I am only guessing
as is nearly everyone
but as the object appears to get smaller and smaller until it vanishes,
then I would say it is obliquely moving away from us.
How do you know that its trajectory wasn’t from beyond the horizon at FL350, flying on a trajectory parallel to the surface having a vector generially in the direction of the camera and then at some distance from the camara the object begins decending and the vector of the object is now directly towards the camara such that the contrail being emitted by the object becomes obscured by the object itself.
Using spherical trigonomety: how far out does an object at FL350 need to be to become visable over the horizon? What angle of decent is required for such object’s most recent contrail to coincide, i.e., overlap, its earlier emitted contrail?
When you see a broomstick held at right angle to your prespective, what happens to the apparent length of the broomstick as it is brought closer to your position? What occurs when the broomstick’s longitudinal axis becomes parallel to its motion at right angle to your position? When the broomstick is orientated in such manner, how does its apparent lenght change as it approaches your position?