05 March 2006

What does a train look like from high above?


Snail’s Tales’ Useless and Frivolous Information Team (STUFIT) proudly presents this photo of what undoubtedly is a train on railroad tracks (where else would a train be?) somewhere not too far from the Union Station in Washington, D.C. The red arrow I added points at the engines (there appear to be 2 of them).

Here is a closer look.


Here is a question for everyone. Is it possible to determine the direction of travel of a train from one aerial photograph only? Don't forget that the engine could push the train from behind while facing either way.

Aerial photos are from Terra Server.

In a future post, STUFIT will show you how to identify railroad tracks in aerial photos.


Duane said...

With sufficient resolution, one could see if the couplings are compressed (the engine is pushing) or in tension (the engine is pulling). In other words, look at the distance between the cars. There is a little play between the couplings. Plus, there is play in each coupling itself. I think modern couplings even have shock absorbers and springs to dampen impact on coupling, accelerating and braking. If one knew the exact length of every car and there were enough cars, one might be able to tell by the length of the train. It would be somewhat shorter being pushed than pulled.

John said...

The other way that one could probably make a reasonable guess about which direction the train is moving is if one could identify the type of train or type of tracks. In a train yard there is a greater likelihood of cars being pushed from the rear to assemble a train or position it to be pulled out of the station. A short-distance passenger train might also be pushed from behind. On ther other hand, a long-distance passenger train (like Amtrak) will almost always be pulled rather than pushed. A freight train on its route (rather than doing yard maneuvers) is also much more likely to be pulled than pushed. I am not sure why this is, but that has just been my experience from observing trains around DC.


Duane: I suppose what you are proposing could be done, although it would require a lot more info than available in this case.

John: I concur with your observations. But then again, one can never bu sure without additional information. The MARC train I take to work is pushed from behind by an engine in reverse in the mornings, but pulled in the afternoons.

deniz said...

It's frustrating to make a guess if you won't be able to tell us the answer later but I think it's moving towards the left of the photograph. I guess this in this instance (with other trains on other tracks it would be different) because there are two tracks joining each other just where the very last car is on he right and I think if the train was travelling in that direction the car would be tilted a little more toward the bottom track which it would be turning on to. Layman's guess or what...

pascal said...

First, I have to disagree with your interpretation of the engines. Locomotive engines (even ones that look rather like boxcars) have several features that make them stand out in top view:


There are three (sometimes two) round fans and several other fans/grilles that make the engine stand out from the boxcars in the same consist. Note also that the "front" of the engine is facing the cars, but the train is moving from right to left - you can't rely on the way the engine is pointing.

Based on the overpass in the larger view, I would guess that the longest, shiniest-roofed cars are likely 89-foot auto-rack cars:


However, I can think of several ways which one could assume the direction of travel, and one particular circumstance in which that direction would be unequivocal.

But, seeing as this is already a long comment, perhaps I'll post it on my own blog... :)


Deniz said:
"It's frustrating to make a guess if you won't be able to tell us the answer later..."

Well, welcome to the world of science! But, sooner or later, we'll find an answer.

I am not sure if the train would be tilted more or less towards one side depending on its direction of travel. But that's an interesting thought, though.

Pascal: If what I marked as the engines are not the engines, then where are the engines? Those 2 do have some darker areas on their tops that the other cars don't have.

No, one can't rely on the way the engine is pointing. That was also my point.

I am looking forward to your solution.

pascal said...

I'm working on an analysis now - one more question: what are the coordinates of the terraserver image? Or, more specifically, who owns the tracks in the photo?

Ain't science fun?

Joe S said...

It seems the draw gates go down at least half a minute before the train crosses, and go up within 5-12 seconds after it passes. If there are roads crossing the track, one might assume that roads that have just been crossed by the train would have draw-gates up and those ahead of the train would have draw-gates down much further down the track. this might also give an indication of the speed of the train.


Pascal: I was mistaken about the location of the tracks. They are not near the Union Station. Also I reoriented the photos (North is to the right). In the 1st photo the train is going under Franklin St., E of VA Hospital in DC.
Lon: -76.9, Lat: 38.93

Those are probably CSX tracks.

Joe S: Your idea would work if the train happened to be between 2 roads each with a draw gate.

Here's an idea from me: If a train is photographed at dusk or dawn, then its headlights will indicate which way it's going. if it's being pushed by an engine in reverse, then it will probably have headlights both in the front & in the back. But the train will be going opposite to the direction the engine is facing.