite keep-alives, that previous pings aren't stored, they are overwritten with the latest data. Actual data is presumably stored wherever the satellite forwards it to but the handshake and keep-alive stuff is rewritten every time a new ping is received..

[HaemishM]

Since "serious" news outlets are considering theories up to and including the literal hand of God, I think we're all comparatively reasonable here.

--Dave

 

[01101010]

Since "serious" news outlets are considering theories up to and including the literal hand of God, I think we're all comparatively reasonable here.

--Dave

 

There's..........something onthewing!   

[Ghambit]

It was explained previously in an article about the ACARS thing and satellite keep-alives, that previous pings aren't stored, they are overwritten with the latest data. Actual data is presumably stored wherever the satellite forwards it to but the handshake and keep-alive stuff is rewritten every time a new ping is received..

Makes sense.  Now to be 'splaining that to the media who keeps harping on the fact there isn't more ping data.

Latest theories has our military and Australia's leaning towards the southern arc now (that's where we sent the sub-hunter).  The search box is much smaller though it's all the way to the southern Indian Ocean.
They derived it simply by using point of last contact, available fuel/range, and point of last sat-ping.  If you overlay all that you come up with the new search box.



Note: this route comes from the NTSB:


I took it a step further and dead-reckoned a possible scenario if they were diverting to Lankgawi and failed on Autopilot.  Note, the above box is 11 days old so it's been shifting steadily Eastward due to currents/wind.  Actual splashdown occured more to the WNW, which supports the claim.  If you line it all up, a failed approach to Lankgawi corresponds exactly to a splashdown in that position.  Similar could be said of Andaman airport, but it's a tad far imo and the approach angle doesnt line up as well:

http://tinyurl.com/nn9lcx4

Note, currents arent terribly swift in that spot.  Matter of fact, just to the north is a doldrum.  It wont have moved far to the east w/o the wind.  Maybe I'll plot more later.

[IainC]

Here's the article about the satellite ping data. Source

I note many posts referring to earlier pings and asking the question why no information has been released concerning anything other than the “last ping.” Maybe I can shed some light on this.

Summary
1) It is most unlikely that data relating to anything other than the most recent ping would be retained onboard the satellite, and
2) Each ping must necessarily contain an unique identifier.

Communication satellites generally use time division multiple access (TDMA) to enable them to service many “clients” simultaneously. This works as follows: the satellite allocates time slots to each client for them to uplink their data, so that in normal operation client A transmits a brief burst of data then Client B, client C and so on. The whole sequence repeats many times per second in so-called frames. For this to work, the sat’ needs to keep a record of all its potential clients and the time slot(s) allocated to each one. This is the TDMA scheduling table and it is maintained by the satellite’s on-board computer.

Periodically, the satellite broadcasts a channel access frame, inviting any new client wanting to use the satellite’s services to identify itself and the services it wants to access. The timing of the replies received during the channel access frame enables the satellite to work out the most suitable time slot(s) to allocate to that client. Note that in doing so, the satellite has implicitly calculated the round-trip time for radio waves to that client.

Once a client (aircraft, sat-phone, etc.) has established channel access it periodically pings the satellite to check its time slot allocations. (Hello this is client XYZ. My time slot is 36ms after the frame. Is that still OK?) If the client has moved significantly, the satellite may notice the timings have changed and update its TDMA tables with a new time slot for that client. Note that all of this TDMA stuff is merely to maintain access to the satellite if needed and has nothing to do with actual data transfer, which is negotiated separately as and when needed.

The operative word is update in that paragraph. When the tables are updated, previous values are overwritten and lost. There would be no reason to log or down-link all the technical details of the TDMA protocol as that information has no commercial value to the satellite operator.

When a client falls silent, its entry will remain in the TDMA tables until expunged to make room for a new client. It would appear that Inmarsat were able to interrogate their bird before this happened and down-link the data for MH370. Unless the satellite was logging useless, out-of-date information, this record would contain only the most recent ping data.