Sigwx

We should wait until an interim report is released. The ministers dumb down information to make it more digestible, the media follow up by clutching at straws and going in the total opposite direction to that indicated by official press releases.

The Satcom components are not housed in the MEC and are powered by the Left Main Bus. Therefore the SATCOM unit and components were likely constantly powered until fuel starvation.

VHF Data works on line of sight at could well have spoken directly to a ground station. The centre or right VHF must be in DATA mode and not VOICE and a station must be in line of sight range while a transmission is made.

JustMaybe

Surely with today's technology it is possible to install gps tracking devices on all aircraft that cannot be switched off by the crew to know where any aircraft is every second of its flight. Is it also not possible to install listening devices in cockpits that will enable all conversations to be recorded external to the aircraft for access by authorities in emergencies?

arlev

You old sage, you - but you're quite right.

And that was my problem - if the crash site is as assumed (or as calculated), then, if the last ping at 8:19 was VHF (and it can't be SATCOM) then the 'line of sight' becomes a problem.

Thanks for your answer.

EsherFlyer

I agree with that based on the ping distances, but it would be highly unlikely to be consistent with the Doppler analysis wouldn't it?

arlev

If the Doppler analysis is accurate, of course. I actually trust the Inmarsat people quite a bit - being British myself .

However, as far as I understand the relevance of the Doppler analysis, it means something like how you interpret the data based on the shift in signal between receipt by MH370 and transmission - that is, the distance travelled which will have had the effect of causing the plane to be at two different distances between the two events.

Therefore, the Doppler effect causes you to postulate the position on the arc.

Two observations:

1. That would still work if the aircraft circled (but I'm not suggesting that it *did* - I am simply pointing out that such an event would throw the crash site further north).

2. Haven't they just disregarded the Doppler work of the initial route by shifting the debris site hundreds of miles from its initial location? That is, assumed a crash site further north based on more fuel burnt rather than on a reassessment of the Doppler maths?

EsherFlyer

I don't quite get this bit. Why do you think SATCOM could not have initiated the partial ping / communication? My reading was along the lines of:

1. SATCOM initiates connection on power up, like a laptop's WiFi connection to an access point;
2. In normal ops on-board avionics kit sends info over that connection, which keeps it alive, but normal kit was turned off so these never happened;
3. Periodically the 'access point' pinged back to say 'You're very quiet today, do you want to stay connected?' and these were detected;
4. At some point something on the plane that hadn't been transmitting decided it would after all. Maybe this was the low-level SATCOM wanting to re-establish a link during power loss / recovery, maybe it was a bit of kit that recovered power or went into special mode that forced transmission.

And just to make sure I'm following again...

Because VHF would only have a range of a few hundred miles at best - typically 200ish for a airplane height?

Quite !

This is perhaps where our understanding of the significance of the Doppler analysis varies. I think the thing they can tell from the measurements is how fast the plane was flying directly away from the satellite. But since it was not on a directly away path this will only have been one part of the velocity vector. I've tried to show what I mean here:



The green and blue diagonals show a distance covered in a fixed period of time, so as they are different lengths they represent different speeds. But if you were travelling along the brown line (which is the direct away from satellite path for the initial positions) and keeping the plane at right angles to you then you would cover the same distance for both the green and blue 'targets', and it is effectively your speed along the brown line that the Doppler effect is measuring (I'm ignoring the fact that the 'direct away from satellite' path sweeps around as time progresses, but I think that is what causes the upwards slope on the Doppler graphs from Inmarsat).

If the target flew in a circle you'd go forwards an backwards on the brown line, and the Doppler measurements wouldn't be linear(ish). So I think they can infer that the plane took a straight-ish course (maybe the earth curvature bit starts to come in as well) south rather than anything northern or circular.

I agree there could have been a circular bit at the end. Can the plane be put in hold over a point where someone wants it to comedown after fuel exhaustion? But why do that unless the 'deepest point in the ocean' theory is valid?

I don't think so. If a lower speed 'diagonal' is now assumed then you end up with something more like the green path than the blue one - which takes you NW-ish. The Doppler measurements gave them the 'along the brown line' bit and that stays the same, so they haven't disregarded or changed it. I think we agreed above that you get a similar effect just by considering the path across the arcs at different speeds. My brain starts to hurt when trying to think about how combining speed across arcs with Doppler shift vectors gives you a better combined result than just one of the methods, but presumably someone's didn't!

As ever, very happy for others to correct / expand on this. I think discussing it helps us all understand what may or may not have happened and to debunk ill-formed theories.

polarbreeze

My money is on the time of crash being 8:19 and the location 34.83°S 102.57°E.
And I don't necessarily buy in to the idea that this was when the fuel ran out. I think it's more likely this location was a deliberate target - the very deepest spot in the ocean.

