Originally Posted by
MG23
Don't forget, the ground station compensates for Doppler from satellite motion, and the aircraft attempts to compensate for Doppler from aircraft motion. The former is probably very precise, since it can sync to the pilot signal from the satellite, the latter is probably not, since I suspect it just treats the satellite as located at a fixed position.
At least, I believe that's what the slides posted earlier today meant, when they said only aircraft->satellite Doppler wasn't corrected.
So you'd be looking at Doppler from satellite motion relative to the aircraft, plus any residual Doppler that the aircraft hadn't corrected for in its own motion.
If the aircraft did any compensating, it could compensate more or less to zero, since it has GPS. From the charts, it does not look like there's any compensating happening on the side of the aircraft. That's why Doppler shift jumps when it turns.
Anyhow, this is the best I could do to "reconstruct" their "reconstruction" of the south track. Scales are different, I have no idea why. Even the zero in the original chart is not a zero. (The "knee" around 20:00 UTC corresponds to the Doppler shift going through zero and changing sign as the aircraft passes the closest approach to the satellite, somewhere near the equator.) But at least qualitatively there's an agreement:
http://i61.tinypic.com/abo2n7.png This corresponds to the aircraft going off at heading 185 at 450 kts from 18:30 UTC and beyond.
I can't quite tell how they came up with whatever north route they used. There seems to be no way to draw a route that agrees with their "predicted north track" and ends up at 40 degree arc. Here's one way to look at it: their "predicted north track" has no apparent sign changes after 18:30 (we can assume that their "100 Hz" is near real zero, and the red line on the official plot does not get near 100 Hz.). Therefore, the aircraft flying along their "north track" has large consistent Doppler shifts all the way. But they began at the ~32 deg. arc at 18:30 and got only to 40 deg. by 0:11.
I can draw an approximate mirror image of the south route and get good agreement with "measured data" if I send it over Bangladesh and Nepal into south Kazakhstan.
About the only real conclusion that can be drawn here is that there was nothing remarkable happening to the aircraft after 18:30. It was not zigzagging or making any sharp turns. If it went south, I get an OK agreement simply assuming a constant 185 deg. heading (even without corrections for magnetic declination a la Cpt Kremlin above). This is not the track of an aircraft that tries to avoid detection.