Pingology.........
The more I look at the Inmarsat BFO chart the more it seems to be wrong...........almost as if the function is inverted , and the higher values towards the top of the chart are when the satellite and aircraft are approaching each other , and the pings nearer the bottom line are when they are moving apart...........e.g.........
.forgoing the maths for a moment , the aircraft was supposed to takeoff on a heading of 341 degrees for 11 nautical miles , before turning right to proceed on their planned route (latest/current departure chart), so the ping at 16.43UTC is about right for the turn at the end of climbout . So the takeoff represents initially , close to a radial arc from the satellite base position , ie , neither flying toward or away from the satellite , but the BFO value increases , so this has to mainly be from the aircaft motion vertically , climbing at an average of 2000 feet per minute .The satellite vertical motion is a much smaller proportion of this and is decreasing as it approaches its high point . So the aircraft is ascending faster than the satellite and the net effect is the distance is closing .....ie.doppler increase . After the turn the BFO increases at close to the same rate (the next ping at 16.55UTC)..... although the aircraft is now flying 032/025 degrees which includes a large component away from the satellite base position., this can only be from the vertical motion still , as the aircraft is increasing speed and continuing to climb......
Then at 17.07UTC the BFO value reduces and the only thing that has happened then , is that the aircraft has reached the top of its climb to the assigned cruising level and leveled off , ie. the vertical motion towards the satellite has stopped , and so the BFO correction value is reduced to suit.........
Then we have the large gap which looks wrong , I suspect that some part of the Satphone system has been powered off here at the same time the radios and transponder went off......
Then when the pings re-appear at 18.25UTC , we have several in quick sequence , which suggest to me that perhaps the satellite has assigned an idle transponder , or at least one where the 'default' value of max. BFO has been set , and so there is a fairly rapid adjustment in order to sync and lock on to the signal from MH370 , which it may regard as a new contact now.......this seems more likely than extreme aircraft motion ....
So now the last five pings would appear to be from motion towards the satellite instead of away from it , so it suggests moving in from outside the final ping arc............until you consider the satellite motion.........
.......... there is the unlikely symmetry in the three pings 18.29UTC , 19.40UTC , and 20.40UTC , and now this matches the relative motion of the satellite too closely to be co-incidence for me........From the data provided by STK , the satellite reaches its highest point close to 19.40UTC , then starts to descend again , so one might almost expect the chart of pings to have the same slope even if the aircraft were stationary and the BFO values were changing purely from the satellite position.....????????........ hmmmmmmmm????.......