Trying to read all posts, it seems that how the "arcs" were found from time data is a little deceptive for many of us. I have tried to make a diagram showing how time stamps can be converted in "arcs": time is a physical value which is measured with a very high precision (I do think it is the physical value which can be measured with the highest).
The sat is very high (35700 km) compared to the "thickness" of the flyable sky (~15 km). And the Earth is a pretty large body (~12740 km wide). From the sat, an even distance to the Earth is a circle, so, if you only know the distance of an object (flying or not), it could be anywhere on a circle. But even a small distance on the Earth surface
outside a given circle can be known by the time an electric signal uses to go to this object and back because the time could be measured
absolutly with a high precision: even for times as long as many days, you could "see" a difference of a fraction of a microsecond.
The sat is on the right (a very BIG sat), orbit and Earth the same scale. The black line around our blue planet is more than ten times too thick relatively at the highest an a/c could fly at this scale...
Hope I didn't make errors there and this could help!
EDIT: after MM43 post (
http://www.pprune.org/rumours-news/5...ml#post8395701) and remarks by MP, I will add this :
BEWARE: The purpose of this post is just to show the order of magnitude of the travel time of the signal between the sat, the a/c and back. And what could be the time difference for localisations distant of 1/360 of the Earth circumference (that is ~111km).
The 40° on the drawing is not the same 40° of "the arc" from Inmarsat data (but not very far) because the 0° is not the pole but the place from where the sat is view on the horizon.
Sorry if I have confuse somebody for that: I just wanted to show it was very possible to extract something useful from this data. For many of us, just thinking, it could be appeared largely too small according to sat distance. But it is not.