And some might say "but -komatiitebif, it's obvious that the car drove into the pole, and it's obvious the person was walking through the snow, but we don't know how long these events took to unfold."
On the contrary, we know that a brisk walk occurs at some 3 miles per hour. We know that people of certain heights have certain lengthed steps.
We can certain estimate the time it took for someone to walk a certain distance.
How about the trains?
Well, if you know what a train is made of, you can actually estimate the rate at which the train must have been moving to take on certain increments of damage. If the train were moving at 1,000 miles per hour, it would be utterly obliterated. If the train were moving 1 mile per hour, the train would take less damage.
And by using general physics, we can determine a lot about the passage of time of objects in the past, even without needing to be present to see the event.
Time is basically something relative, it is defined by the passage of events. And so long as we understand the nature of the object of the event, so too can we determine passages of time of the event itself.
If I throw a ball 20 feet in the air, I can turn my back away from that ball, and not look at it at all, and based on basic physics, I can tell exactly when that ball will hit the ground and how much time it will take without needing to be present to see it.
Old Earth geology is the same thing. We calculate rates of orogenesis, we have trace fossils like foot tracks, nests with eggs, complex burrow networks etc. And simply by understanding the objects of the events (by understanding rocks and dinosaurs and other ancient life) we can quite easily derive how long events took to unfold, even without a time machine.