Let me add my other concern about the INMARSAT analysis. Based on the very cursory explanation given (and echoed by the media) they used doppler analysis to determine the frequency shift of the “ping” signal and then determined that the southern route was the best fit model for the data observed. What I’d like to understand is how exactly INMARSAT measured the frequency of the ping with such precision?
Speed of airplane = 600 miles per hour
Speed of light = 670,000,000 miles per hour
SA/SL = 9e-7 * 1,600,000,000 HZ = 1400Hz
This requires not just capturing the data (digital message from Flt. 370) but the exact frequency of that signal at the satellite’s uplink receiver. Was this capability onboard the satellite, and why would it be, since it’s not needed for normal communications purposes and satellites generally don’t carry any extra mass. Also, given that the satellite drifts in it’s geo-sync orbit in a figure-8 pattern, it was moving toward and away from Flt. 370 at the time also.
I can accept that INMARSAT would be able to factor out the station-keeping drift of their satellite, but to fully accept this analysis, I’d really like to see a breakdown of the data format from the satellite to ground that shows exactly how and with what precision the frequency of each received signal is measured, and communicated to the ground so that doppler shift could be calculated.
It’s free publicity for INMARSAT.
Your questions about the Inmarsat receiver's capabilities are valid and are similar to mine.
I have a friend who is a very experienced RF engineer and I've spoken with him several times about this question of the Inmarsat data. One possibility is another Inmarsat satellite with overlapping coverage had record of fl-370 engine communications.
Another possibility is that every logfile of 777 engine reports in which positional data was also reported was compared against fl-370 engine reports. Inmarsat engineers may have then been able to determine that based on speed and direction, the received signals would have certain characteristic phase (Doppler) shifts.
Also, consider that the satellite sits almost three earth diameters above the Indian Ocean.
That means only a fairly small component of the airplane's velocity would be to or from the satellite. In order for the plane's full velocity to be manifested in the Doppler shift, it would have to be at the edge of the satellite's visibility and flying either directly towards or directly away from the satellite.
A handheld GPS computes frequency shifts to about the same precision day in and day out. I use the example of a handheld to illustrate that no unusual computing power is necessary — no basketball-court sized supercomputer, just specialized off-the-shelf hardware.