engineers were able to narrow the location of the plane down
with an initial analysis of the Doppler effect on the signal
from the flight’s pings and the plane’s approximate altitude.
How and why would they have a record of this to be able to study it?
I'm curious about that too, it seems very unlikely that the actual digitized RF waveform was available for analysis. My guess, based on working with somewhat similar systems in the past, is that the ping packet content was also tagged by the satellite with the measured carrier frequency (or offset from nominal) and this value had enough precision that it let them calculate the Doppler effect. The reason I would say this is because I've done similar things in the past for telemetry applications, it can be really handy for debugging when things aren't working right.
Inmarsat is a "bent pipe" system, meaning that the satellite just amplifies and sends down to the base station everything that it receives. So the operator's RF fanciness is limited only by the noise factor in the satellite's amplifiers.
Ah, that makes a lot more sense, though nevertheless impressive precision on the part of the base station.
However, given the corridors, it still seems like the suggestion is that the Doppler measurements said it was moving rightward or leftward, when it should only say how much the aircraft was getting closer or getting further away. Looking at the map alone, it doesn't seem like that information would help distinguish which corridor it was in.
The satellite was also moving (south or north), and the measured redshift/blueshift could be used to determine if the plane was north or south of the satellite.
Geostationary satellites don't actually stay completely stationary for any length of time without active stationkeeping. They very quickly develop a north-south motion which makes them trace out an analemma, a sort of figure eight on the ground.
I first wrote geosynchronous and approximately geostationary, then decided it was too cumbersome. What sort of magnitude does the analemma have? Also is the effect amplified by it being a "bent pipe"?
Even if it isn't intentional, in scenarios such as measuring a Doppler shift I can see how it can be a useful tool. Almost analogous to microsaccades we make with our eyes.
What is a likely way for that to work with something launched in 1996? If the plane was going 50 MPH away or toward satellite, that's ~7e-8 * c. If I remember, the frequency would be shifted by a factor of the square root of (1 - 7e-8)/(1 + 7e-8), or its inverse.
Would the satellite always relaying that level of precision to the ground? If it is, why would they have prioritized that over what would have presumably be a dramatic increase in capability if it were discarded instead? Could they have designed it to store that much precision in rotating logs available on request? And furthermore, how is a relativistic Doppler shift on those magnitudes discernible from the frequency shift naturally occurring from the plane's transmitter?
I suppose my point is that while it's easy to understand the idea of a train whistle changing pitch as it passes by, how they would do it in practice sounds like magic, being unfamiliar with the capabilities and the equipment that geostationary satellites generally carry.
Apparently it was hourly and Inmarsat calculated two arcs (north and south) which it could have taken, but then focus seems to have moved to the southern one in last few days.
I wonder if they flew test flights along both arcs and compared the data.
They didn't need to fly test flights. That would be big $$ and take time.
But apparently, their hardware is installed on plenty of airplanes, so they just needed to study normal flights north and south in the region to establish a baseline to which to compare MH370's data.
And it was that data that lead them to conclude that the southern route was likely the one MH360 actually took.
I should have been more clear -- they compared routes and aircraft velocities that could be expected to produce the same radial velocity and Doppler effects, given the satellite's geostationary position.
So the idea is that there's a shift on the intra-ping timescale? So there are numerous packets and each packet would have the frequency of that particular packet attached to it? Obviously I have no clue what the protocol for the ping looks like, just trying to understand how it would work.
Also curious how they matched it to planes flying the southern route, since I wouldn't have guessed there were any other planes flying on the southern route (which goes...nowhere).
Logged ping, base carrier frequency 1375.003762 MHz.
Logged ping, base carrier frequency 1375.003761 MHz.
Logged ping, base carrier frequency 1375.002961 MHz.
Then with some maths and comparing to the known positions of the plane (pre-contact loss) and known positions of the satellite you can work out which way the plane would have had to move to get the observed results once contact was lost.
(numbers in example made up, and likely aren't appropriate for satellite communications but you get the idea)
Well sure, but I don't know if doppler information between pings buys you much new info.
The pings, as I understand it, came an hour apart, and continued for 7-8 hours total. With just the pings themselves they were able to extract the distance from the sat, which is what led to those arcs they published last week. I dunno if doppler shifts from pings an hour apart add much to that.
Conservatively there are tens of thousands of commercial flights per day. They could use each flight as a test case to understand and refine the model of the original analysis and come up with something more accurate.