Why can't it be SATCOM? It's my understanding that it was indeed SATCOM.

And a very small part at that because the plane's track was predominantly tangential to the arcs - so the doppler information is at best a second order effect as far as estimating distance south. It seems that the main value of the doppler was to distinguish between the northern and southern path, and that's how they eliminated the northern path.

EsherFlyer

I think as they go further south it gets more radial. No idea how dominant that makes it in the calcs, but there is a noticeable lift in the frequency offset in the Inmarsat graphs.

cassiewoofer

That's what I thought when I first saw the map of the sea bed! Also when I first heard about the 08:19 half 'ping' it sounded to me like a last desperate attempt to 'shake hands' as water started ingress into the electronics.
This ship towed ping (black box) detector is on it's way out there, shall I say a bit prematurely having not found any debris yet....... let's hope they're going to survey that area first even before they find debris..... even if just to eliminate it.
I'm on a very slow laptop at the moment has that 8000m hole in the seabed got a name?
So could the whole thing be virtually intact at the bottom of 8000m of water....:-O

EsherFlyer

I'm not sure if you're aware of Great Circle Mapper at http://www.gcmap.com. You can use it to play with routes (as seen on a flat projection) if you like to see if your theory fits.

For example to plot a path via IGARI-VAMPI-IGREX to your end point put the following in:

6° 56' 12N 103° 35' 6E-6° 10' 56N 97° 35' 8E-9° 43' 28N 94° 25' 0E-34.83°S 102.57°E

(Based on waypoint info from http://www.fallingrain.com/waypoint)

If you can get info on other known / assumed routes you could overlay those with a comma separator. So for example if you wanted to compare that with a PER-SIN route just add ",SIN-PER". Although I expect you'd have to figure individual co-ords to input .

polarbreeze

Diamantina Deep

arlev

Please excuse the point format:

1. SATCOM does not initiate the ping from 1:11 to 8:11. It only responds to the ping that it receives. I've seen no information about what happens when power is lost or switched off and the power comes back on to indicate - either way - that the plane would initiate a connection via SATCOM. I only understand that as being solely the responsibility of the satellite. Because it is a response, the time from transmission to receipt - including a pre-determined delay response - is the data used to determine the distance of the plane from the satellite. And, we don't know that the 'half ping' was a re-establishing of a link. In short, we don't know what it was - or why.

2. Yes, VHF transmission would be a problem if the 'half ping' was transmitted that way.

3. Vectors. I understand the Doppler calculation (whether rightly or wrongly) to arrive at a distance travelled from receipt of the ping to the retransmission and the, er, 'effect' that that is known to have had on it. The Doppler calculations were added to the data *after* the arcs were determined. They should give some idea of direction being travelled on the arc. That's how I understood it. Therefore, if you move the crash site, you change the angle at which the plane has to fly 'across the arc' - and, therefore, disagree with the Doppler reassessments.

If velocities, it would still work the same. By moving the crash site further north, you cause the angle at which the plane 'crossed' the arc to be less shallow and the vector (as yourself) would have to be different than those calculated and used unless a different velocity for the aircraft was proposed - or, making the airplane speed fit the place where you want the crash.

4. Circling. If the circle is large enough so that one 'lap' is less than one hour, then the time between a one hour interval *could* (by sheer chance) look like it had travelled between two arcs suggesting a diagonal path when, in fact, the difference between the two arcs is somewhere closer to 90 degrees. As I said previously, I don't adhere to it - but I don't discount it.

5. The shifting of the crash site I dealt with above. I guess that it depends on exactly *how* the Doppler calculations interpreting the path of plane at the arc intersections. I must add one bit here, previously posted. The new 'crash/debris' site does have the effect of reducing the distance flown to something that *extremely roughly* I worked out would give a speed somewhere around 500mph.

Now I must go and have dinner...

lewko

How much money are you wagering on the exact location? IMHO, the FDR location and data will probably prove you wrong. It's just too much of a stretch to think an experienced pilot would have executed a number of fuel burning maneuvers to position the plane on a route to a far away deep where it was questionable that the fuel would last long enough to reach it.

I agree the 'partial' ping might be the systems resetting after the fuel ran out or in the final seconds before the plane sank, (not sure if that's SATCOM). That might give the time the flight ended which could narrow the search area.

I wonder if the pilot set a pure directional heading or a waypoint as the last heading he programmed in. This makes a difference in the course because the computer navigation system adjusts for wind drift differently.

iquitos

The fact the have found no debris yet raises the possibility it did not go down anywhere near where they think it it did.

arlev

Or even that it's not on the southern part of the 8:11 arc